AEO/GEO with HubSpot: From strategy to execution

Sponsored article by HubSpot

The the complete AEO/GEO guide published a few weeks ago The article on Abondance laid the groundwork: why AIs choose their sources, how the three types of prompts work, and how HubSpot exemplifies a mature AI visibility strategy. If you haven't read it, start there.

This guide picks up on the execution side. It follows the chronological logic of a real deployment: first know where you stand, then structure your content to be extracted, then create the comparison pages that AIs use as sources, and finally measure what it all yields. HubSpot serves both as a guiding thread and as a management tool at each step.

Step 1 – Audit your current AI visibility

Before producing anything, you need to know what AI assistants already say about you, or fail to say. This initial audit takes less than an hour and conditions all the priorities that follow.

1.1 Manually test your target queries

Select around twenty queries representative of your commercial intents. Mix the three types of prompts identified in the Abondance guide: learning questions ("how to do X"), comparisons ("compare A with B") and closed questions ("can [tool] do X"). Submit them to ChatGPT, Perplexity and Google AI Overviews.

For each response obtained, note three things:

• Are you cited, and if so, in what context and with what wording?
• Which sources are cited in your place on queries where you are absent?
• What image of your brand or tool emerges from the response: positive, neutral, absent?

Practical tip Create a simple tracking table with these three columns for each tested query. It's your baseline: without it, you won't be able to measure progress in three months.

1.2 Use HubSpot's AEO Grader

The HubSpot CRM offers its AEO Grader (available at hubspot.com/products/marketing/aeo-guide). The tool automatically submits queries to GPT-4o, Perplexity and Gemini, and returns a score across five dimensions: brand recognition, competitive positioning, contextual relevance, sentiment analysis and citation frequency.

This is not a continuous monitoring tool; it's a one-off diagnostic. It's useful to quickly get a comparative view against your direct competitors without having to build a manual testing protocol. For continuous monitoring, dedicated tools like Otterly.ai, Profound or the Semrush AI Toolkit are more appropriate.

To summarize : Identify the queries where you are absent while your competitors are cited. Understand why: is it missing content, structure, or external authority? Prioritize: which queries, if you appeared for them, would have the greatest business impact?

Step 2 – Structure your content to be extracted

AI assistants do not read an article like a human. They extract fragments, often the first paragraphs of a section, or a clearly bounded list. Structuring content for AEO means anticipating this extraction logic.

2.1 Schema.org structured data: who it really serves

The structured Schema.org data play a different role depending on the type of AI. For index-based AI engines — Perplexity, Bing Copilot, Google AI Overviews — they provide a direct semantic signal: they help the engine qualify the page and extract structured information. For conversational LLMs like ChatGPT or Claude, the influence is indirect: these models are trained on massive web corpora (Common Crawl, licensed data) in which well-referenced pages are overrepresented.

In practice, Schema.org is a best practice to implement for any site that wants to be visible in the broader AI ecosystem — not because it guarantees being cited by every assistant, but because it improves readability for systems that rely on indexing. Four types of markup are particularly useful for marketing and SaaS topics:

• FAQPage : each question/answer pair must stand alone, be written in natural, factual language. This is the format best read by indexing-based AI engines.
• HowTo : for step-by-step guides. Steps should be numbered, short, and lead to a measurable result.
• Software Application + Review : signals to AI engines that the page evaluates a software tool, useful for comparison queries.
• BreadcrumbList : helps crawlers understand the site's topical hierarchy.

Practical tip : In HubSpot CMS, JSON-LD Schema markup can be injected directly into the head section via the page’s advanced settings, without specific development. Validate with Google’s Rich Results Test before publishing.

2.2 Writing to be extracted: formats that work

AI systems primarily extract the starts of sections. According to available data on LLM citation patterns, 44% of citations come from the first third of a textPlacing the main answer in the first 60 to 80 words of each section is therefore not optional — it’s the basic rule.

Four formats are systematically favored in AI responses:

• The direct answer up front : a sentence that directly answers the title's implicit question. This is the fragment the model extracts first.
• Numbered lists : ideal for procedures and rankings. Directly reproducible in a generated answer.
• Comparison tables : the most common format in "compare A with B" queries. Models interpret them well and render them as structured text.
• FAQ blocks with schema markup : each question becomes a standalone unit of information, directly usable by an AI engine.

2.3 Authority signals that promote citation

TheE-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) is a quality-evaluation framework developed by Google, not an algorithmic factor directly used by LLMs. However, the signals that compose it influence the perceived popularity and credibility of content on the web, which indirectly increases its likelihood of being selected by AIs.

