Transactional emails are automated messages triggered by a specific action a customer takes, sent to help that person complete or follow up on something already in progress. An order receipt, a password reset confirmation, or a shipping notification are all examples of transactional emails because each one delivers information tied to an action the recipient just took.

That’s what sets transactional email apart from marketing email: a promotional campaign persuades a broad audience to take a new action, like making a purchase, while a transactional email responds to an action that’s already happened, delivering information the recipient needs rather than trying to sell them something.

Because transactional emails are expected, timely, and directly relevant to the recipient, they tend to see higher open rates than marketing emails. But that reliability only holds when they’re set up correctly: triggered at the right moment, formatted clearly, and compliant with email regulations.

This guide breaks down what transactional emails are, how they work behind the scenes, and how to set them up across the customer journey from order confirmations to renewal reminders. It also covers compliance essentials and how to monitor transactional email performance.

Table of Contents

What is a transactional email?

A transactional email is a one-to-one operational message triggered by a specific user action or system event. Some examples of transactional emails are:

• Password reset emails
• Order confirmations
• Payment-failure notices
• Suspicious-login alerts

The email system sends each transactional message to one recipient because something relevant happened to that person or account. “One-to-one” describes that event-to-recipient relationship; however, emails can still be generated automatically at scale.

Transactional emails shouldn’t be confused with bulk emails, where the sender chooses a segment and distributes substantially similar content according to a campaign schedule.

Transactional Email vs. Marketing Email: What’s the Difference?

Marketing email creates demand or prompts a new commercial action. Transactional email helps users complete, confirm, or monitor an interaction already underway. Automation and personalization can apply to both. However, the sending infrastructure for marketing and transactional emails has to be separated.

A marketing campaign can cause volume spikes, complaints, or bounces. Shared infrastructure between marketing and transactional email can let those problems linger or filter password resets, receipts, and account alerts. For this reason, it’s not uncommon for large email sender platforms like Yahoo to recommend separating marketing and transactional traffic by IP address or DKIM domain.

Separation also protects customer trust. A marketing opt-out can stop promotions without suppressing essential transactional notifications. Customers also learn that operational emails will lead with the information they need, rather than hide it beneath a sales pitch.

Featured resource: The Difference Between Marketing Email and Transactional Email

Transactional Email Examples Across the Customer Journey

The strongest transactional email examples solve one immediate customer need. They confirm access, document a payment, report progress, or surface relevant account activity. Let’s walk through real-world examples of transactional emails at each stage of the customer lifecycle, organized by type.

Account and Security Notifications

Account and security emails protect access from signup onward. Common examples include password resets, login alerts, and two-factor authentication codes. Each message should identify the event, explain the next step, and state when a link or code expires. Login alerts should also show the device, time, or approximate location. That context helps users separate a legitimate warning from phishing.

For example, HubSpot nudges all their users to protect their accounts with two-factor authentication.

Reddit’s password reset is another example of account and security notification emails. Not only does Reddit make a user follow the link to reset their password, but the company also advises what to do in case the user didn’t make that request.

Orders, Billing, and Renewals

Order, billing, and renewal emails reduce support tickets because customers don’t need to hunt through an account portal to find this information. An order confirmation records what the customer bought and what happens next. A receipt proves payment, while a failed-payment notice helps prevent service interruption.

Hospital Medicover informs patients about their paid appointments and how to prepare for the procedure.

Renewal emails are sent before the anniversary purchase date in case a user wants to cancel the subscription. Unfortunately, not all companies send renewal emails, but those that do typically include the plan name, renewal date, charge amount, and cancellation route.

Formula 1 TV, a paid monthly streaming service, lays out its renewal and cancellation terms right in your first purchase confirmation email.

Shipping and Service Updates

Shipping and service update emails cover what happens after a customer buys something. They typically include tracking updates, delay alerts, and support confirmations that keep customers informed about order status and help requests. Every message should answer three questions:

1. What happened?
2. What happens next?
3. Where can the customer get help?

Climbing equipment company Clevo provides a great example of shipping update emails. In one email, Clevo provides an order tracking link, the shipping status, and the receipt and contact information should any issues occur.

Rossmann, a retail company selling household and baby essentials, exceeded my expectations in handling order issues. Apparently, Rossmann had trouble syncing its warehouse inventory with its online store, because I ended up paying for a few items that were actually out of stock.

To resolve the issue, Rossmann sent an automatic email informing me of my incomplete order and the steps I could take: keep the rest of the order with a partial refund or cancel it altogether. Rossmann’s automatic email is a brilliant way to kill two birds with one stone: keep the customer happy while cutting costs by keeping a support agent out of the loop.

Project, App, and Community Activity

Task assignments, mentions, review requests, and community digests keep users current without requiring constant app checks. These messages qualify as transactional notifications when they report activity tied to the recipient’s account or preferences.

Digests need extra care. Product promotions can shift their purpose toward email marketing, so account activity should remain the main content. For mixed messages, the FTC’s CAN-SPAM guidance tells U.S. senders to examine the subject line and content order. Promotional framing can make the entire email commercial. Asana’s task mention is the simplest example of transactional email in this category.

How to Set Up Transactional Emails With a Connected CRM Stack

A connected CRM setup makes sure each transactional email uses the right customer record, the right template, and a clear send history. Below, we’ll cover an example of how HubSpot’s transactional emails connect CRM data, tracking, and API sending paths. With that setup, marketers will own the message while developers control the product event that triggers it.

Choose your trigger and data source.

Start by defining the exact event that should send the email.

For example, a receipt should receive an email only after a payment is confirmed as successful. Before the email goes out, make sure the payment processor confirms the charge. After that, an integration should send required details, such as the amount, currency, receipt ID, and timestamp. This helps prevent emails from being sent for failed or incomplete events. Next, decide which system owns each piece of data:

• Your app or payment system should own transaction details
• HubSpot should own customer details, like name or language
• You should also store a unique event ID and send a timestamp so you can track what happened and avoid duplicates

Pick a sending path.

HubSpot supports three transactional sending paths:

• In-app sending when the trigger and required data already live in HubSpot.
• SMTP API for system-triggered transactional email sending from internal or third-party systems.
• Single-Send API that combines HubSpot email templates with external trigger data.

Single-Send associates each email with a contact and creates one when no match exists. SMTP API can make contact creation optional. Choose SMTP API when your own application renders the email or when creating a CRM contact would add noise.

Pro tip: Evaluate any transactional email API service against three facts: event owner, template owner, and contact-creation rule. That decision prevents fragile integrations and months of CRM cleanup.

Orchestrate with workflows.

Use a workflow when HubSpot is the system that should decide when the email is sent. HubSpot workflows can handle branching, delays, record updates, and email actions in the same automation. For example, you can branch by language, plan, or region before sending the message.

Before launching, test the workflow with a real record. HubSpot’s workflow test tool lets you preview how a record would move through the flow without actually executing the actions. When you test, the path is shown in the workflow preview so you can spot logic mistakes before going live.

Note: For time-sensitive emails, such as password resets or receipts, a direct API path is often better. That reduces delay and keeps the application in control of the send.

Template, test, and version control.

First, build the transactional email in HubSpot and keep it focused on the specific event, such as a receipt, password reset, or invoice update. Next, add the key details the email needs, such as the customer name, receipt number, amount paid, or reset link. If a personalization token has no value, it can appear blank, so check important fields carefully before sending. Then, test both the message and the automation:

• Send yourself a test email.
• Use HubSpot’s WorkflowTest feature to preview how a record would move through the workflow without executing the actions.

Finally, use HubSpot’s version history to track changes to the workflow over time, and revert to a prior version if needed — note that reverts aren’t supported for workflows containing webhook or custom code actions. Keep API schemas and integration code in source control with each release tied to a named template version. This helps you trace changes, reproduce issues, and roll back safely when a deployment breaks email rendering or event processing.

