By software category · · By Gerald De Oleo

Why ChatGPT Recommends Your Competitor's Attribution Tool

The short answer: AI engines recommend the attribution and marketing analytics tools that the web describes clearly, consistently and in the buyer’s terms. In this category, most of the buyer’s guides those engines learn from are written by the vendors themselves. If your competitors wrote the guides and you’re not in them — or you’re in them under the wrong label — the model learns the category from their version of it.

This post is for founders and marketing leads at attribution, revenue analytics and marketing measurement companies. If a prospect told you “we found them in ChatGPT” about a competitor, this is why that happened, and where to start.

The shortlist is being built before your first call

Your buyer is a head of growth, a VP of marketing or a revenue ops lead. The question they used to type into Google now goes to a chatbot, with context attached: “best attribution tool for a B2B SaaS with six-month sales cycles on Salesforce.”

That shift is measurable. G2 surveyed 1,076 B2B software buyers in March 2026: 51% now start software research with an AI chatbot more often than with Google, up from 29% a year earlier. Sixty-nine percent chose a different vendor than they originally planned based on AI guidance. One in three bought from a vendor they’d never heard of before.

In attribution, those decisions carry real budget. One 2026 pricing review puts Dreamdata’s plans at roughly $12,000 to $40,000 or more per year. A single AI answer that leaves you off the shortlist can cost you a five-figure contract you never knew was in play.

“Does anyone really choose attribution software from a ChatGPT answer?”

Not from the answer alone. But the answer decides who gets evaluated. By the time your competitor’s sales team gets the demo request, the model has already done the first round of elimination — and your team was never told the round happened.

This is the hardest part of the problem to see from the inside. A lost deal in a CRM has a record. A buyer who never contacted you because ChatGPT didn’t mention you leaves nothing behind.

Who wrote the buyer’s guide

AI engines don’t invent their recommendations. They synthesize the pages that exist about a category — comparison lists, reviews, roundups, vendor pages — and the pages that rank are the pages most likely to be retrieved.

So I looked at what currently ranks for best B2B marketing attribution software. When I ran the search in September 2026, nine results came back. Six of them were published by companies that sell attribution or marketing measurement software:

The other three were TheCMO, Gartner Peer Insights and Hey Sid, a B2B marketing company.

“Isn’t that just normal content marketing?”

Yes — and it works. There’s nothing wrong with a vendor publishing a comparison list. Search results vary by person and location, so your results may look slightly different, but the pattern holds across every attribution query I checked.

The point isn’t that these lists are bad. It’s that they are the raw material. When a model answers a buyer’s question about attribution tools, it’s working from a category description that your competitors largely wrote. If you haven’t written your side of it, you’re being described by their version — or not at all.

The tool every competitor includes

Here’s the most useful thing in those nine results.

One company appears in at least eight of them: Dreamdata. It shows up even in lists written by its direct competitors.

When your competitors put you in their own buyer’s guides, you’ve won something more valuable than a ranking. You’ve won the category description. A model reading nine sources sees agreement from pages that have every reason not to agree — and agreement across independent sources is exactly the signal these systems use to decide what’s true about a market.

I don’t know Dreamdata’s internal strategy, and I’m not suggesting any one tactic produced this. But the observable pattern is consistent: a clear category — purpose-built B2B revenue attribution — described the same way by almost everyone, and attached to a specific situation, long and complex buying cycles.

Every tool gets a “best for” slot

Read the lists closely and you’ll notice they aren’t really rankings. They’re assignments.

Cometly’s guide makes this explicit. For teams already on HubSpot CRM, it points to HubSpot’s native attribution. For enterprise teams on Salesforce with complex offline touchpoints, Bizible. For sophisticated ABM programs, Dreamdata or CaliberMind. When phone calls drive conversions, Ruler Analytics. For AI-driven insights with account identification, Factors.ai.

AttributeIQ does the same thing with categories: revenue attribution for long sales cycles, paid media and campaign attribution, and advanced marketing measurement.

Every tool gets one line. If your positioning can’t be summarized in one line that a list-writer can assign to a situation, you get left out — by the writer, and then by the model that reads the list.

Notice Ruler Analytics. It owns a narrow slot — phone-call attribution — and that narrow slot gets it named in a guide written by a competitor. Specific beats comprehensive when someone else is deciding how to describe you.

“But our product does all of those things. Why would we pick one slot?”

You don’t have to limit the product. You have to give the web one sentence to describe you by.

A platform that does ABM attribution, paid media attribution and call tracking, and describes itself as “the revenue intelligence platform for modern GTM teams,” gives a list-writer nothing to assign. The competitor who says “B2B attribution for teams on Salesforce with six-month cycles” gets the slot.

Pick the situation where you’re genuinely strongest, lead with it everywhere, and let the other capabilities live one level down. I cover this in more depth in How to Get AI Recommendations, where clear category and explicit fit are the first two of five requirements.

Five reasons the model picks your competitor

In this category specifically, the gap usually comes down to some combination of these.

1. They own a slot. You describe a vision. “Revenue attribution for B2B teams with long sales cycles” can be matched to a buyer’s question. “The operating system for data-driven marketing” can’t.

2. Their integrations are explicit. Yours are a logo wall. Buyers prompt with their stack. The questions look like “BI tool that connects to Snowflake and Salesforce” — LLM Pulse lists integration questions like this as one of the main ways SaaS buyers query AI engines. Fibbler’s own description is a good example of doing this well: it names LinkedIn Ads and Google Ads on one side, HubSpot, Salesforce, Attio and Pipedrive on the other, and states that it’s an official LinkedIn Marketing Partner. Every one of those is a checkable fact a model can match to a question. A row of logos on your homepage isn’t.

