HBR Analytic Services survey: 11% report revenue gains from go-to-market AI

Three out of four survey respondents say their organization used more AI for selling, marketing, customer support or revenue enablement over the past year.

The numbers come from a new Harvard Business Review Analytic Services report. In it, 75% of respondents say AI use for revenue-generating activities at their organization rose over the past year. Yet only 11% say AI has delivered increased revenue or new revenue streams. The rest of the survey tells you more than the headline gap does, because it shows where AI actually fits into a revenue team's day. That may matter more than the adoption rate to a chief revenue officer or revenue operations lead who is asked what the AI budget has returned.

Respondents at organizations using AI for revenue enablement report the most value of the four functions measured. AI is also concentrated in the work done before and after a customer conversation.

Least-used function, most reported value

The survey measured four functions. Marketing leads adoption at 54% of respondents' organizations. Sales and customer support follow at 44% each. Revenue enablement comes last at 39%. That's the content, training and coaching that prepare sellers to meet buyers.

Look at value instead and the bottom of the adoption list moves to the top. Among respondents at organizations using AI for revenue enablement, 71% report significant or moderate value, the highest of the four functions measured. The announcement doesn't give comparable value scores for the other three, so nobody outside the study knows how big that lead is.

For teams that put their first AI dollars into marketing content, the gap between 39% using AI for revenue enablement and 54% using it for marketing suggests enablement has less competition for attention.

There are good reasons to hold that loosely. Seismic sells revenue enablement software, so the most flattering result lands in the sponsor's own category. Value here is self-reported, not measured in bookings. And the 39% already using AI for enablement may be the most committed adopters in the sample, which would raise their scores no matter which function they worked in.

Meeting prep versus the live call

Timing produces the sharpest split. The study counts as AI adopters any respondents at organizations that use or are considering AI in any of the four revenue functions. Within that group, 54% say their revenue teams use real-time AI to put together meeting prep materials, and 50% use it for post-call summaries. Only 19% rely on it in the moment to suggest next-best actions to employees. Just 18% use AI agents to support revenue staff in real time.

Half of AI adopters use it to summarize a call after it ends; fewer than one in five let it suggest what to do while the call is still happening.

Based on the findings, the report says organizations have significant room to progress from preparation to AI-guided execution.

Timothy Dickson, chief digital and information officer at Regal Rexnord, said in the announcement that the company is aiming its AI spending at work that helps teams respond to customers faster and close deals sooner. Both goals can be measured inside the sales cycle itself.

Rules and confidence trail the rollout

The biggest gap in the survey has nothing to do with software. Adopters overwhelmingly say revenue teams need a strong governance framework to use AI safely and well. Far fewer are building one. Strategy is missing too: asked for the top barrier to using AI more effectively, 45% of adopters named the lack of a clear AI strategy.

Adopters picked critical thinking and judgment as the most important skill for revenue staff to get value from AI (55%), followed by knowing when to use AI and when not to (49%). Beth Tracton-Bishop, director of research at HBR Analytic Services, made the same point in the announcement. As AI moves deeper into revenue work, employees need to know when to trust what it produces and when to push back.

In a 2025 arXiv preprint, Thomas Übellacker of Maastricht University interviewed seven AI implementation experts about how people's views of AI's limitations shape an organization's readiness to adopt it. Seven interviews is a small base. In this survey's terms, governance is how one seller's sense of when to ignore the machine becomes a rule for the whole team.

What would move the 11%

The revenue figure may simply be early. A 2025 projection from the Penn Wharton Budget Model found that generative AI's contribution to annual productivity growth peaks in the early 2030s, at 0.2 percentage points in 2032. Its authors cautioned that the estimate rests on limited data about AI's early effects. That's an economy-wide model, not a sales survey. Even so, it suggests a 2026 snapshot comes well before the years when the model expects the biggest gains.

Today, 19% of AI adopters say their revenue-generating teams use AI to recommend next-best actions in the moment, and 18% use AI agents to assist revenue employees in real time. If those shares climb and the share reporting revenue gains rises with them, then the reading here holds, and the money was in the live call. If real-time use grows while revenue gains stay near one in ten, the limit is somewhere else, most likely in the strategy and governance gaps adopters already admit to.