Transactional Email Deliverability Checklist for Getting to the Inbox

Sinch Mailgun’s 2026 research analyzed more than 400 billion emails and found that nearly 18% failed to reach the inbox. For transactional messages, that failure can block account access, delay revenue, or generate avoidable support cases. Let’s look into the non-negotiable technical and operational steps that keep transactional emails out of spam folders and off blocklists.

Authenticate your sending domain.

Set up SPF, DKIM, and DMARC before sending production traffic.

• SPF identifies servers authorized to use your envelope domain.
• DKIM adds a domain-linked signature.
• DMARC checks whether an authenticated domain aligns with the visible From address.

Gmail requires SPF or DKIM from all senders. Senders exceeding 5,000 Gmail messages daily need SPF, DKIM, and DMARC. Gmail also requires TLS, valid forward and reverse DNS, and RFC 5322 formatting.

After publishing the records, send test messages to Gmail and Outlook. Gmail’s “Show original” view should report PASS for SPF, DKIM, and DMARC.

Review DMARC reports to identify forgotten tools or unauthorized senders. Repeat this check whenever you add new email sending services. One missing provider can break authentication for an otherwise healthy domain.

Separate critical traffic.

Send transactional notifications through a separate authenticated subdomain, “From” address, and delivery stream. High-volume programs should also consider a separate IP pool and queue. Gmail recommends different IP addresses for account notifications and promotional messages. Marketers who use HubSpot for transactional email sends can configure a dedicated IP reserved for transactional traffic right inside the ecosystem and start using it immediately.

Monitor reputation and failures.

Configure your transactional email platform to return a status event for every message. These events tell you where the email succeeded or failed. Record each event against the same message ID so support can trace the email from trigger to outcome.

Amazon Simple Email Service, or Amazon SES, is one platform that provides these events. It is an AWS cloud service that any company can connect to its website or application. It has no connection to selling products on Amazon or using Fulfillment by Amazon.

Use HubSpot’s delivery records and reporting if your company already sends through HubSpot.

Transactional Email Design That Puts the Message First

In Sinch’s 2025 survey of 3,180 shoppers, 75% expected order confirmations within five minutes. Speed gets the message into the customer’s hands. Clear design helps get the key message across at a glance.

Lead with the facts.

The subject line should identify the event and its status: “Payment received,” “Password reset requested,” or “Order confirmed.” Avoid sensitive account details in the subject.

I like when transactional emails specify the sender in the subject line. It helps me recognize valuable emails in my cluttered inbox instantly.

Inside the email, put the status, amount or date, next step, and support route above the fold. Give the customer one primary action. Secondary links should never compete with tracking an order or securing an account.

Build clarity, accessibility, and mobile capabilities.

Build accessibility into the template before adding branding. In HubSpot’s drag-and-drop email builder, start with a responsive, single-column layout. Place the status, essential details, primary action, and support link in reading order. HubSpot’s templates are responsive by default, and the editor lets you control how modules stack on mobile.

HubSpot’s image modules also include fields for adding alt text. These controls help marketers create an accessible foundation without editing HTML. You still need to write descriptive button labels, avoid color-only status indicators, and maintain 4.5:1 contrast for normal text.

Know when to add a light cross-sell.

A light cross-sell can fit after a completed, low-risk transaction. Keep it related to what the customer already bought and place it below all operational details. The subject line, preview text, and primary button should remain transactional. Gmail’s current sender guidance tells teams not to mix promotional content into sales receipt messages, so marketers need to be extremely careful if they plan to do so.

Definitely skip promotions in password resets, login alerts, failed-payment notices, refunds, and service incidents.

Transactional Email Compliance: Consent, Unsubscribes, and Data Rules

Every marketer needs to know and understand legal and regulatory requirements that apply to transactional emails.

Consent and Unsubscribe Basics

Do transactional emails need an unsubscribe link? In the U.S., a message that fits the FTC’s narrow definition of transactional or relationship email generally does not need an unsubscribe link, although its routing information must remain truthful.

The UK ICO similarly excludes purely administrative service messages from direct marketing, but adding promotional content can change that, so you’ll have to add an opt-out link.

Subject Lines and Content Placement

Write a subject line that states the customer event, such as “Payment received” or “Your renewal date changed.” Under FTC guidance, a mixed-purpose email may be treated as commercial when its subject line promotes an offer or its transactional content does not appear mainly at the beginning. Put the status, amount or date, required action, and support route first in the email template. Keep the sender name, Reply-To address, and routing details accurate as well.

Data Retention and Audit Trails

Be careful in what information you log. Don’t retain reset links, security codes, or full payment information since GDPR requires companies to delete it once it no longer serves a documented purpose. Plus, there’s a huge risk that exposed logs could give a hacker account access or sensitive financial data.

Log only the message ID, recipient, trigger, template version, send time, and delivery result.

Transactional Email KPIs: What to Measure and What to Act On

Transactional email deliverability is healthy when messages arrive quickly, reach the inbox, and help recipients complete the expected action. Measure each email type separately because a delayed password reset is more urgent than a delayed order confirmation.

Inboxing and Engagement Signals

For each transactional email type, track delivery, inboxing signals, and task completion.

• Delivery confirms that the receiving server accepted the message. Check it in your transactional email provider’s event logs.
• Inboxing signals reveal reputation or spam-filtering problems. Use Google Postmaster Tools to pinpoint these issues.
• Completion shows whether the email helped the customer verify an account or complete another task. Task completion is checked in product analytics.

Bounces, Complaints, and Blocks

A hard bounce signals a permanent failure, so suppress that address instead of sending again. Complaints show that recipients consider the message unwanted, while blocks often point to authentication, reputation, or content problems. Review these metrics by message type and mailbox provider to identify where the failure occurs.

Delivery, Latency, and Retries

Delivery, latency, and retry metrics show whether transactional emails leave your system fast and what stops them when they don’t. Find these events in your provider’s delivery logs.

• Time to send: Measure the time between the customer’s action and the provider accepting the email. If it increases, check the CRM workflow, integration, or sending queue.
• Time to deliver: Track how many messages reach the receiving server within your target time. If password resets arrive after their links expire, shorten the sending path or investigate provider delays.
• Retries: Watch how often transactional email providers temporarily reject messages. A sudden rise can indicate throttling, full inboxes, or sender-reputation trouble.
• Final failures: Act on hard bounces, rejections, and rendering errors. Suppress invalid addresses and fix missing template data or authentication before resending.

Frequently Asked Questions About Transactional Email

Do transactional emails need an unsubscribe link?

Transactional emails don’t need an unsubscribe link when the email is strictly necessary to complete or manage a requested service, such as a receipt or security alert. The U.S. and EU generally apply unsubscribe requirements to marketing rather than purely operational messages. However, definitions and exemptions vary by jurisdiction, especially under Canada’s CASL. Review the message’s purpose, content, and recipient market before removing the link.

Can I include promotions in transactional emails?

Yes, sometimes you can include promotions in a transactional email carefully. The promotion must not overtake the operational purpose. Under FTC guidance, a promotional subject line or marketing content placed before the transaction details can make the email commercial. UK guidance is stricter: promotional content may turn a service message into direct marketing. For global sends, keep cross-sells in a separate marketing email.

Should transactional emails use a dedicated IP?

Not automatically. A dedicated IP isolates transactional email reputation, but it needs steady sending volume and proper warm-up. Amazon SES recommends shared IPs for low or irregular volume and dedicated IPs for larger, predictable programs. Choose based on sending patterns.

How do I log and troubleshoot failed transactional emails?

First, store a message ID, trigger, recipient, template version, timestamps, and final status for every send. When an email fails, confirm that the trigger fired, required data reached the template, and the provider accepted the request.

Second, read the error event: retry temporary delays, suppress permanent bounces, and fix authentication or rendering errors before resending. Amazon SES event logs and HubSpot’s Single-Send API status show the type of send history a provider should expose.

