The final slide.
What happens to the best researcher in the building when we get to that final moment, and someone asks: "So what should we do?"
Every business has one: the researcher everyone turns to when the numbers look strange. We'll call her Naledi. She's a composite of a lot of very good researchers we've worked with over the years, and you may recognise her. You may be her.
Naledi is rigorous. She checks whether a pattern is real before she lets anyone build a story around it, and she goes looking for the thing that doesn't fit. When she finds it, she says so. People trust her work, and they should.
Then comes the final slide
Picture the debrief. The work is good. The findings are clear, properly evidenced and well presented. The room is engaged, people are asking smart questions, and Naledi knows the material inside out. Then someone asks, "So what should we do?" and something shifts. What started in the working deck as we should stop investing here has become an area to explore. Naledi talks through the options and gives the caveats, and the meeting ends well.
Two weeks later, the actual decision gets made in a room she isn't in.
What insights hold, and what very often gets slipped
In Human Influence terms, Naledi's Pattern is strong. She can tell the difference between something interesting and something real. Under pressure, she stays close to the evidence and won't let the room drag the work toward the story it would prefer to hear. Her Meaning is strong too. She knows what the evidence means. Ask her privately, or sit with her own team after the meeting, and she'll probably tell you exactly what she thinks should happen next.
So she has a view. The real test is whether that view survives contact with the room. It's hardest when the conclusion is inconvenient, when someone senior disagrees, or when saying it plainly means moving from here is what we found to here is what I think you should do about it. That is often where Meaning starts to slip.
It happens to good researchers for a reason
Research trains caution, and that's mostly a good thing. The instinct that makes Naledi test a pattern twice also makes her deeply uncomfortable claiming more than the evidence can support. Somewhere along the way, though, methodological caution can become organisational caution, and taking a position starts to feel like overreach. In plenty of organisations the downside is asymmetric, too. The researcher who goes too far gets challenged for it. The stakeholder who ignores the evidence and makes the wrong decision rarely gets accused of failing at research. So the language softens. We recommend becomes you might consider, and stop doing this becomes an opportunity to explore. It feels more responsible, and sometimes it is. Sometimes it simply moves the judgement somewhere else.
Meaning doesn't disappear when the researcher steps back
Someone still has to decide what the evidence means, and if the researcher doesn't, the room will. The room often arrives with an existing commercial agenda, an existing hypothesis and a fairly strong opinion about what it would like the answer to be. So the research gets heard. It gets quoted, and it might even appear in the next strategy deck. But the decision itself barely moves.
That gap matters even more now. AI is making it dramatically easier to gather, organise and summarise evidence, which moves the scarce capability further downstream. Finding the evidence is becoming the easy part. The harder question is whether we can work out what matters, take a position on it, and carry that judgement into the decision.
A question worth asking
Think about your last three pieces of work.
How many changed a decision? And for the ones that didn't, where were you when the decision was actually made?
See where your influence holds, and where it leaks
The HI Score is a short diagnostic for people working in insight and research. It looks at how Question, Pattern and Meaning show up in the way you work, and at what happens to them when the work gets pressured, challenged or pulled into a real organisational decision.