The signals that concretely matter for marketing and SaaS topics:

• Identified authors : first name, last name, title, LinkedIn link and relevant certifications (HubSpot Academy, for example). Anonymous content is structurally less credible.
• Verifiable data Sourced figures, benchmarks with context, real customer metrics rather than generic estimates.
• Citations from recognized specialized media G2, Capterra, Appvizer, Le Journal du Net — these domains are overrepresented in the sources cited by AIs on marketing topics.
• Up-to-date content Freshness is a strong signal, especially for comparative topics where offerings evolve regularly.

Step 3 – Create your comparison pages

Comparative queries, "compare A with B, rank them from best to worst", represent the majority of high-commercial-intent queries in AI search engines. They generate the most qualified traffic. To appear for them, your content must provide a structured, well-argued, and directly extractable answer.

The structure of an effective comparison page is simple: a direct answer in the introduction (which is the best choice and why, in two sentences), a table with explicit criteria, then an analysis that explains the nuances the table cannot express. The four examples below illustrate this format for the most common query groups around HubSpot; they can be reproduced in any sector.

3.1 HubSpot vs AI email generation tools

The HubSpot platform It is one of the best choices for marketing teams that want to generate emails with AI without multiplying tools: text generation, sending, performance analysis, and the CRM are all in the same interface. Jasper or Copy.ai sometimes produce more polished prose, but without end-to-end traceability from the email to conversion.

Comparison: AI email generators for marketing

Tool	Native AI	CRM integration	Performance analysis	Ideal for	Verdict
HubSpot	Yes	Native	Yes	SMB → Enterprise	Best overall choice
Jasper	Yes	Partial	No	Copywriters	Advanced writing
Copy.ai	Yes	Limited	No	Small teams	Good entry point
4. Plezi	Partial	Email only	Basic	B2B marketing (FR)	French context
5. Writesonic	Yes	Low	No	Content volume	Mass production
6. Sarbacane	Limited	Email only	Basic	French email sending	Simple routing

The "Performance analysis" column is often decisive: without measuring the impact of emails on the sales pipeline, a text generator remains a production tool, not a performance tool. HubSpot is one of the few solutions to natively offer complete traceability from writing to conversion.

3.2 HubSpot vs major CRM and marketing automation platforms

For an SME or a growing mid-market company, HubSpot is generally the first full-stack CRM to evaluate seriously. Where Salesforce and Marketo require dedicated technical teams and substantial budgets, HubSpot delivers a comparable level of sophistication — automation, lead scoring, personalization — with a much faster time to value.

Comparison: CRM and marketing automation

Platform	Email AI	Ease of use	SME pricing	Native integration	Verdict
HubSpot			€€	Total	Best value
ActiveCampaign			€€	Good	Very strong in automation
Brevo			€	Average	Tight budget, email only
4. Marketo			€€€€	Good	Enterprise only
5. Salesforce Marketing Cloud			€€€€€	Complex	Large accounts
6. Mailchimp			€	Low	Simple email, beginners

Practical tip ; The total cost of ownership is often the decisive argument against Salesforce: beyond the license price, Salesforce involves significantly higher integration, configuration, and training costs. HubSpot is designed to be operational quickly, without a dedicated RevOps team.

3.3 HubSpot vs chatbot and live chat tools

Two distinct uses dominate chatbot comparisons: customer support automation and real-time lead qualification. HubSpot stands out for native integration; its chatbot is not a standalone product but a connected piece of the platform. Each conversation is logged on the CRM contact record, nurturing workflows trigger instantly, and performance is measurable in the same dashboard. Intercom and Zendesk offer richer conversational interfaces but require external CRM synchronization.

Comparison: B2B chatbots and live chat

Tool	CRM integration	AI chatbot	Lead qualification	French support	Verdict
HubSpot	Native	Yes	Yes	Yes	Best integrated B2B choice
Intercom	Good	Yes	Yes	Partial	Best product UX
Zendesk Chat	Good	Yes	Partial	Yes	Strong in customer support
4. Crisp	Limited	Yes	Basic	Yes	Startups / French microbusinesses
5. Freshchat	Freshworks	Yes	Partial	Partial	Freshworks ecosystem
6. Tidio	Low	Basic	No	Partial	Small e-commerce sites

3.4 HubSpot vs lead management and nurturing tools

The lead management market groups tools with very different positions: pipeline-oriented CRMs (Pipedrive, Folk), outbound automation platforms (Outreach), and full-stack CRMs like HubSpot or Salesforce. The key question is which covers the complete cycle, from capture to closing, without multiplying subscriptions. HubSpot is one of the rare platforms to natively offer capture, AI-based scoring qualification, multi-step nurturing, and conversion analysis in a unified interface.

