Everyone was half right

Ask the product team for a feature and the honest answer was usually "we can build that," and they were right. A decade of capable engineering was sitting in the platform. What the company could not do was agree on what to build next, or why, or first.

The roadmap was set by whoever argued hardest that week. Every function held a piece of the truth and none of them held the whole picture, so the loudest piece won. That is a specific kind of stuck. Not a broken company. A capable one that had stopped being able to decide.

“The problem was never the feature list. It was the way the whole thing was pointed.”

Stuck rarely announces itself. It shows up as a calendar full of meetings that resolve nothing, a backlog growing from every direction at once, and a quiet agreement to stop having the hard argument because it never goes anywhere.

Sales knew what was losing deals. Product knew what was fragile. Customer success knew what made customers quietly give up. Each of them was right, and each was solving a different problem, and nobody had put the three next to each other.

The tell was in the language. Ask what the top priority was and you got a different confident answer from every seat.

“Give it enough time and the workaround becomes the process. Nobody remembers it was ever a workaround.”

An audit is only as good as what it refuses to skip. In the first two weeks we looked through four lenses at the same time, because a problem that looks like design is often architecture, and a problem that looks like architecture is often a decision nobody made.

Business and vision first: where the company thought it was going and what a win in eighteen months actually looked like. Then the technical architecture, to learn what was truly load-bearing and what had calcified. Then the product itself, evaluated the way a buyer evaluates it, not the way it was built. Then the functional and usability audit, watching real people do real work, where the gap between "the product can do this" and "a person can do this" shows up in seconds.

The point isn't thoroughness for its own sake. The four lenses rarely agree, and the disagreement is the finding.

Fourteen conversations in a month, and we said at the start we weren't there to confirm anyone's hypothesis, including our own. Three questionnaires went out first: business direction, technical constraints, and how the product actually got sold. They agreed on something none of them had said out loud. The problem was orientation, not capability.

Then we went past the walls. The customers who lived in the product every day, the customer success people who absorbed the friction, and the partners and vendors who built on top of it and had opinions the company never heard.

The sharpest thing we found wasn't a complaint. It was the quiet. The longest-tenured users had stopped reporting problems. They said everything was fine, and they were also writing their own documentation and standing up their own tools next to the product. Quiet had been read as satisfaction for years. It wasn't satisfaction. It was surrender.

“You raise it, it drops into some void, and you've no idea if anyone's looking at it.”

Three days, in person, and not a readout. This is where an audit stops being a document and becomes a decision. A working structure with three phases: absorb, prioritise, plan. Build one shared diagnosis before anyone defends a corner, score the work together, walk out with a sequence.

Everyone who owned a piece of the truth was in the room: product, engineering, the AI lead, customer success, operations, pre-sales, the chief executive. Each got a floor they owned.

The exercises moved people from being presented to into building something together. A walk-the-walls read of the evidence, so the findings belonged to the room. A ten-minute paper-bridge build that got competitive fast, on purpose. Then a scoring instrument where each function weighed the work differently by design, because the place where product and engineering disagree is the conversation, not a number to average away.

The recommendations were voted. The room scored every initiative, argued the ties out loud, and left with an order it had set itself. That is the difference between a report a company nods at and a plan a company owns.

“Give it enough time and the workaround becomes the process. Nobody remembers it was ever a workaround.”

The product was not behind on features. It was pointed at the wrong thing. It had been built to match the shape of its database instead of the shape of the work, so it leaned on a handful of people to make it usable, and the value was trapped in those people instead of living in the system.

Every function's private worry turned out to be the same worry wearing different clothes. Saying it plainly was the whole point. A company that disagrees with itself cannot act, and the reason it disagreed was that nobody had assembled the fragments into a single picture and held it up to the light.

Once the picture was on the wall, an argument that had run for years resolved in an afternoon, because it turned out everyone had been half right.

Side by sideSlider

Before

After

Six functions, six versions of the problem. Now one picture the whole room agrees on.

A diagnosis that doesn't turn into a sequence is just a nicer way to feel stuck. So the last piece was a roadmap the room had already voted for, sorted into what was doable now versus what needed building, and honest about which was which.

Every section was classified: ready, ready with caveats, or not ready. That sounds procedural and it is the opposite. It is what lets a company start Monday on the safe two-thirds instead of waiting for a perfect plan that never comes.

The sequence was ordered around one idea the room had agreed on: the product should hold up on its own and carry its own work. So the first moves were the ones that shifted the most weight off the people and into the system. Not a wish list. A path, with a reason for the order.

The engagement fit inside a month, and the last week was spent on outcomes, not analysis, which is the whole design of it. A company that had argued about direction for longer than that walked out with a direction it had set itself, in the time it usually takes to schedule the meetings.

1. Week 1
   Immersion

   Business and vision sessions, technical architecture study.

2. Week 2
   Evidence

   Product review, functional and usability audit, conversations with customers, CS, and partners.

3. Week 3
   Alignment

   Synthesis, then the three days in the room where recommendations get scored and voted.

4. Week 4
   Outcomes

   The sequenced roadmap and the readout.

What changed wasn't a screen. It was that the arguing stopped and the building started, pointed the right way.

If a capable product has stopped moving and no one can agree on why, the problem is almost never the feature list. It's that the truth is scattered across the people who each hold a piece of it, and nobody has put the pieces together and made the company look.

That's what the audit does. In 30 days you get a straight answer about where you actually stand, a roadmap you can act on, and the thing you can't buy any other way: a leadership team that agrees on both.

It's built for mid-to-large organizations with a real product and a real market, where the interface and the direction have drifted apart and the cost of the wrong roadmap is measured in quarters. It's the step before you spend the money, so you spend it on the right thing.

“More and more customers want to work through agents and APIs, not just a screen.”

That's why this matters more now, not less. The companies pulling ahead are the ones getting honest about where they stand before they commit to the next big move, because the move is expensive and the market isn't waiting.

A note on this story. This is a real 32°N engagement, shared with the client's name, product, and identifying details withheld, and some specifics generalized, to protect their confidentiality. The quotes, including the note from their chief executive, are genuine, taken from the actual sessions and correspondence with identifying information removed.