Why teams ignore AI insights
AI can now tell a team what to change. Often nobody has been given the job of deciding whether to.
By Nathan Barling Product Marketing Principal 5 min read
Figure 1
Many say AI helps them. Fewer say it reaches company earnings
Share of respondents agreeing, 2026
Bar chart. 80% of respondents say AI improved their own productivity; 37% say it contributed to company earnings.
On Monday morning, the model shows that one channel is past the point of diminishing returns and another is underspent. The fix looks obvious. But those channels are run by different teams, and moving budget between them goes through a review that happens once a quarter. So the insight gets logged and the money stays where it is.
The same week, a planner overrides a different recommendation because she doesn't trust the number behind it. She might be right. She might be wrong. Nobody writes down that she did it, so nobody follows up, and nobody finds out which. Next month the model says the same thing, and it goes the same way.
The shape will be familiar.
An AI insight is a request for a decision. Too often, nobody has decided who answers it. So it sits, or it gets skipped, and nothing records that it happened.
By an insight, we mean any time an AI system tells a team to change a price, a segment, a bid, or a budget. It's often right. What happens next depends on who can act, and whether the people around them agree on what the model is for. That's the core of decision intelligence: the decision matters as much as the data behind it.
What the numbers say, and what they can't
No survey we found measures how often teams ignore AI insights. The closest evidence is indirect.
McKinsey's 2026 survey of 1,719 respondents found nearly nine in ten organizations using AI regularly, and about 80% saying it improved their own productivity. Only 37% said it contributed to company earnings, about the same share as a year earlier. Grant Thornton's April survey of nearly 1,000 senior US leaders points the same way. Only 52% of boards have set clear AI governance expectations, and 22% of operations leaders say they have a fully developed AI strategy.
Usage isn't impact. A tool can be used every day and still change no decision. The sharpest signal comes from the lead handoff, where AI scoring already lives.
In a September 2026 survey of 192 practitioners, fielded by Heinz Marketing for Docket (which sells AI sales agents), 78% said sales rejects more than half of marketing qualified leads. Nearly 90% still report lead volume to leadership. Only 7% of those are confident it predicts pipeline.
Figure 2
The lead handoff, 100 respondents at a time
Each square is one organization or respondent, scaled to 100
Three 100-square grids. 78 of 100 organizations say sales rejects more than half of marketing qualified leads. About 90 of 100 respondents report lead volume to leadership. 7 of every 100 who report it are confident that volume predicts pipeline.
A more precise lead score doesn't settle what counts as qualified.
What a model forces teams to decide
Teams can work around a missing definition for years. Marketing counts one kind of customer, sales counts another, and the difference gets settled in meetings. Nobody writes it down.
A model can't work around it. To optimize anything, it needs one definition of a customer, one definition of a good lead, and one number to maximize. Writing those down is what context engineering and a semantic layer are for. For many teams, it's the first time anyone has had to.
It goes wrong in a few ways. The definition is missing, or two teams define a qualified lead differently. The model chases one number while the people deciding care about another. Or the person accountable for a result can't change the price, the segment, or the budget behind it. That last one is the hardest to see, because everyone involved is doing their job.
Insights also leave a record. A disagreement over budget used to happen in a meeting and leave nothing behind. A logged insight has a timestamp and an outcome, even when the outcome is that nobody acted.
That may explain part of the distance between McKinsey's numbers. A person can adopt a tool without asking anyone. Changing a price, segment, or budget takes several functions agreeing.
Diagram. Adopting a tool, one person decides. Changing a price, segment, or budget, marketing, finance, and sales have to agree.
Who can say yes
Write down who is held accountable for a result and who can actually change it. The two lists often differ. This table is an illustration, not survey data, and yours will look different.
| Decision | Accountable for the result | Authority to change it |
|---|---|---|
| Pricing and promotions | MarketingRevenue target | Finance Sales |
| Who counts as a customer | MarketingAcquisition target | Sales Product |
| What makes a lead qualified | MarketingPipeline target | Sales |
| Moving budget mid-quarter | MarketingGrowth target | FinanceQuarterly approval |
| What counts as marketing-influenced revenue | Marketing | Finance |
How to tell whether this is your problem
Not every ignored insight is a decision problem. In Collibra's 2026 survey, 72% of AI decision-makers said a poor data foundation was the root cause when initiatives fell short. In Forrester's June survey, US agencies named accuracy and bias (63%) and legal concerns (62%) as top barriers. Those need their own fixes.