What the Best Transactional Email Service Gets Right

There are two ways to define this: from the customer’s standpoint and from a technical one.

Customers never see your email stack. They see whether the password reset arrives, or whether the receipt lists the correct amount. That level of accuracy can’t happen without a robust technical setup working behind the scenes.

The best transactional email service connects the two: accurate CRM data, controlled templates, protected sending infrastructure, and searchable delivery logs, all working together so the customer-facing result stays reliable.

That’s where HubSpot’s email marketing software becomes indispensable. It keeps customer data, automation, and technical settings in one system, so marketers can control the email experience while developers control the trigger. The result is fewer handoffs and more satisfied customers.

If you’ve opened Claude’s Customize settings and weren’t sure whether a task called for a Skill, a connector, or a Plugin, you’re not alone.

The names sound similar, and people often use them interchangeably in casual conversation even though they do distinct jobs. This guide walks through what each one actually is, how they relate, and which one fits a given workflow, using a real example along the way instead of a hypothetical one.

TL;DR: Skills teach Claude a procedure, MCP connectors give it authorized access to an outside tool or data source, and Plugins package both together so a team can install everything at once.

Table of Contents

What are Claude skills, plugins, and MCP connectors?

The easiest way to keep these three straight is to separate access from procedure from distribution. Each layer answers a different question.

MCP connectors are access.

A connector links Claude to an outside service — your files, your calendar, an issue tracker, a CRM — through a live connection called an MCP server, which the service itself runs so Claude can reach it. Once you turn a connector on, Claude can read from that service and take action in it, inside your existing permissions.

Skills are procedure.

A skill is a set of written instructions, sometimes with scripts and reference files attached, that Claude loads when your request matches what the skill covers. A skill teaches the how: your team’s report format, your review checklist, the brand voice you always want applied. A skill can tell Claude how to use a connector’s data, but it can’t reach a live service on its own. That’s what the connector is for.

Plugins are distribution.

A plugin bundles skills, MCP connectors, commands, and sub-agents into a single package that installs once and can be shared with a whole team, instead of asking every person to add each piece separately.

If you want a broader primer on how plugins work as a category of software, What are plugins? covers the concept outside of Claude specifically. Inside Claude’s ecosystem, a plugin can hold any mix of the pieces above, and installing it adds everything in one step.

A short glossary, since these terms get thrown around loosely:

• Skill: written instructions Claude follows when a task matches the skill’s description.
• Plugin: an installable package of skills, connectors, commands, and sub-agents.
• MCP (Model Context Protocol): the open protocol that lets Claude talk to outside tools and services in the first place.
• MCP connector: one live connection, built on MCP, between Claude and a specific service.
• Sub-agent: a specialist Claude can delegate part of a task to; this only runs in Cowork and Claude Code, not in regular chat.

How do these three layers fit together?

Picture a request moving through Claude in three stages: a skill can shape how Claude approaches the task, a connector lets Claude reach outside the conversation to get or change real data, and a plugin is the optional wrapper that ships a skill and a connector together as one package.

None of the three sits “above” the others in a power ranking.

A connector without a skill still works fine, as Claude will simply use its own judgment about how to phrase requests to your CRM.

A skill without a connector also works fine — it just can’t touch live data, only whatever you paste into the conversation or whatever else Claude already has access to.

A plugin is only useful when you actually want several of these installed and shared as a unit; a single skill or a single connector doesn’t need one.

Which should I use: a skill, plugin, or MCP connector?

Match the problem to the layer, not the other way around:

• Use an MCP connector when you need live access to an outside tool or data source. If the task involves anything Claude can’t see just by reading the conversation — your calendar, your ticketing system, your CRM — a connector is the only layer that reaches it.
• Use a skill when you need a repeatable procedure, a template, or a brand voice applied consistently. Anything you find yourself correcting Claude on the same way every time — a report format, a QA checklist, a tone guide — is a good candidate.
• Use a plugin when you need a team-ready bundle that other people can install in one step. Once you’ve got a skill and a connector that work well together, packaging them into a plugin means your team doesn’t have to hunt down and configure each piece separately.

Combinations are common, and often the right call. A connector supplies the data, a skill decides what to do with it, and a plugin is how both reach everyone else on the team at once.

What does a combined workflow look like?

Take a weekly pipeline review, for example. On its own, none of the three layers gets you the full workflow:

A connector supplies the pipeline context. With the HubSpot Claude connector turned on, Claude can pull deal stages, recent activity, and engagement history straight from HubSpot, scoped to whatever a given user is permitted to see in HubSpot already.

A skill defines the review format. A skill can tell Claude exactly how your team likes a pipeline review structured — which fields matter, how to flag stalled deals, what counts as “needs attention” — so the output looks the same every week regardless of who asks for it.

A plugin packages it for the team. Bundling that skill with the connector into a plugin means a sales manager installs one thing, rather than separately connecting HubSpot and then hunting down the right skill, and rolling it out to a whole team is a single install instead of five.

 

This pattern is worth building toward because most teams end up connecting to more than one AI tool at once. HubSpot’s own changelog announcing the connector noted that over half of its customers already use multiple LLMs to get work done, which is part of why having your CRM context available consistently, regardless of which layer is doing the work, matters more than picking a single “winning” tool.

What are examples for marketing, sales, and service Teams?

Marketing

• Access: a connector supplying campaign performance and content-analytics context.
• Procedure: a Skill that applies your team’s brand voice to any draft, consistently.
• Distribution: a Plugin that packages both so every writer on the team gets the same setup.
• Best for: teams that produce a high volume of on-brand content and want consistency without a manual style review every time.

Sales

• Access: a connector supplying account and pipeline context.
• Procedure: a Skill that defines a standard follow-up sequence or note-taking format after a call.
• Distribution: a Plugin bundled as a “rep toolkit” so new hires get the same setup on day one.
• Best for: teams onboarding reps regularly, where you’d otherwise have to explain the same CRM habits to each new person.

Service

• Access: a connector supplying ticket and knowledge-base context.
• Procedure: a Skill that defines how tickets get triaged and routed.
• Distribution: a Plugin that packages the triage skill with the ticket connector as an agent toolkit.
• Best for: teams that field a high ticket volume and need consistent triage regardless of who’s on shift.

Pro tip: Start with the smallest layer that solves the actual problem. If a single skill already gets the format right, don’t reach for a plugin just because it feels more “official.” Package things into a plugin once you’ve confirmed the skill and connector actually work well together and more than one person needs them.

How do I set up and govern these layers?

Setup and governance are worth covering together, since the controls at each stage carry over into the next.

Start with one repeatable task rather than trying to stand up your whole stack at once. Pick a process you already do by hand every week, build or find the skill for it, connect the data source it needs, and see whether the combination actually saves time before you package anything for a wider team.

Test the skill before anyone relies on it.

Anthropic’s own skill-creator skill runs a skill against test prompts you define and reports back so you can revise the description or instructions before rolling it out.

Authorize connectors deliberately.

When you add an MCP connector, you review and approve the access it’s requesting, and you can set tool-level permissions afterward: Always Allow, Needs Approval, or Blocked, for individual actions rather than the connector as a whole.

Verify plugins before installing from outside the directory.

Anthropic runs automated validation and a security scan on every plugin version submitted to its public directory, plus a human review before a new listing goes live. That review doesn’t extend to a plugin someone uploads from an unofficial marketplace URL or a local file, so only add those from sources you actually trust.

Pilot before you roll out organization-wide.

On Team and Enterprise plans, you can share a plugin with a handful of specific people first, confirm it works the way you expect, and only then set it to install by default or make it required for everyone.

Set permissions deliberately, not just once.

The HubSpot Claude connector, for instance, respects whatever permissions already exist in HubSpot. A rep only sees the pipeline data they’re already allowed to see in HubSpot itself, so existing HubSpot permissions constrain the context available through the connector rather than the connector introducing a separate permission system.

Keep versioning and rollback in mind.