3. They give pricing context. You say “contact sales.” Buyers ask about cost at their scale. The same LLM Pulse guide lists questions like “total cost of Mixpanel vs Amplitude at 10 million events.” When a buyer mentions a budget and your pricing page offers nothing, the model has no way to include you in a price-constrained answer.

4. They appear in lists they didn’t write. You only appear in your own. Your own comparison list helps the model find you. Being named in someone else’s is what makes the model believe you.

5. Their description is consistent. Yours changed. Attribution companies reposition often — from “marketing analytics” to “revenue attribution” to “GTM intelligence.” Each rename leaves a trail of G2 descriptions, directory profiles and old reviews using the previous label. The old descriptions don’t expire. They keep competing with the new one, and they often win on volume.

The broad-query trap

There’s a strategic mistake worth naming before you fix anything.

Broad category questions default to incumbents. An AI answer captured in August 2026 for “best growth marketing platforms and tools for B2B SaaS” led with HubSpot Marketing Hub, ActiveCampaign, Adobe Marketo Engage and Salesforce’s marketing automation. That’s not surprising. On a question that broad, the model reaches for the names with the most evidence behind them.

A specialist attribution tool will rarely win that question. But buyers don’t only ask broad questions. They ask the constrained ones — with their CRM, their sales cycle, their ad channels and their budget attached. That’s where a specialist can win, because the incumbents’ generic descriptions don’t match the specifics.

“So should we stop trying to appear for ‘best marketing attribution software’?”

No — but measure it separately, and don’t build your strategy around it. The broad query is where you’ll see the least movement, and the most expensive competition.

Build for the questions where your specific fit is the answer. “Attribution for LinkedIn Ads into HubSpot pipeline.” “Multi-touch attribution for a 20-person B2B marketing team.” “Salesforce-native attribution for six-month sales cycles.” Those are narrower, and they’re the ones buyers ask right before choosing.

Test it yourself this week

You don’t need a tool to see where you stand. You need about an hour.

1. Write 10–15 questions in your buyer’s words. Not your homepage’s words — the language your prospects use on discovery calls. Include constraints: CRM, sales cycle length, ad channels, team size, budget. For example:

Never include your company name. Asking “is [your company] good?” introduces the brand artificially and tells you nothing.

2. Run them in ChatGPT, Perplexity, Gemini and Google’s AI Mode. The engines draw from different sources, so track them separately.

3. Log five things for every answer. Were you recommended? Where in the shortlist? For the right use case? Described in the right category? And who was recommended instead?

4. Open the sources. For the answers you’re missing from, look at which pages the engine cited. Those are the pages shaping the answer you should own — and in this category, several of them will be your competitors’ buyer’s guides.

“Can’t we just publish our own ‘best attribution tools’ list with us at number one?”

You can, and most of your competitors already have. It helps retrieval — it gives the model a page about your category that mentions you.

But as evidence, it’s weak, because every vendor’s list puts the vendor first. A model reading six lists that each crown their author learns very little from any of them. What moves recommendations is appearing in lists you didn’t write.

If you publish comparison content, make it honest: name the situations where a competitor is the better choice. Models give more weight to comparisons that admit trade-offs than to ones that win on every row.

What to fix first

The order matters. Each step makes the next one worth more.

1. One sentence, everywhere. Decide the situation you’re best for, write it in one line, and make it identical on your homepage, your G2 profile, Gartner Peer Insights, LinkedIn and every directory listing. Fix the old descriptions — especially after a rename.

2. One page per integration that matters. A dedicated page for each major CRM, ad platform and data warehouse you connect to, stating exactly what the integration does. This answers the stack-constrained questions directly.

3. Pricing context. Even without full public pricing, give ranges, what drives the price and who each plan is for. Something the model can reason with when a buyer mentions a budget.

4. Honest comparison pages. One per real competitor, stating when each is the better choice.

5. Get into lists you didn’t write. Independent reviewers, analyst peer-review platforms, category pages on G2, and industry publications. This is the slowest step and the one with the biggest effect on whether a model believes your description.

6. Measure monthly. Rerun the same questions, per engine, and watch the five columns. When something moves, trace it back to the source that moved it.

It’s worth reading AEO vs SEO for why ranking alone doesn’t produce any of this, and AEO vs GEO for why the engines don’t agree with each other.

Quick answers

Why doesn’t ChatGPT recommend my attribution tool? Usually because the web doesn’t describe you clearly or consistently enough for the model to match you to the buyer’s question. In attribution, the most common causes are a vague category, integrations that aren’t stated explicitly, no pricing context, absence from third-party comparison lists, and old descriptions left over from a repositioning.

Which attribution tools show up most in the sources AI engines draw from? In the comparison lists ranking in September 2026, Dreamdata, HubSpot’s native attribution, Adobe Marketo Measure (Bizible), Factors.ai, HockeyStack and Cometly recur most often. Which ones a given engine recommends varies by the exact question and the engine.

Does publishing our own “best tools” list help? It helps a model find you, but it’s weak evidence on its own, because every vendor’s list tends to rank its author first. Being named in lists you didn’t write matters more.

How do I check what AI says about my attribution tool? Run 10–15 unbranded buyer questions with real constraints across ChatGPT, Perplexity, Gemini and Google’s AI Mode. Log whether you’re recommended, your position, whether the use case and description are right, and who appears instead.

How long does it take to start appearing? It depends on the engine and on how much outdated information exists about you. Fixes to your own site show up sooner in engines that search the web on every question. Changes in how third parties describe you take longer. Nobody can honestly promise a date for a specific engine.

I run the audit for attribution and marketing analytics companies: the unbranded questions your buyers actually ask, across every major engine, with the specific pages shaping each answer. Before we talk, I run a free Shortlist Check — three or four of those questions in one engine — so the call starts with something real rather than questions. Book a call.