AI adoption has plateaued as a differentiator
The State of Sales 2026 surveyed 4,050 sellers across 22 countries in August and September of 2025. The headline number is no longer surprising: 87% use artificial intelligence somewhere in their sales process. And among them, 54% work with AI agents — not just assistants that answer questions, but systems that execute tasks autonomously.
The number is significant not for what it reveals about adoption, but for what it implies about the performance gap. When 87% of sellers are using the same category of tool, that tool stops being a differentiator. It becomes the floor — the starting point that no longer separates the best from the rest.
And yet, the performance gap between top performers and the average hasn't closed. In some teams, it has widened. What that suggests is that the variable that matters isn't whether the team uses AI. It's what conditions surround that AI so it can produce something different.
The gap that actually matters: data hygiene
The State of Sales 2026 gives a fairly concrete answer to that question. Among top performers, 79% prioritize data hygiene as an active part of their process. Among the general average, that number drops to 54%. A 25-percentage-point difference that, in practice, explains a great deal.
The logic behind the number isn't complicated. AI doesn't invent quality — it amplifies what it finds. If customer data is current and structured, if call notes are consistent and specific, if qualification criteria are applied uniformly across the team, AI can do more with them: identify real patterns, prioritize with precision, generate recommendations that actually make sense in context.
If the data is outdated, incomplete, or inconsistent, AI produces incomplete and inconsistent outputs faster. It doesn't improve the material — it processes it. And the faster the processing speed, the larger the scale of the original problem.
Structured and consistent call notes, qualification criteria applied uniformly across the whole team, customer history updated after every interaction.
Ad hoc or incomplete notes, qualification criteria that vary by rep or by deal, customer context that depends on someone remembering to update it.
Teams that understand this difference don't work more hours. They work with better materials. And when AI enters the picture, that difference gets amplified.
A generation without feedback or practice
The second gap has a different profile. The report cross-references data by generation and the result for Gen Z is striking: 46% report not receiving enough feedback from their managers, and 47% say they don't have access to roleplay or structured practice.
This number matters for a specific reason in the context of AI adoption. The tool can accelerate someone who already has the skills — help a good seller prepare better for meetings, identify patterns in their conversations, suggest next steps with more precision. What it can't do is replace the calibration that comes from real feedback, or the practice that comes from roleplay.
What the data describes is an onboarding and development problem that coexists with — and becomes more expensive alongside — AI adoption. Reps have the tool. They don't have the scaffolding to grow within it.
The combination creates a specific risk: teams that trust AI recommendations without having developed the judgment to question them. That use the tool as a shortcut instead of as an amplifier for a skill that's still developing. It's not a technology problem — it's a problem with how teams are developed around the technology.
The 48% who can't execute cold outreach
The third gap is more operational. Almost half of the sellers in the report — 48% — say they don't have the ability to execute cold outreach effectively.
This isn't a problem that gets solved with a tool. Cold outreach requires real understanding of the buyer's problem, personalization that goes beyond filling in template variables, and the judgment to choose the right moment, channel, and message. AI can help scale execution when that skill exists. It can't create it from scratch.
What the data describes is a pattern that keeps repeating: technology was adopted before the foundational skill was consolidated. The result is scale that amplifies ineffectiveness instead of effectiveness. More messages, worse results, more noise in buyers' inboxes — and less willingness from the market to respond to outreach in general.
AI amplifies what exists — it doesn't replace what's missing
All three gaps point to the same underlying pattern: AI as a multiplier works, but first it needs something to multiply. The problem isn't the tool — it's the conditions under which it operates.
Without data hygiene, AI processes fragmented material — it produces inconsistent outputs faster, not better ones. Speed amplifies the problem; it doesn't solve it.
Without feedback and practice, reps use AI without the judgment to evaluate its outputs — speed without calibration, confidence without foundation.
Without a foundational skill in outreach, automation scales messages that don't work — more volume, less effectiveness, more damage to the sender's reputation.
Top performers in the report aren't the ones who use AI the most. They're the ones who combine AI adoption with the habits that make it produce something different: clean data, continuous development, and the judgment to know what to automate and what not to. It's not a difference in tools — it's a difference in conditions.
What it means for sales leaders
The report doesn't say AI doesn't work. It says it works in very different ways depending on the conditions under which it operates. And those conditions — data, development, foundational skill — aren't the tool's responsibility. They're the responsibility of the team that uses it and the manager who runs it.
What that implies is concrete: the return on AI in a sales team isn't managed in the tech stack. It's managed in the quality of data the team captures in every conversation, in the frequency and consistency of feedback reps receive, and in whether they have the skill that AI is supposed to amplify. Without those three conditions, more AI produces more of the same, faster.
That's exactly what Performy runs automatically across 100% of calls: how clean and structured the information from each conversation is, what skill gaps show up consistently in each rep, and where foundational execution needs reinforcement before it's worth automating. Not on the sample someone decided to review this week — on everything that happens across the team.