Organization admins on Team and Enterprise plans can see a plugin’s full version history and revert to an earlier version if a new one causes problems, without losing any of the version history in the process.

Know where audit logs live if you need them.

On Enterprise plans, Organization Owners can export audit log data from Organization settings and Data and Privacy, and the same data is available programmatically through Anthropic’s Compliance API for teams that need to feed it into their own compliance tooling.

None of this amounts to a security or legal guarantee. It’s a description of the controls that exist today, and you should evaluate them against your own team’s requirements rather than treating any of it as a compliance sign-off.

Frequently Asked Questions About Claude Skills, Plugins, and MCP Connectors

Do I need a developer to use Skills, Plugins, or MCP connectors?

No, not to use ones that already exist. Adding a connector from Anthropic’s directory or installing a plugin is a point-and-click process from Claude’s Customize settings. Writing your own skill from scratch is also mostly writing plain instructions in a text file — no code required unless you want to add an optional script. Building your own MCP connector or submitting a plugin to the public directory is the part that takes engineering work.

What is the difference between Claude Skills and MCP in one sentence?

A Skill teaches Claude how to do a task; MCP is the protocol that lets Claude reach outside tools and data to get the information it needs to do that task in the first place.

When should I standardize a process as a Plugin rather than a single Skill?

Once more than one person needs the same skill, or the same skill needs to travel together with a specific connector, package it as a plugin. A single skill is fine as long as you’re the only one using it and it doesn’t depend on a connector someone else still has to set up separately.

How do Claude Skills compare with tools or slash commands?

A “tool,” in Claude’s terms, is usually something a connector exposes — a specific action Claude can take in an outside service. A slash command is a named shortcut, often bundled inside a plugin, that runs a specific skill or action when you type it directly rather than describing the task in plain language. A skill is the instructions themselves; commands and tools are two different ways of triggering something.

Is there an example of a CRM-grade MCP connector I can try?

Yes. The HubSpot connector for Claude is a working example you can install today. It requires an active HubSpot account on any tier and a paid Anthropic plan (Pro, Max, Team, or Enterprise), and it can create and update records like contacts, companies, deals, and tickets, though it doesn’t delete anything on any object.

Solve with the right tool.

Skills, plugins, and MCP connectors aren’t competing tiers of the same feature. They’re three different jobs, and the workflows that actually stick tend to use more than one at a time. If you want to see the access layer in action with a real CRM rather than a hypothetical one, the HubSpot connector for Claude is a reasonable place to start, and it pairs naturally with whatever skill or plugin your team builds around it.

Summary

Why the distance between what AI can do and what GTM teams are allowed (and organized) to do is compounding in your competitors’ favor, plus five ways to close it. The real constraint isn’t model capability anymore, it’s unresolved IT, legal and platform governance plus workflows that were never built with AI in mind, and closing that gap takes deliberate ownership and documentation, not another pilot.

The volume of chatter about AI among marketing executives and practitioners keeps growing, and more and more of it is about governance and permissions. This includes mandated systems, unapproved platforms and why the tools people use (or want to use) on their own time can’t touch the systems they work in every day.

The workarounds are increasing corporate risk and making AI’s true potential for your organization farther away from reality.

I’m writing this from day one (day zero?) of MAICON 2026 (The Marketing AI Conference) in Cleveland. More than 2,000 GTM professionals are here, including ~150 CMOs at this morning’s AI for CMOs Summit (recap of my panel on 2027 budgeting here).

Throughout the day it’s been fascinating to hear how companies are embracing AI…as well as those unfortunately falling farther behind (whether they realize it or not).

Some of the people in those conversations see the problem acutely. They’ve built real workflows themselves, then a security review or a single “approved” platform shut them down. They feel every month of delay compounding, because they know what the teams who did get connected are building in the meantime.

And then there are the ones who are blissfully unaware that their competitive advantage is vanishing.

For a couple of years the honest answer to “why isn’t AI doing more for us” was that the models weren’t good enough yet. That answer stopped being true months (if not quarters) ago.

Today’s constraint has moved into three gaps: between what the models can do and what GTM teams actually do with them, between how fast the technology moves and how fast your company’s governance adapts (or doesn’t) and between how work really gets done and how well it’s designed for AI to do any of it.

Marketing is using a fraction of what’s already there

Ethan Mollick recently called the first gap the capability overhang: the distance between what current models can do and what almost anyone is doing with them. His fix is largely individual, which is the half of the problem a motivated person can solve alone on a Saturday.

The marketing numbers show how wide that gap runs. Only 28 percent of organizations in Jim Lecinski’s research with Kellogg are rethinking their workflows to accommodate AI, and just 10 percent of marketing leaders say they’re recouping major value from their investments in it. OpenAI’s own enterprise data shows the same pattern inside a single license. Among monthly active users, 19 percent have never tried data analysis, 14 percent have never used reasoning and 12 percent have never used search.

From what I see, a lot of marketing teams are still using AI mostly to draft. First drafts, variants, summaries. That’s useful, and it’s why 85 percent of marketing and product users in the same OpenAI report say they’re executing campaigns faster. But that’s also the shallow end of the pool.

The deeper capability sits one layer down. Nearly every system your team runs on already exposes an API, and a growing number ship MCP servers (the emerging standard for plugging AI assistants into other software). That includes analytics, your CRM, marketing automation platforms, data and intent sources and more. A marketer cleared to connect those systems is running analysis, debugging campaigns and building new ones against live data, while the marketer who isn’t is pasting a CSV into a chat window and hoping the LLM can read their mind.

What holds that second marketer back is usually permission. OpenAI’s data hints at how common that is: roughly one in four enterprises still hasn’t enabled the connectors that give the model secure access to company data at all.

Governance hasn’t kept up

Getting connected, not individually but for your full team or organization, runs through three corporate constraints, and marketing hits all three.

IT policy.

Connecting a model to a production system means a security review, and security reviews are built (for good reasons) to slow down exactly the kind of change you’re asking for.

Legal.

Most marketing work worth automating touches customer data. Some of the caution around that is of course well justified, but a fair amount of it is absorbed from headlines rather than from the company’s own risk profile.

The mandated platform.

Somebody chose an enterprise LLM license, usually inside a companywide software bundle your team had no vote in. That choice is now the ceiling on what your team can build.

Legal’s caution is rational enough. There’s still no comprehensive federal AI law in the United States, and the 109 AI laws enacted across 29 states in the first half of 2026 make it even harder to understand. When the rules aren’t formalized and clarified, the defensible move for company counsel is to assume the strictest version of every rule and wait. Nobody gets fired for the campaign that didn’t get built, or for the data leak that never happened.

So people route around it. UpGuard surveyed 1,500 security leaders and employees across seven countries, and more than 80 percent reported using AI tools their company hadn’t approved (including nearly 90 percent of the security professionals). Seventy percent said they were aware of colleagues inappropriately sharing sensitive data with those tools.

A CMO this morning at the MAICON Summit said every one of his employees has a second laptop they use for real AI work. That’s terrifying.

The risk is real and the policies exist for a reason, even as I’d argue those same policies have become the binding constraint on marketing’s output.

The work itself was never designed for AI

The third gap is the least visible, and it was there long before AI showed up.

Many marketing organizations quietly run on institutional knowledge: who owns which step, where the current version of the plan lives, what “done” means for a campaign and how a new request gets weighed against what’s already committed. Shockingly little of it is written down (let alone followed by all), and for a long time that was survivable. Smart people filled the gaps. They asked around, reconstructed campaign status from Slack threads, meetings and memory, and guessed right often enough that little visibly broke (at least until the team got big enough that things started to).

Now we’re asking AI to “guess” at what we want inside processes that only exist in people’s heads. It guesses with a fraction of the context the person who used to fill that gap had, and at a speed and volume that turns small inconsistencies into big ones. Unclear ownership, inconsistent planning and no shared view of what’s in flight were already costing teams before AI, and AI magnifies every one of them.

