You Can't Command a Peer. You Can Only Convince One.

Leadership is influence, nothing more, nothing less.

John C. Maxwell, author of The 21 Irrefutable Laws of Leadership

Seven countries, seven Net Promoter Scores (NPS)¹, and almost nothing in common between them. One market was reporting patient satisfaction scores in the 80s. Another sat in the low 20s. The rest were scattered somewhere in between, and the first instinct — the obvious one — was to assume the gap reflected something real about the quality of care in each market. It didn't, or at least not mostly.

The market in the 80s was gathering its scores through in-person surveys, selectively, from patients who'd already had a positive enough experience to be approached. The market in the low 20s was doing the opposite: sending an SMS survey to every single patient, satisfied or not, with no filtering at all. The two numbers weren't measuring the same thing. They were barely measuring the same kind of thing, and no amount of comparing them to each other was ever going to produce a useful insight, however confidently each market presented its own number in a regional meeting.

Five principles came out of untangling it, and they apply to any metric you need several teams to report the same way.

The job was never to pick a number. It was to make the numbers mean the same thing.

Alongside my day-to-day role, I'd been asked to align patient experience metrics across the group's regional footprint. None of the gap was deliberate dishonesty. Each market had simply built its own measurement practice over time, optimized for whatever made sense locally, with no one ever asking whether those practices could actually be compared to each other.

The work started with the unglamorous part: going market by market to understand exactly how each one currently measured satisfaction, before trying to change anything. The differences went further than survey method. NPS is built on a 0 to 10 question, but some markets had adopted a 1 to 10 scale and others 1 to 5, so even the arithmetic behind the headline number wasn't shared. Only once that picture was clear did it make sense to socialize what a comparable methodology should look like, and even then every proposed standard had to be checked again afterward, market by market, because a new instruction doesn't reliably become a new practice unless someone confirms that it did.

Underneath sat a second problem: how patients experience care varies by culture as much as by hospital. A standard built around what one market considered good service could land as irrelevant, or actively wrong, in another. Unifying the measurement couldn't mean flattening that difference. It meant finding what was genuinely comparable underneath it, and being honest about where that line sat.

Leading people who don't report to you

None of this came with formal authority. I couldn't instruct a country CEO or a hospital director in another market to change how they measured anything, and leading as though I could would have failed quickly. The only real lever was showing people why a shared standard was worth adopting voluntarily, one relationship at a time.

What did the most to turn compliance into buy-in was a regular rhythm of getting everyone in the same room: a monthly regional session, and an annual in-person summit. Each country was given airtime to share its own initiatives, and we brainstormed solutions together rather than receiving them from me. By the time we proposed a common service values program, a "service values 101" for every market, much of the resistance had already eased, and it was later implemented across the countries.

None of it would have gone as far without a senior sponsor who believed in the work and was willing to nudge country heads who were slow to engage or unwilling to align decisions. The sponsor didn't replace the work of making the case myself, and it never turned into an instruction handed down on my behalf. But it opened conversations that peer-to-peer persuasion alone sometimes couldn't, and the story would be tidier, and less true, without it.

What changed once the numbers meant the same thing

The scores didn't converge neatly, and I'd be suspicious of any version of this story where they did. They shifted only slightly once the methodology was aligned. What changed was what the scores could be used for. Instead of ranking the markets, we used them as the basis for understanding what each country's own problems were.

With the measurement more uniform, the differences that persisted pointed to something other than method: possible cultural biases in how patients in different markets respond to the same question. Some markets may simply be more generous scorers than others, and no amount of standardizing will remove that. That was more useful than a clean convergence would have been. For the first time, the group could start separating the part of each gap that came from how the question was asked from the part that might reflect how patients actually experienced care, or how they chose to rate it.

Five principles worth borrowing

1. Agree on what you're measuring, and be transparent about it. NPS asks for a score from 0 to 10, yet some markets had adopted 1 to 10 and others 1 to 5. Write down the definition, the scale and the formula, and make sure every party is working from the same page.
2. Measure the same way everywhere. A shared definition isn't enough if the method differs. For NPS, that meant surveying every patient rather than letting a team choose whom to ask, but the principle holds for any metric: the approach has to be consistent across parties.
3. Use the scores to understand problems, not to rank. A market defending its number learns less than one asking what the number says about its own patients.
4. Let every voice be heard. Regular airtime to share initiatives and brainstorm solutions together reduced resistance when we later proposed a common standard.
5. Find a senior champion. Leading without authority is easier with someone who has standing and believes in the work. A champion can't do the convincing for you, but they make the conversations possible.

Beneath all five sits the lesson that's stayed with me from that project, which has very little to do with NPS specifically. It's that influence, absent any formal authority to compel anything, depends entirely on being able to show people a shared standard is worth more to them than whatever private version they're already comfortable with.

You can't command a peer into agreement. You can only make the case clearly enough, and patiently enough, that they choose it themselves.