A quick test separates the cases. Pull the last 10 insights your team ignored.
The insight was wrong
Model or data problemIt was right and nothing happened
Decision problemPeople overrode it because they didn't trust the numbers
Data problem, and a reasonable overrideIf people overrode it because they didn't trust the numbers, the override was a reasonable response. The fix is the data, and data governance is where it starts.
What it looks like when it works
McKinsey's high performers are the roughly 6% of respondents who attribute 5% or more of their earnings to AI. About three-quarters say they have fundamentally redesigned workflows, against about a quarter of everyone else. They're twice as likely to have defined ways to measure AI's impact. A survey can't show cause, but the pattern is clear: they changed how the work gets decided, not just which tool people use.
We saw a version of it in our own work. For Kohler's Sterling brand in 2022, PMG ran strategy, creative, and media together and built a dashboard to guide optimizations and budget. Within weeks of launch, the work delivered 34% higher cost effectiveness than benchmark.
Where to start
Days 1 to 30
Pick one decision where insights get ignored. Get both sides in a room and agree on one definition and one number to optimize.
Leave with: a written definition and one number
Days 31 to 60
Name an owner for that decision. Set the override rules: who can override, how fast, and where the reason gets logged.
Leave with: a named owner and override rules
Days 61 to 90
Review the log. Some overrides will be right. A 2026 study of human-in-the-loop pricing found a hybrid can beat people or the model alone, but only when human judgment adds something the model can't see. Where overrides were wrong, make the insight the default. Where they were right, fix the model or the data.
Leave with: a reviewed log
Start with the insight your team ignores most often, and find the person who could have approved it. Ask what would have had to be true for them to say yes.
Terms in this article
- Decision intelligence
- Using data and AI to support specific business decisions and next steps.
- Human-in-the-loop AI
- AI that keeps people involved at key points and approves consequential decisions.
- Context engineering
- Supplying an AI model with business definitions and rules so its answers match company practice.
- Semantic layer
- Shared business definitions between raw data and the tools that use it.
- AI guardrails
- Controls that define what AI can access and do, what gets logged, and what needs sign-off.
- Enterprise AI adoption
- How widely and effectively people use AI at work, measured by usage, cost, and results.
Frequently asked questions
Why do teams ignore AI insights?
Often because the person who sees the insight can't change the price, segment, or budget behind it, or because teams disagree on what the model should maximize. Data quality and trust matter too.
Are ignored AI insights usually wrong?
Not necessarily. Pull the last 10 ignored insights. If they were wrong, it's a model or data problem. If they were right and nothing happened, it's a decision problem.
How do you get a team to act on AI insights?
Agree on one definition and one number to optimize, name an owner for the decision, and set override rules, including where the reason is logged. Review the log after 90 days.
What do high-performing AI adopters do differently?
McKinsey found its high performers were likelier to have redesigned workflows (about three-quarters, against about a quarter of others) and twice as likely to have defined ways to measure AI's impact. Survey data shows an association, not cause.
Sources
- McKinsey, The state of AI in 2026: On the road to ROI, August 2026.
- Grant Thornton, A widening AI proof gap is emerging, April 2026.
- Forrester and 4As, Nine in 10 US marketing agencies use AI to cut costs at the expense of creativity, June 2026.
- Docket and Heinz Marketing, The Trust Gap, Part 1: Sales and marketing alignment, October 2026. Survey fielded in September 2026. Docket sells AI sales agents.
- Collibra and Harris Poll, AI failures often trace back to poor data foundation, September 2026. Reported by CIO Dive. Collibra sells data governance software.
- Charles River Associates, Human-in-the-loop dynamic pricing, July 2026.
- PMG, Plumbers Getting It Done.
Last updated November 6, 2026. Statistics are quoted from the sources above and were checked on that date. Survey results are self-reported unless noted.