We call this orchestration debt: the cumulative burden of misaligned processes, fragile or incomplete workflows, disconnected systems and inconsistent governance across your marketing and revenue engine. Like financial debt, it accrues interest, and AI raises the rate.

Gartner found 84 percent of marketers report high levels of “collaboration drag” in cross-functional work, and organizations with high drag are 37 percent less likely to hit their revenue goals. When almost everybody is dragging, dragging starts to feel like the job.

And the people at the top are often the last to feel it. We’ve been running our marketing orchestration self-assessment with a growing number of marketing leadership teams, and a pattern keeps showing up when the CMO and their key leaders score the same operation independently. Increasingly, CMOs rate their team’s collaboration materially higher than the leaders closest to the work do. On strategic alignment and planning, leaders on the same team often land more than two points apart on a five-point scale.

I don’t think either side is wrong. The CMO sees outcomes and a team that shows up, while the leaders below see the Slack threads, meetings and memory it took to produce them. That gap matters more with AI than it ever did, because the CMO is usually the one approving the AI-forward programs, based on a view of the operation the people doing the work don’t share.

Meanwhile, the gap is compounding in your competitors’ favor

The cost of waiting shows up as work that never happened rather than as a line item that went wrong.

In OpenAI’s data, frontier firms generate about twice as many messages per seat as the median enterprise and seven times as many messages to custom GPTs, the tools teams build to standardize their own workflows. McKinsey’s latest State of AI survey found 80 percent of people using AI in their roles say it has improved their productivity, yet only 37 percent of respondents attribute any bottom-line impact to it and just 6 percent qualify as high performers.

Individuals are getting real value, and a small group of companies have figured out how to turn that into business value. If your competitor is in that group and you aren’t, every month your request sits in the queue is a month they spend building on live data you still can’t touch. And I’d argue it compounds, because every system they connect and every process they write down makes the next approval easier and the next build faster.

What connected under real governance looks like

Five things that move it faster than another pilot:

1. Pick one system. I know the pace of innovation makes picking a single LLM hard right now. But you can’t afford to wait. Picking a system with rich capabilities (and that’s more than a couple LLMs right now) connected properly beats six tools bolted onto the old process, and it gives you a reference case for the next request.
2. Bring the governance with you. “We can’t use that” almost always means “we can’t put that data there,” and those are different problems with different fixes. Walk into legal with the scope already written: which objects and fields, read or write, which named users, what gets logged and retained, what happens to the data at the vendor. A request that arrives with controls attached gets a different answer than “can we use this?”
3. Write down how the work actually moves. Marketing orchestration is a disciplined approach to how marketing work gets done, from planning all the way through execution. Before you point an agent at a process, map who owns each step, where status lives and how a new request gets weighed against what’s already committed. Then have your CMO and key leaders score the operation independently and look at where they disagree before you look at the average. (Our free marketing orchestration self-assessment uses 12 realistic workplace scenarios and takes about eight minutes, if you want a starting point.)
4. Put a number on the wait. Nobody measures what the queue costs in campaigns not run or analysis not done. Bring that number, even a rough one, and frame it against what a connected competitor is doing in the meantime. It turns the conversation from a favor you’re asking into a tradeoff the business is making.
5. Give someone the job. In a lot of orgs, nothing gets approved because getting things approved is nobody’s role. Name a person, give them the time and let them own the relationship with your legal and security teams.

Change here is genuinely hard, mentally and structurally (let alone untangling the license bundles and data agreements already baked into everything you run). And remember, your colleagues writing the restrictive policies are doing their jobs, often with rules written for an earlier era of technology.

But this conversation can’t keep getting deferred to the next planning cycle, because your competitors are already pulling ahead.

This post originally appeared on Matt Heinz’s Substack.

AI Overviews have shifted how audiences discover brands. Gone are the days when users would click through multiple search results until they found what they’re looking for. Now, AI Overviews answers questions directly on the results page, meaning brands are either cited or ignored without users ever visiting a website.

The changes brought on by AI Overviews mean marketers need to adjust the way they track brand visibility. Here’s what you need to know about tracking AI Overviews and what data actually shows.

Table of Contents

Why Tracking AI Overviews Matters Now

Traditional rank-tracking tools cannot measure the exposure generated by AI Overviews, which now sit above featured snippets as the highest-visibility position.

AI Overviews consolidate multiple sources into a single answer, making inclusion — or exclusion — a meaningful driver of both brand perception and traffic. In my experience, this shift catches marketers off guard precisely because their rankings look fine, making the problem invisible until you know where to look.

Additional reasons for tracking matters:

Volatile citation patterns — Unlike organic rankings, AI Overview citations don’t correlate reliably with ranking position. A page ranking fifth may be cited while the top result is ignored. Tracking reveals which content attributes actually earn inclusion.

Query category variation — AI Overviews appear more often for informational and how-to queries than transactional ones. Knowing which target keywords trigger an overview considerably shapes your content strategy.

Competitive intelligence — If a competitor is consistently cited for your target terms, that’s a strategic signal you can’t act on without visibility.

Measurement gaps — AI Overviews suppress clicks even when rankings hold, creating the “ranking fine, traffic dropping” paradox. Tracking helps explain it.

How Tracking AI Overviews Works From Baseline To Benchmarks

The way I think about it, tracking AI Overviews is less like checking a dashboard and more like building a case — one data point at a time. It starts with a baseline: a structured snapshot of which queries trigger AI Overviews, which cite your content, and which surface competitors. Without that starting point, there’s no meaningful way to measure change.

From that baseline, each SERP capture adds a data point that either reinforces or challenges what you recorded before. Over weeks and months, those observations become benchmarks — reference points that reflect normal performance and allow you to distinguish meaningful shifts from routine fluctuation.

That distinction is what separates reactive content decisions from informed ones. I’ve put together a five-step checklist below outlining how to put this into practice.

The 5-Step AI Overview Tracking Checklist

Step 1: Define your keyword set — Identify informational and how-to queries most relevant to your business. Prioritize keywords where you currently rank. Transactional keywords can be deprioritized for this exercise.

Step 2: Establish a baseline — Run your keyword set through your tracking tool and record the initial state: which queries trigger an AI Overview, which cite your domain, and which cite competitors. This snapshot becomes your reference point.

If you’re not sure where to start, HubSpot’s AI Search Grader is a free, one-time diagnostic that shows how your brand currently appears in AI-generated search experiences — a useful starting point before you begin logging data over time.

Step 3: Select your tracking cadence — Weekly is a practical starting point; daily is appropriate for high-priority terms or during active Google update periods.

Step 4: Log citations and mentions separately — Record whether your content receives a citation (a linked source) or a mention (a reference without a link). Keeping the distinction between linked sources and linkless references preserves data integrity over time.

Step 5: Review and act monthly — Aggregate weekly data into a monthly review. Look for patterns: a lack of citations, competitors gaining ground, or content types consistently earning inclusion. Use findings to inform content updates and gap analysis.

Tracking AI Overviews Across New Keywords And Topics

I’ve found that expanding AI Overview tracking to new keywords requires a different mindset than traditional rank tracking. A single observation — cited once, or absent once — tells you very little. What matters is the pattern across multiple data points over time.

Selecting Priority Queries

Filter your keyword set against two criteria: search intent and business relevance. Informational and how-to queries are the highest-priority candidates. From that filtered list, prioritize by volume, competitive importance, and relevance to your content pillars.

A practical tiering approach:

• Tier 1 — High-volume informational queries where you currently rank on page one. AI Overview presence or absence has the most immediate impact on visibility here.
• Tier 2 — Mid-volume queries aligned with content you’re actively producing or updating. Tracking these early measures to determine whether new content earns citations as it matures.
• Tier 3 — Emerging topics you’re monitoring for strategic interest. Establishing a baseline now positions you to act quickly if AI Overview generation increases for these terms.

