Every acquisition deck contains the same optimistic line: we will migrate the users to the new platform. It sits there in twelve-point font, a single bullet, as though moving people from one product to another were a data-transfer task rather than a test of whether they will stay at all. When Flywire acquired Cohort Go, I was handed a version of that bullet: move more than four thousand education agents onto Flywire’s new portal, without breaking the trust that made them worth acquiring in the first place. What I learned is that trust does not survive migration by accident, and it certainly does not survive a single message broadcast to everyone at once.
A migration is a churn event you scheduled yourself
Here is the framing that kept me honest. These four thousand agents were not prospects I was trying to win; they were a working relationship I could very easily break. Every one of them had a functioning habit, a portal they knew, and a workflow that fed their own business. My new platform, however superior on a feature grid, arrived as an interruption to people who had not asked for one. The metric that mattered was never “migrated,” a status you can claim the moment you flip a switch. It was “activated and still transacting sixty days later,” which is the only number that actually protects revenue, and the only one that proves trust held.
A status you can claim the moment you flip a switch.
The only number that protects revenue — and proves trust held.
One message would have failed four thousand people
The single biggest mistake available to me was the tempting one: write a clean, global migration message and send it everywhere. It would have been efficient, consistent, and quietly disastrous, because these agents were not one audience. They were education agencies across LATAM, APAC and India, each operating in a different commercial culture, each with a different reason to trust the old portal and a different fear about the new one. A migration message that ignores that is not neutral; it reads as a company that does not understand your market, which is precisely the doubt you cannot afford to plant mid-transition.
So I treated the global narrative as a spine, not a script. The strategic story stayed constant — the same product, the same commitment, the same direction — but the proof points, the emphasis, and the objections I pre-empted were tailored to each region’s buyer.
India — built at scale, sharpened on value
Agencies moving large volumes of students needed efficiency, speed and reliability at volume, and reassurance that nothing would slow the throughput their businesses depend on. The message led with what the new portal removed from their day, not what it added to it.
LATAM — trust runs through the relationship
A migration announced as a corporate fait accompli lands badly where the human connection is the account. The emphasis was on continuity and partnership — the people they dealt with were not going anywhere, and the portal was an investment in the relationship, not a replacement for it.
APAC — many markets, not one region
APAC is not one market pretending to be several; the mistake is treating it as a single region because it fits neatly on an org chart. The tailoring here was the most granular — emphasis, examples and even rollout pace adapted market by market.
None of this was a translation project. Translation swaps the words and keeps the assumptions; localisation changes what you choose to prove. The spine gave the migration coherence so it still felt like one company; the regional tailoring gave it credibility so it felt like a company that actually knew each market it was asking to move.
Translation swaps the words and keeps the assumptions. Localisation changes what you choose to prove.
Self-serve was how the tailoring reached everyone
I was, for practical purposes, a one-person function against a four-thousand-person audience across multiple time zones. Tailored messaging is only worth writing if it actually reaches the agent at the moment of doubt, so the delivery had to be as considered as the words. I built guided onboarding directly into the portal with Appcues, meeting agents at the exact point of confusion, and produced AI-generated Scribe tutorials so the how-to library was searchable at the point of need rather than trapped in a webinar nobody would attend. The AI was leverage, not the story; it let one person deliver region-aware onboarding at a scale that would otherwise have needed a team. Every question the product answered for itself was a ticket that never reached the queue, which is how you scale enablement without scaling headcount.
Enable the room before you enable the market
The tailoring could not stop at the customer. The sales and partnership people who owned these relationships were my first audience, because their confidence, or their anxiety, travelled straight down the line to the agent. So I equipped them with region-aware enablement built for the questions agents would actually ask: what changes, what stays the same, what is better for the agent’s own business, and what to say when someone pushes back, in the framing that fits their market. When the person holding the relationship can answer the hard question in the buyer’s own language, the migration stops feeling like something happening to the agent and starts feeling like something their trusted contact is guiding them through.
What transfers to any migration
Strip out the specifics of education agents and payment portals, and the pattern holds for any forced move.
Localise what you prove, not just what you say
A global spine keeps the story coherent; regional tailoring keeps it credible. Translation is not localisation.
Measure activation, not migration
A flipped switch is not an adoption. Prove the user came back and did the thing that makes you money.
Enable the room before the market
The people who own the relationship carry your message the last mile. Arm them in each market’s language, or watch their uncertainty become the customers’.
The four thousand moved, and, far more importantly, they kept transacting. What I took from it is a conviction I now bring to every launch: the moment you ask people to change something that already works for them, you are defending a relationship, not managing a project. Send everyone the same message and you tell each market you do not quite understand it. Tailor the proof to the buyer in front of you, market by market, and trust survives the very thing that was most likely to break it.