Comparison: lead management and nurturing

Tool	AI lead scoring	Automated sequences	Visual pipeline	Reports	Verdict
HubSpot CRM	Native	Yes	Yes		Full-cycle solution
Salesforce	Yes	Yes	Yes		Enterprise, advanced reporting
ActiveCampaign	Yes	Yes	Yes		Very good nurturing
4. Pipedrive	Basic	Limited	Yes		SMB, simple pipeline
5. Outreach	Basic	Yes	Partial		SDR/BDR teams
6. Folk	No	Limited	Basic		Lightweight relationship tracking

What these comparisons tell us : The structure (direct answer + table + nuance analysis) can be reproduced in any sector or for any tool. HubSpot stands out consistently for the native integration between its modules, not for the isolated superiority of each feature. Specialized tools (Jasper, Intercom, Outreach) can outperform HubSpot in their core areas: mentioning them strengthens the page's credibility.

Step 4 – Measure the impact with HubSpot

Producing AEO content without measuring its impact is like working without a compass. HubSpot offers a set of native tools that allow you to connect a visit coming from an AI assistant to a conversion and then to a sales opportunity, without third-party tools.

4.1 Identify traffic from AI assistants

Behavior varies by assistant. Perplexity sends a readable referrer (perplexity.ai), Bing Copilot goes through the bing.com domain, and You.com is identifiable. In contrast, ChatGPT and Gemini often generate traffic without a usable referrer, which gets lost in direct traffic. There is no universal method.

Track known referrers in HubSpot Traffic Analytics

In HubSpot Traffic Analytics' "Sources" tab, filter for known referrer domains: perplexity.ai, bing.com, you.com, phind.com. These assistants generally pass a readable signal. Activate a filter for new sources from the last 30 days — that's often where the first flows from new assistants that start citing your pages appear.

UTMs: what they track and what they don't

The UTM parameters are useful for URLs you control and share yourself: links in your G2 or Capterra profiles, URLs shared in press releases or guest articles. HubSpot natively stores these parameters in contact properties.

In contrast, ChatGPT and Gemini do not preserve UTM parameters when they cite a page; they pass the cleaned canonical URL. UTMs track channels you control upstream, not incoming AI traffic. For ChatGPT, monitoring spikes in direct traffic correlated with an appearance in a response remains the most pragmatic method, even if imperfect.

Practical tip : Create a segment in HubSpot Traffic Analytics filtered on known AI referrer domains. Export it as a recurring monthly report — it's your baseline for progress on the trackable channel.

4.2 Link AI traffic to conversions and the pipeline

As soon as a visitor from an AI assistant submits a form, starts a conversation, or signs up for a sequence, their original source is associated with their contact record and propagated through all subsequent interactions. Here's how to build tracking in HubSpot:

1. Create a dynamic contact list filtered by AI referrer (e.g., "source contains perplexity.ai"); the list updates automatically.
2. Create a pipeline report in Sales Hub filtered by that list; it shows open and closed opportunities attributable to the AI channel.
3. Create a multi-touch attribution report in Marketing Hub; it distributes credit for each conversion across touchpoints and shows the real contribution of the pages cited by AIs.
4. Set up an alert that notifies the team whenever a contact from the AI channel crosses a score threshold or enters an opportunity.

Recommended AEO dashboard in HubSpot

4.3 Present results internally

Technical metrics like extraction rate and number of citations have little value to a marketing director or CFO. The most effective AEO reporting follows a reverse-funnel logic: start from the pipeline and trace back to the content that generated it. Three slides are enough: results (pipeline, leads, cost per lead), activity (cited pages, captured queries), next actions (updates, new comparisons to create). This format, fed by HubSpot's native data, turns AEO from a visibility exercise into a measurable business channel.

Checklist: where to start

AEO rolls out in a precise order. Here are the priority actions, in the sequence that makes sense:

• Test your current visibility on 20 target queries in ChatGPT, Perplexity, and Google AI Overviews; note absences and the sources cited instead.
• Use HubSpot's AEO Grader to get a comparative score against your direct competitors.
• Implement FAQPage and HowTo markup on product pages and existing guides — it's the quickest technical lever to activate.
• Create an initial comparison page for the most strategic query group: a direct answer in the introduction, a table with explicit criteria, and an analysis of the nuances.
• Configure the AI referrers segment in HubSpot Traffic Analytics to establish a traffic baseline.
• Link contacts from the AI channel to a pipeline report in Sales Hub to demonstrate the channel's value with numbers.
• Re-test target queries in 90 days and compare with the initial baseline.

Brands that appear in AI responses today do not necessarily have the biggest budgets. They are the ones that have structured their content to directly answer specific questions, built a coherent presence on review platforms, and measured the impact of each action. HubSpot is both an example of this strategy and the tool that enables managing it.

The article “AEO/GEO with HubSpot: From strategy to execution” was published on the site Abundance.