Identifying AI Overviews Without Over-Interpreting Single Observations

AI Overview appearances are not static. A query triggering an overview today may not tomorrow. Treating any single check as a reliable signal produces reactive decisions based on noise.

A more reliable approach:

• Run each query multiple times before concluding. Check the same keyword across at least three to four separate sessions before treating its AI Overview status as consistent.
• Separate presence from citation. Whether an AI Overview appears and whether your content is cited within it are two distinct observations. Track them independently.
• Note the date and context of each observation. Flag data points against known Google update windows to distinguish algorithmic volatility from content evaluation changes.
• Look for directional consistency over four or more weeks before treating a change as meaningful. A one-week drop that recovers is volatility. A four-week decline warrants investigation.

How do I find AI Overviews in search for my topics?

Understanding how to trigger and interact with AI Overviews manually is the first step in tracking them effectively. The most reliable way to manually trigger AI Overviews is to phrase queries as complex questions. Instead of “best hair conditioner,” try “How do I choose the best conditioner for curly, low porosity hair?” Why and how prompts are more likely to surface an overview than short, vague queries.

Tracking UGC Sources Cited in AI Overviews

AI Overviews don’t draw exclusively from brand-owned content. Google increasingly surfaces user-generated content — Reddit threads, YouTube videos, forum discussions, and third-party reviews — particularly for queries where firsthand experience carries more credibility than brand messaging. UGC platforms are playing an increasingly important role in how AI tools discover and evaluate content.

What to Monitor and Where

• Reddit — Subreddit threads appear frequently for product comparisons, how-to queries, and category research. Monitor for threads that reference your brand or core topics, and note whether the cited sentiment is neutral, positive, or negative.
• YouTube — Video content is cited when it provides procedural or explanatory depth that text sources don’t. Track which channels and video types are surfaced for your priority informational queries.
• Forums and community platforms — Quora, niche forums, and G2 and Trustpilot discussion sections appear for industry-specific and product queries. Monitor for brand mentions and competitive framing.
• Third-party reviews — Review content from G2, Capterra, and similar platforms is cited frequently for product and software queries. A high concentration of citations from review platforms signals that peer evaluation is being treated as authoritative for that query type.

Contributing Without Crossing Into Promotion

UGC platforms have established norms around authenticity. Branded or promotional content is routinely flagged or removed, which is counterproductive to earning organic citations. Contribution should follow these guardrails:

• Participate where you have expertise. Seeding threads or responses primarily to earn citations is not appropriate.
• Disclose affiliation when relevant. Undisclosed promotional participation violates most platform policies.
• Focus on completeness and accuracy over advocacy. Cited content tends to be specific, well-structured, and directly responsive to the query.
• Do not manufacture or incentivize reviews. If fabricated reviews are surfaced in an AI Overview, reputational exposure extends beyond the review platform.

Tracking AI Overviews With Visibility And Citation KPIs

The five KPIs below provide a structured measurement framework, followed by an honest account of where attribution breaks down.

Coverage — The percentage of your tracked keyword set that triggers an AI Overview. This establishes the size of the opportunity and should be tracked over time to detect whether Google is expanding or contracting its overview generation in your category.

For teams tracking visibility across supported search engines, HubSpot’s AEO tools provide an ongoing workflow for monitoring brand presence and surfacing optimization recommendations — though coverage is engine-specific and should be applied only where it’s been verified.

Mention Frequency — How often your brand, product, or content is referenced within the AI Overview text, regardless of whether a link is present. Monitor separately from citation data to avoid conflating two distinct forms of visibility.

Citation Coverage — The percentage of tracked keywords for which your domain receives a linked citation. A gap between high mention frequency and low citation coverage may indicate your content is being paraphrased rather than sourced — worth investigating at the page level.

Cited-Page Distribution — Which specific pages are earning citations and how citation volume is distributed across your content. A heavy concentration on one or two pages signals both dependency risk and an opportunity to identify and replicate what those pages do well.

Business-Impact Proxies — Direct revenue attribution from AI Overview citations isn’t currently possible. Useful proxies include organic click-through rate trends on cited pages, changes in branded search volume, and referral traffic from domains that discuss AI Overview appearances.

These don’t confirm causation, but directional movement across multiple proxies in the same period is meaningful context.

HubSpot Smart CRM doesn’t directly attribute traffic to AI Overview citations, but it’s where downstream signals — shifts in branded contact acquisition, lead-source patterns, or pipeline influence — eventually surface. Useful for connecting AI Overview tracking to metrics that leadership recognizes, without overstating what the data can confirm.

Attribution Limits

Google Search Console does not categorize AI Overview impressions or clicks as a distinct traffic source. There is no mechanism to confirm whether a user clicked through from an AI Overview citation or a standard organic result.

Mention-level visibility generates no trackable referral signal. The metrics above are directional indicators, not closed-loop attribution — decisions made from this data should be treated as informed hypotheses.

How do I track traffic from AI Overviews?

Direct traffic attribution from AI Overviews isn’t available through any native analytics platform. GSC doesn’t segment AI Overview clicks from standard organic results, and GA4 doesn’t receive a distinct referral parameter that identifies an AI Overview as the source.

In practice, tracking requires inference and proxy measurement. Isolate pages with consistent AI Overview citations and monitor their organic traffic independently. Click-through rate compression — stable or growing impressions alongside declining clicks — is consistent with AI Overview influence, though not confirmation of it.

Segmenting by query type adds context. Comparing click-through rates between informational queries that trigger overviews and transactional queries that don’t can surface whether AI Overview-prone terms are underperforming relative to their impression volume.

Some teams track branded search volume alongside citation presence, linking AI Overview activity to a business outcome that leadership can readily understand. Demonstrating measurable impact from AI-driven visibility is an increasing priority — and proxy-based frameworks are currently the most honest way to approach it.

Tracking AI Overviews To Find Competitive Gaps

Position one doesn’t guarantee citation. A competitor ranking below you may be cited more consistently for your shared target queries. Closing that gap starts with a structured comparison across five dimensions, applied to a reproducible sample of keywords.

Building a Reproducible Sample

Select a fixed sample of 20 to 50 priority informational queries that trigger AI Overviews with reasonable frequency. Run every competitive analysis against this same sample, on the same cadence, using the same tool or capture method. Changing the sample between cycles introduces variables that make comparison unreliable.

Query Coverage — For the queries in your sample, how often does each competitor appear in an AI Overview relative to your domain? Track as a percentage of the sample per domain and update on your standard cadence to detect whether gaps are widening or narrowing.

Mention Share — How frequently each domain is referenced within the AI Overview text across the sample, with or without a link. Compare ratios rather than raw counts to avoid distortion across tracking periods.

Citation Sources — Where each competitor’s cited content lives: blog posts, product pages, documentation, or third-party sources. A competitor earning significant UGC citations signals that the category conversation is happening in spaces where brand content alone can’t dominate.

Cited Content Types — Which formats are earning citations: long-form guides, how-to articles, comparison pages, video transcripts, and data-driven research. Consistent competitor citation through a specific format is a format signal, not purely a quality signal.

Competitor Presence in Your Priority Queries — Queries where your domain is absent but a competitor is cited are your highest-priority content gaps. Examine what cited competitor pages have that yours don’t: length, structure, specificity, data, recency, and directness of answer.

Repeating this analysis monthly against the same sample transforms a one-time audit into an ongoing competitive intelligence practice.

What content gets cited most often?

If I had to whittle what earns an AI Overview citation down to one principle, it would be this: answer one question completely rather than many questions partially. Beyond that, certain attributes consistently correlate with inclusion.

Structure — Content that directly and completely resolves a specific query outperforms broad topic coverage. Step-by-step and numbered formats perform well on how-to queries. Clear headers and logical hierarchy make it easier for Google to extract and attribute specific passages.

Depth and specificity — Narrow and thorough outperform broad and shallow. Original data, proprietary research, and precise statistics earn citations because they offer something generic sources can’t replicate. Accurate definitions and technical explanations earn citations on informational queries where Google needs a reliable reference point.

