The Meeting Changes When They Can See

Most senior leaders I work with spend their meetings asking questions. Not because they want to, and not because they enjoy the sound of their own voice, but because the meeting is the only place the information lives. A VP walks in, sits down with a direct report, and the first twenty minutes are just extraction. Where are we on the campaign. What did the agency say. Why is that number lower than last month. The direct report reports, the VP nods and writes things down, and by the time anyone has a thought worth having, the calendar invite is already telling them to wrap up.

That is the normal shape of it, and it is so normal that nobody questions it. The leader is a receiver of information. They get a deck from an agency on Monday, they get a verbal update from their team on Wednesday, and somewhere in the gaps between those two things they are supposed to form a strategy. Except forming a strategy takes time to think, and all of their time is being spent on the retrieval, so the thinking gets squeezed into the car ride home.

I have started to believe that the single most useful thing we do for a senior leader has almost nothing to do with the fancy parts of AI. It is just making the unseen seen. When you take the stuff a leader normally has to ask for and put it somewhere they can actually look at it, something changes in them that I did not expect the first few times I watched it happen. They go quiet. They start noticing things nobody flagged. And the next conversation they have with their team is a completely different conversation.

Top Cup is the clearest example I have of this.

Top Cup makes aluminum cups, and they have figured out how to do genuinely beautiful custom printing on them. If you have ever picked up a craft beer can and thought, quietly, that whoever designed this can was better at their job than most people are at theirs, that is roughly the level of print we are talking about, except now it is on a reusable cup instead of a thing you throw away in ninety seconds. It is a good product with a real reason to exist, which already puts them ahead of a lot of companies I meet.

The problem was not the product. The problem was that Top Cup sells everywhere. They sell business to business, they sell direct to consumer, they run ads across a pile of different platforms, and every one of those channels kept its data in its own little room with the door closed. The VP of Marketing knew all of it was happening. She just could not see it in one place, and because she could not see it, every question she wanted to ask started with the retrieval problem I described above.

So we pulled it together. Every ad source, every sales channel, all the different feeds, into one dashboard that a human being can actually read. That part is not glamorous and I will not pretend it was. It is plumbing.

Here is where I got it wrong the first time, and I got it wrong in the most predictable way possible. I built the first version myself, alone, the way a consultant does when he is quietly sure he already knows what the client needs. It was fine. It was an okay enough dashboard. It was also not something the team actually wanted to open, and it did not get good until I stopped building for them and started building with them. We sat down together, they told me which numbers they actually cared about and which ones I had put front and center for no reason other than that they were easy to chart, and we added and cut features until it was theirs instead of mine. The difference between the version I made and the version we made was the difference between a dashboard that technically worked and one people used on a Monday. I know this lesson. I have known it for years. I still had to relearn it in front of Top Cup.

Once the plumbing was done and the thing was actually built with the people who would live in it, we had a single source of truth, and that is where it got interesting, because now we could point Claude at the data and the code underneath it and actually ask it things.

One of the questions we asked was about sequence. Top Cup works with influencers and creators, and they also run paid ads, and the open question was whether the order mattered. Should the paid spend come first, or should you let the creators go and then layer the ads on top of what they stirred up. This is the kind of question an agency will happily give you an opinion on, and the opinion will sound confident, and you will have no real way to check it. Claude went into the database and pulled out the pattern that was actually working. Not a hunch, not a best practice from a blog, the thing their own numbers said.

Within a very short window we had a repeatable playbook. And the part I care about most is what that did to Top Cup's relationship with its agencies. Before, the pattern was the one everybody knows: the agency sends a report, the report says the agency is doing a good job, and the client nods and approves it, because what else are they going to do, they do not have their own version of the truth to argue with. Now the VP walks in with her own read. She can say, I see the return on ad spend dips in this channel when we sequence it this way, let us talk about how we fix that. It stops being a receiving exercise and becomes delegation. Less “tell me what happened” and more “here is what I noticed, now go make it better.”

I want to be careful not to oversell this. It did not fix everything, and it was never going to. What it did was move the objective stuff into a place where it could be seen, which frees up the subjective stuff to go where it actually belongs, which is the creative side of marketing where a human should be spending their judgment. You do not want your smartest people spending their judgment on whether the numbers are right. You want them spending it on the work only they can do.

Here is the part that still gets me a little. The whole thing runs on almost nothing. A handful of tokens and a Claude account that costs about twenty dollars a month. For that, a VP of Marketing effectively got a data analyst, a sales operations person, and a strategist sitting next to her, and those are three roles a company that size could never have justified hiring for a single marketing function. We did not give her a tool. We gave her the three colleagues she had been quietly missing.