From invisible design work to high-impact business influence

Nobody knew how talented they were

The shaping conversations were happening in DMs. When they finally did share publicly it was a wall of text with no clear ask, and they wondered why nobody engaged. The result: nobody knew how talented they were. The work had a perception of being slow.

“Where are they? I haven’t seen them post in a week.” — Leadership

They thought good work was enough

They were operating on the belief that doing good work was enough. At a large company, maybe. At a startup where you want a promotion, that belief will quietly cost you the promotion. Work does not speak for itself. You need to be your own marketing team.

This was the belief I had to dismantle. Everything else followed from it.

Nine months to Lead, if they wanted it

The aim was promotion to Lead within nine months. I told them the timeline. I also told them I would not drag them there. People have to want it. I cannot do it for them. So before anything else I confirmed they actually wanted the promotion and were willing to do uncomfortable work to get it. They did.

Design out loud

I created a roles-and-responsibilities matrix so the expectations for a Lead role were explicit. I wrote performance goals that covered their own objectives but forced them to address the weaknesses they had been routing around.

Then I reframed success. Quality of the work in their own opinion was off the table as a measure. The new measure was business impact and how well that impact was communicated.

The communication system was a methodology I call ‘design out loud’:

• One Loom video in a public channel per week.
• Written updates no longer than five lines.
• When feedback is needed: clear ask, clear context, clear timeframe. Tag the stakeholders directly. Chase them.
• Share with execs directly (the only DMs allowed about work) when you know it is an area or strategic project they care about. If you do not know what they care about, set up a monthly skip-level to find out.

Early on, the hardest part was getting them to share at all. One of those nudges got this reply:

My own role moved through three phases. First I directed them: I told them what to post, where, and when. Then I coached them: I asked questions instead of giving answers. By the end I was asking for their recommendation, and getting a better one than the one I had.

I built the system, then handed it over

This designer was one of three reporting to me. The same move I made with this designer, I made with the whole team: build the system first, then hand it over.

I started the logging on everyone’s behalf. Each designer got a ‘brag doc’, and I wrote a Claude skill that pulls the evidence together so capturing it is not one more chore. Once it was running I handed each of them their own and explained why it mattered; the point was never for me to own their record, it was for them to.

I also kept a private bank of the good things people said about their work, and read it back to them on their work-anniversaries.

Recognition reached the top

Executive-level engagement and recognition. Ownership of projects they believed in. More confident in making decisions. Sharing opinions more freely. Able to push back on decisions and reach agreements rather than capitulate. The fear of being wrong in public did not go away. They had tactics that worked faster than the fear.

A number nobody asked for

On their own initiative, with no brief, they ran a retrospective impact analysis on a feature change that had shipped months earlier (the kind of self-organised, data-led work they would not have been confident to start nine months before). They posted it to a public channel. The COO recognised it unprompted in front of leadership; an exec read the numbers off their analysis and calculated a roughly 40% relative drop in churn among the customers the change targeted, and others started picking over the findings and deferring to their read of the data.

The honest part is what made it land. Churn had also fallen for customers the change did not target, so the drop could not be pinned cleanly on this one piece of work; a broader retention push was running at the same time. They surfaced that themselves rather than claiming the headline number. A weaker analyst takes the 40% and moves on; surfacing the result that complicates your own story is exactly why senior people started trusting the judgement behind it.

One person’s habit became how we work

The designer started sharing how to get the data, how they socialised the idea, and how to report on effectiveness. A self-organising system of data-informed and design-led work. I codified it into a 20% workstream and a dedicated report to leadership so the methodology stopped being one person’s habit and became how the team operates.

The matrix kept evolving

The roles-and-responsibilities matrix went through its own iteration. V1 was a row-per-competency snapshot built for use in 1:1s. Printable, scannable, behavioural rather than aspirational (“proactively communicates capacity-based deadlines cross-functionally” rather than “strong communicator”). V2 added a Frequency of Evidence column, because a behaviour without a cadence expectation isn’t verifiable. It also split Strategic Thinking into three separate rows (Defining the Problem, Initiating Work, Managing Up) so one strong behaviour couldn’t mask a gap in another. V2 emerged during the next designer’s promotion planning exercise, which was the right test: the matrix should change in response to real coaching, not in response to a planning meeting.

The full rebuild has its own case study: how a real promotion case broke v1, and how v2 spelled out the evidence for every row.

I used the matrix live in 1:1s, not just in performance reviews. When a designer navigated a situation well I named the competency. When they missed I used the matrix to frame the gap in terms of behaviour, not judgement. A coaching trends tracker, updated weekly, made progression visible across the team, and showed measurable week-over-week improvement for the designers I was coaching.

The promotion came after I was laid off

Nine months to Lead was the target. Before it was up, I was laid off. They were promoted to design manager soon after; my leaving helped open that path. I am confident the Lead track we built would have got them there on its own.

The promotion mattered. The quieter part matters more: a designer whose work had been invisible now runs work that leadership defers to, and captures the impact themselves. That outlived me, which is the only real test of whether the coaching worked.

Four principles I now run on

The satisfaction I used to get from IC work has been replaced by watching my team succeed. I now have a clear framework for design teams to get recognition and do meaningful work that they define.

I used this coaching as an opportunity to push the wider Product Design strategy: better signal through better measurement. Four principles came out of it that I now run the team on.

Design from evidence. Every design decision should be traceable to customer data. Design’s first job is to get the data.

Measure, ship, measure. An AI prototype with baseline and success metrics beats a perfect hand-crafted mockup without them.

Effectiveness over experience. A tool that sounds and looks great but fails to do what the customer needs is a useless tool.

Transparency builds trust. Showing how AI performs, why it made certain choices, and what customers can do about it builds trust, which lowers churn.

The next two designers I am coaching have different problems. One is over-confident. The other skips framing and goes straight to execution. The framework is the same: behavioural goals, a frequency of evidence, “design out loud” cadence, weekly coaching tracked against the matrix. The structure scales.