Authority and experience — Content reflecting practitioner knowledge aligns with Google’s stated quality criteria. Established topical authority on your domain correlates with more consistent citation than equivalent content on domains without that track record.

Recency — Matters more for fast-moving categories. Content covering AI, marketing technology, or regulatory topics that no longer reflect current conditions tends to lose citation presence and requires active maintenance.

Content that consistently fails to earn citations includes thin pages that restate common knowledge, conversion-oriented pages where informational value is secondary, and content that requires surrounding context to be useful as a standalone answer.

HubSpot’s Marketing Hub SEO tools can help at this stage — surfacing page-level recommendations around structure, internal linking, and on-page completeness that support citation candidacy without directly measuring AI Overview performance.

Tracking AI Overviews Daily Without The Noise

Daily tracking is feasible only when structured to surface meaningful signals. Natural AI Overview variability means undisciplined daily monitoring produces more anxiety than insight.

Recommended Monitoring Cadence

• Daily — Tier 1 keywords only: highest-volume informational queries and terms tied to active campaigns or brand visibility.
• Weekly — Tier 2 keywords. Enough data points to identify directional trends without daily overhead.
• Monthly — Tier 3 exploratory terms and full competitive gap analysis.

Threshold Alerts

Establish thresholds that define when a change warrants investigation:

• A drop in citation coverage of 10% or more across Tier 1 keywords within a single week
• A competitor appearing in AI Overviews for three or more priority queries where they were previously absent
• Complete citation loss on a historically stable high-priority keyword for two consecutive tracking periods

Calibrate thresholds to your category’s baseline volatility — wider for high-fluctuation categories, tighter for stable ones.

Ownership

Assign explicit responsibility for three functions: data capture and logging, threshold alert review, and monthly synthesis and reporting. On smaller teams, these can be one person, but they should be explicitly assigned. Without clear ownership, tracking data accumulates and goes unacted upon.

Publish → Monitor → Interpret → Iterate

• Publish — Add target queries to your active tracking set immediately when content goes live or is significantly updated.
• Monitor — Capture data on your defined cadence. Record AI Overview presence, citation status, and competitor appearances. Flag changes against known Google update windows.
• Interpret — At weekly and monthly intervals, look for patterns rather than reacting to individual data points. Use your thresholds to distinguish volatility from directional change.
• Iterate — Translate interpretations into specific content hypotheses. Test them, then re-enter the monitor phase to measure whether the change produces a different citation outcome.

This cycle doesn’t guarantee citation. It builds a practice of informed experimentation that improves content quality and AI Overview visibility over time.

What To Do After Tracking AI Overviews

The hardest part of this process, in my experience, is resisting the urge to treat a tracking finding as a conclusion rather than a starting point for a hypothesis. Translation from observation to action means forming specific hypotheses about why content is or isn’t earning citations, then testing them through targeted changes. No change guarantees inclusion.

Forming Hypotheses From Findings

• Answer clarity — If your page covers a topic but doesn’t address the direct query, it may not answer the question explicitly enough. Test by adding a direct, concise answer near the top of the page before expanding into supporting detail.
• Evidence — If competitor citations consistently come from pages with original data or research and yours don’t, your content may lack evidentiary weight. Test by incorporating proprietary data, specific figures, or original research.
• Entities — If your content is topically relevant but not surfaced, your page may not clearly establish its relationship to the key entities in the query. Test by naming relevant entities explicitly and stating their relationships to your core topic directly.
• Structure — If cited competitor pages use numbered lists, defined terms, or Q&A formatting, and yours don’t, the gap may be structural. Test by reformatting to match the organizational patterns in the cited results for your priority queries.
• Distribution — Citation isn’t solely a content-quality function. If your page isn’t widely linked or referenced, it may lack the signal strength that supports citation. Test by improving internal linking, earning external references, and including the page in your active promotion workflow.

What These Hypotheses Cannot Promise

Content improvements increase citation candidacy — they don’t guarantee inclusion. AI Overview composition can change with algorithm updates, independent of anything your content does. Frame recommendations to stakeholders as improving quality and candidacy, not as a mechanism for securing placement.

How do I get my pages into AI Overviews?

There is no submission process or direct pathway for inclusion. Google determines citation sources through its own evaluation of quality, relevance, and authority.

What’s within your control when optimizing content for AI Overview is making content a stronger candidate. Answer the target query directly and completely. Use clear structure — headers, numbered steps, explicit definitions — that makes passages easy to extract. Build topical authority through depth and consistency over time. Keep content current where recency matters. Ensure pages are technically sound and fully crawlable.

External signals matter too. Content that earns links, mentions, and discussion carries stronger authority signals into Google’s evaluation — a longer-horizon factor, but meaningful for domains building citation presence from a low baseline.

Treat these as investments in content quality that serve multiple purposes — organic performance, reader experience, and citation candidacy — rather than as a checklist engineered specifically for AI Overviews. The latter leads to optimization theater; the former leads to content that is more useful and more likely to be treated as a reliable source.

Tools For Tracking AI Overviews And Where HubSpot Fits

No single platform currently provides complete AI Overview tracking coverage. Understanding what each tool category does well — and where it falls short — helps teams build a stack that’s honest about its gaps.

Where HubSpot Fits

• Website Grader — A one-time baseline diagnostic for page-level technical health. Surfaces’ crawlability issues that may limit citation candidacy, but it is not an ongoing tracking tool.
• HubSpot’s AI Search Grader — Where verified coverage exists, it provides visibility into how your brand appears in AI-generated search experiences. Apply it to confirmed use cases rather than treating it as a comprehensive AI Overview tracker.
• Marketing Hub SEO Tools — Page-level optimization recommendations supporting citation candidacy — structure, on-page completeness, internal linking — without directly measuring AI Overview performance. Most useful as a content improvement layer alongside dedicated tracking tools.
• Smart CRM — Does not track AI Overview performance directly, but provides downstream reporting context where citation influence may eventually appear in branded search behavior or contact acquisition patterns.

No HubSpot tool can replace a dedicated AI Overview tracking platform for teams that need consistent citation monitoring at scale. HubSpot contributes to the content quality, baseline assessment, and downstream reporting layers — not to the core visibility-tracking layer that purpose-built SERP monitoring tools occupy.

Frequently Asked Questions About Tracking AI Overviews

Can Google Search Console track AI Overviews?

Not directly. GSC reports on clicks, impressions, and positions for traditional organic results, but does not break out AI Overview appearances as a distinct data type.

Some indirect signals — impression spikes on informational queries without corresponding click growth — may hint at AI Overview influence, but GSC alone cannot confirm whether your content was cited. Third-party tools purpose-built for AI Overview tracking are required for that visibility.

How often should you check AI Overview visibility?

Most teams prefer weekly as a solid baseline. Daily checks may be warranted during known Google update periods or after significant content changes. For high-priority keyword sets that drive conversions or brand awareness, more frequent tracking helps catch drops before they compound into measurable traffic loss.

What is the difference between an AI Overview mention and a citation?

A mention occurs when your brand, product, or content is referenced within the AI Overview text. A citation happens when Google links directly to your page as a source. Citations carry a link and the potential for referral traffic; mentions may influence brand perception without driving any click-through. Tracking both separately gives a more complete picture of how your content is being used.

How do you track AI Overview citations over time?

The core approach involves running a defined set of keywords through a tracking tool or via manual SERP capture on a consistent cadence, recording whether an AI Overview appears, whether your domain is cited, and where in the overview your content is referenced.

Platforms like SE Ranking, BrightEdge, and Semrush have introduced AI Overview tracking features that automate much of this. For teams without those tools, a structured spreadsheet logging captures by date and keyword is a functional starting point, though it doesn’t scale.

