The machines crowned HubSpot, not Salesforce.

Who the machines recommend

1Finding 1: The machines crowned HubSpot, not Salesforce

Salesforce is the biggest CRM company in the world by a wide margin. In the machines’ answers it wins three situations, best overall, enterprise and B2B pipeline, and loses everywhere a smaller company shops, where HubSpot, Pipedrive and Zoho split the field. For every four mentions HubSpot gets, Salesforce gets three. Market share and machine share are different leaderboards, and a growth plan that tracks only one of them is flying half blind.

2Finding 2: A giant can be invisible

Microsoft Dynamics earns more CRM revenue than everyone below Salesforce, and when a small business, a startup or a budget-conscious buyer asks the machines for a CRM, it barely exists: 6 percent of answers for a small business’s first CRM, 7 percent for a fast-growing startup, 2 percent on a tight budget. It appears at full strength in enterprise, where it matches Salesforce answer for answer, and in 57 percent of answers naming the best CRM overall. Deliberate position or slow leak of tomorrow’s mid-market pipeline: that is Microsoft’s call. The numbers just say the machines have already made theirs.

3Finding 3: Who owns which buying situation

Zoho owns the budget question outright. Attio surfaces almost exclusively for startups, proof that a young brand can own a situation without owning the category. This table, not the overall average, is the map a CRM marketing team should plan against, and none of it is visible in a single score.

4Finding 4: Search visibility can be rented. Memory has to be earned.

The most instructive brand in the category is monday CRM. In the machines’ memory it barely registers: 3 percent, three situations. Turn search on and its presence nearly doubles and spreads to seven situations, fed by its own pages and comparison content, sometimes read back to buyers nearly word for word. Freshsales shows the same pattern in milder form.

Both halves matter for planning. Search-fed presence responds to content operations and moves in months; it is real, and it is rented, because it lasts as long as the publishing does. Memory is the compounding asset: what the machine says when it looks nothing up, shifting slowly, the way reputations do. A brand strong in search and weak in memory should know which one it is buying with each euro. A brand strong in both, like HubSpot, owns its position. The monday CRM sample audit shows this gap for one company.

5Finding 5: The websites that feed the machines’ answers

When assistants search before answering, they cite sources. The single biggest source of AI answers about CRM is the vendors’ own websites: hubspot.com (93 citations), pipedrive.com (64), salesforce.com (61), zoho.com (60). The tech press trails far behind.

Two details are worth a board slide. What a brand publishes on its own domain is now answered back to buyers by machines, which makes those pages a sales channel, not a formality. And two small vendors’ comparison articles, from lessannoyingcrm.com and sybill.ai, are cited more often than Forbes when machines answer questions about the whole category. Writing the page that answers the buyer’s question is a seat at the table, and it is demonstrably available to small players.

A recorded answer, in context

One answer, verbatim, from the archive behind this report. The study’s results come from repeated questions; a single answer shows the format buyers see.

“What CRM should a small business start with?”

Recorded test question

For most small businesses, I’d start with HubSpot CRM.

• Best overall starter CRM: HubSpot CRM
• Best for sales-focused teams: Pipedrive
• Best budget-friendly all-in-one option: Zoho CRM

ChatGPT via the OpenAI API, from memory, August 2026. Opening excerpt, not the entire response.

What we would do about it

A taste of the advisory layer:

• If you are not HubSpot: stop fighting the average. Zoho owns budget, Attio owns startups; the winnable unit in this channel is the situation, and the table above prices every door.
• If you are Dynamics-shaped: decide whether enterprise-only is your strategy or your neglect. If it is neglect, mid-market memory takes quarters to earn and starts with the sources that feed those situations, not with advertising.
• If you are renting visibility like monday CRM: it works, keep paying for it, and know the lease terms. Pair the content engine with the slower memory work, or accept that the position resets whenever the publishing stops.
• If you are small: one comparison page that honestly answers a buying question can out-cite Forbes here. That is the cheapest observed entry ticket in this report.

The paid version is specific to you: your numbers per situation, the false claims AI makes about you, the exact sources to earn, priorities with owners and effort grades, and the tactics to skip. That is the audit, and the Program that carries the work out. To see the full depth on a real brand, read the sample audit of monday CRM built from our CRM dataset.

How we measured

ChatGPT, Gemini and Claude, queried through their official developer interfaces in August 2026, in the model tiers most buyers actually use. 1,007 recorded answers across eight buying situations, in two modes: from memory and with web search, reported separately. The design was frozen and hashed before the first question; measurement stability r = 0.93, and differences under 15 points per situation are noise, treated as such. These are controlled tests, not private customer conversations, and results can differ from the consumer apps.

Definitions, limits and the pre-registration protocol: methodology v0.2. The brands measured did not participate in, commission, or endorse this report, and we never sell placement in AI answers. Commissioned audits never change published scores.