I have been on a historical fiction kick lately. I just finished To Kill a Mockingbird and went straight into Go Set a Watchman, both by Harper Lee. Book recommendations are from my middle school and high school kids, they have some amazing picks which I have not read when I was in school, plus it gives me an opportunity to discuss the book with them, it is interesting how much you can learn from your kids.

If you haven’t read them back-to-back, here is what struck me. In the first book, Atticus Finch makes the right argument in court, and the town barely moves. Most of Maycomb isn’t evil. People are comfortable. They worry about what the neighbors will think. They protect the way things have always been.

In the second book, twenty years later, the world around Maycomb is changing fast. And the man who once stood up for what was right is now the one resisting. His world shifted, and he held tight to the version of himself that fit the old one.

 

I couldn’t stop thinking about it. Because I see the same thing playing out on marketing teams right now, with AI. The tools are ready. The business case is clear. And smart, capable people still dig in.

People resist change when it threatens who they are. For a lot of marketers, AI does exactly that. So if your team is pushing back, it’s worth asking what they’re really protecting.

The Real Problem Isn’t Technology

CMOs are buying AI tools. They’re sending the memo. They’re running lunch-and-learn. And adoption still stalls.

Why?

Listen closely to what your team says. “The output isn’t good enough.” “It takes longer to fix than to write it myself.” Sometimes that’s fair critique. Sometimes it means something else: Where do I fit now?

Your strongest writer built a career on writing. Your best analyst built one on analysis. Now a tool appears to do both in seconds. Of course, that feels personal.

Key Lesson: The fear that “AI will take my job” is a change management problem.

Don’t Drop AI Into a Broken System

Here is something we talk about a lot at Heinz Marketing. AI is an accelerator. It speeds up whatever system you put it in. If ownership is unclear, workflows are messy, and your tools don’t talk to each other, AI won’t fix that. It adds more handoffs, more rework, and more collaboration drag.

That friction confirms your team’s fears. “See? This just makes my job harder.”

Not sure where your foundation stands? Our Marketing Orchestration Self-Audit walks you through 12 real workplace scenarios in about 8 minutes. You’ll see where collaboration, roles, tools, and planning are breaking down, so you know where AI will help and where it’ll just add friction.

Key Lesson: Orchestrate first. Get clear on who owns what and how the work flows. Then add AI where it removes real bottlenecks.

5 Moves CMOs Should Make

1. Let Your People Stay the Hero

Nobody wants to hear their skills are obsolete. So don’t say it, and don’t imply it. Frame AI as the thing that takes the grunt work off their plate, so their judgment matters even more.

Key Lesson: Lead with what your team gets to do more of.

2. Start With Their Pain, Not the Tool

“Look what AI can do” is a demo. “You spend four hours every week pulling this report. What if it took twenty minutes?” is a reason to care.

Key Lesson: Find the highest-friction, highest-volume tasks first. That’s where adoption starts.

3. Build It into the Daily Work

Atticus gave one great speech, and Maycomb stayed the same. One big kickoff or one impressive demo works the same way. Change sticks through small, repeated wins inside the workflows people already use.

Key Lesson: Skip the one-time launch. Put AI inside existing workflows and let the wins stack up.

4. Make Early Adopters Visible

Peers showing peers moves faster than leaders telling. And your champions may surprise you. The people with the most invested in the old way are sometimes the slowest to move.

Key Lesson: Watch what people do and give your early adopters a stage.

5. Leave Room for Doubt

Every team hits a letdown moment. The shine wears off. AI gets something wrong. People feel let down. That’s the moment people stop idealizing the tool and start learning how to use it well.

Key Lesson: Make it safe to say “this isn’t working yet.” Teams that feel heard adopt faster than teams that feel pushed.

What a Growth Mindset Looks Like

A fear mindset says, “AI is coming for my job.” A growth mindset says, “AI frees me up for the work I’m best at.”

You can hear the difference in everyday questions. Fear asks, “What happens to me?” Growth asks, “What could I do with the time back?”

That shift comes from people trying it, getting a real win, and seeing their own value grow.

How We Approach It at Heinz Marketing

Our AI orchestration work starts with the system. We align teams, workflows, and ownership first. Then we find practical, responsible places where AI removes bottlenecks. People and process always come first and then comes the tool.

We assess and train, then test and refine. Change management and AI training are built in from day one. Because the tech is the easy part. Bringing your people with you is the real work.

Your Team’s Scout Moment

In Go Set a Watchman, Scout’s disappointment hurts. But it’s also the moment she starts thinking for herself.

Your team’s first frustration with AI can be the same kind of turning point, if you lead them through it.

How is your team responding to AI right now? And if you’re ready to bring AI into your team’s day-to-day, start with the foundation. Take the Marketing Orchestration Self-Audit. It’s free, it takes about 8 minutes, and you’ll get three quick wins you can act on this week.

Summary

After years of running marketing teams, building software for marketers, and sitting in the CMO seat, I realized I care more about making marketers’ lives better than I do about marketing itself. That’s what brought me to Heinz Marketing as President and COO.

By Jenn Steele,COO and President at Heinz Marketing

“I like marketers more than I like marketing.”

I was sitting at breakfast with a couple of fellow CMOs when I said that. They had flown up from Austin the day before to get together with some Seattle members of CMO Coffee Talk. I had just sold Kissmetrics after coming in as a turnaround CEO, but my heart belonged to marketing…or so I thought. Saying that sentence made me realize that the work itself wasn’t the point; I cared more about the people trying to do it.

I’ve come at most of my career a bit sideways. I got a biology degree from MIT and somehow wound up an IT executive at Boston law firms. Then I got into blogging and social media, burned out of IT, and in 2009 joined HubSpot as employee ~80 and an Inbound Marketing Consultant. I’ll admit that I was better at helping our customers set up their DNS than I was at helping them figure out what keywords to target at first, but I grew into it.

From there, I spent most of my career in B2B marketing: running marketing teams, sitting in the CMO seat, dealing with boards and investors, turning around and selling a company, and eventually founding a software company built for marketers. But it turns out there’s only so much you can do for marketers from inside software companies.

The problems I kept wanting to help with were the messy ones. How do you make the team function better? How do you build something the company can actually execute? How do you get better work out without adding five more tools, three consultants, and a ceremonial spreadsheet? How do you help a CMO succeed in a job where the average day can feel like twelve competing priorities wearing a trench coat? (I know. I couldn’t resist)

That’s why I kept writing and speaking, usually some version of, “Here’s how I messed this up. Please learn from it and make new and more interesting mistakes.”

Along the way, I met Matt Heinz (as do most B2B marketing execs!), attended CMO breakfasts, and got involved with CMO Coffee Talk. I partnered with Heinz Marketing, subcontracted for them when I was between roles (as most of us are at times), and hired them to build out my demand generation function when I was a CMO. I fell in love with the whole team.

So I always looked for opportunities to work with Heinz Marketing. When AI agents came on the scene as more than just hype, I jumped at the opportunity to put Heinz Marketing’s IP into agents for the consultants to use. That pulled me even closer to the team and the work. And when I heard that my friend Maria, Heinz Marketing’s longtime COO, was retiring, I made sure I kept in touch.

That created an opportunity I hadn’t been actively looking for, but that made a crazy amount of sense. Because while there’s only so much I could do for marketers at software companies, there’s quite a bit I can do for them here.

• I can work with marketing teams to bring in AI in a way that won’t suck.
• I can help our consultants solve harder operating problems with clients, not just hand over recommendations and hope everyone finds time to implement them.
• I can help CMOs find ways to extend their tenure…or help them find ways out.
• I can take more of the company off Matt’s plate so he can spend more of his time helping the B2B marketing community.

Heinz Marketing’s mission is to “Positively impact careers and lives by empowering ownership in driving results that matter,” and that’s pretty darn close to what my personal mission has become. And I already knew I loved working with the people here.

That’s why I joined Heinz Marketing as President and COO. I’ve spent a lot of my career trying to make marketing work better. Here, I get to make marketers’ lives better, too.

(Jenn is the one in the middle in the picture)