Protect the Core Magic: Acquisition Integration
Freeze the loop users love, rebuild everything behind it, and narrate the seams honestly
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 93%
Hardiman's integration doctrine, learned on Wordle: the question isn't only whether to acquire, it's whether you're ready to integrate. You identify the core magic — the exact loop users are emotionally attached to — and declare it untouchable. Then you fix only what the acquisition's amateur infrastructure got wrong (the things users lose), migrate to your stack, and expand surfaces. Because a migration is mid-flight for a long time, you assume something will go wrong publicly and pre-commit to radical transparency about the product development process.
Origin
Alex Hardiman and the NYT Games team (GM Jonathan Knight) integrating Wordle, created by Brooklyn software engineer Josh Wardle as a gift for his partner, after a Times reporter's column surfaced it in January 2022.
Core principles
- 01Acquisition readiness is integration readiness, not just deal rationale
- 02Name the core magic before you touch anything — the game board and the core loop stay unchanged
- 03Fix what users are losing, not what engineers find ugly
- 04The best integration is invisible: the only thing users notice is that their history follows them
- 05During a mid-flight migration you will not control everything — so control the honesty
How to run it
- 1
Buy for DNA fit, fast
Acquire when the target shares your existing products' DNA and slots into a category your strategy already requires. Wordle shared DNA with Spelling Bee and the Crossword Mini (Josh Wardle was openly inspired by them), and games was a strategically essential category — a counterpoint to news that feels like time well spent, not empty calories.
Pro tip The GM of the category should already be tracking the target before it becomes news. Jonathan Knight had reached out to Josh before the Times column even published; the whole deal took weeks.
- 2
Ask the readiness question
Before closing, ask explicitly: are we ready to actually integrate this? Hardiman flags this as the real lesson for most companies — the opportunity analysis is the easy half.
- 3
Declare the core magic untouchable
Identify what users are emotionally attached to and freeze it. For Wordle: the game board experience and the core loop of the game remain unchanged. Everything else is fair game.
Pro tip Test the emotional range of the attachment — Hardiman notes Wordle resonates with an 8-year-old and an 88-year-old alike. Broad attachment means narrow permission to change.
Watch out Users will still accuse you of changing the core (many believe the Times made Wordle harder — it didn't). Expect the suspicion regardless.
- 4
Fix the thing users are losing
Wordle had no back end; stats and streaks — the social currency people shared — lived in the local browser, so a new phone or browser wiped a user's whole history. The Times connected Wordle to a free NYT account so stats and streaks were preserved.
Pro tip Frame the account requirement around a benefit the user already wanted (my streak survives), not around your CRM.
- 5
Rewrite into your stack and expand surfaces
Rewrite the product in your tech stack and put it on your major surfaces — the news home page, the games app — so users who come for Spelling Bee or the Crossword also find their Wordle.
Watch out A migration means two populations exist at once: some users on the original build, some on yours. Any 'just change it on the back end' fix will only reach half of them.
- 6
Pre-commit to transparency about the seams
When something goes wrong mid-migration, come out and explain the mechanics — including what your product development process actually is and why you couldn't just fix it. Exposing the process demystifies the rumour.
Pro tip State plainly what you are not trying to communicate. The Times said explicitly it was not editorialising and that Wordle is meant to be a fun diversion from the news.
Watch out Transparency does not convert everyone. Some people will never believe the simple explanation — publish it anyway.
In the wild
Just as the Supreme Court's draft ruling on Roe v. Wade leaked, an engineer on the Games team noticed the next day's Wordle solution was 'fetus' — a word Josh Wardle had loaded into the game months earlier. Because the migration was mid-stream, with some users on the original game and some on the new version, the Times could not change the word on the back end for everyone, only for some.
→ The Times went public: explained it was mid-integration, that the word was pre-loaded by the founder months earlier, and that Wordle was not meant as commentary. Hardiman: 'everyone understood.'
The team rewrote a back-end-less browser game into the NYT stack, attached it to a free NYT account to persist stats and streaks across devices and browsers, and surfaced it on the news home page and the games app — all while leaving the game board and core loop untouched.
→ Hardiman: the experience was seamless and 'the only thing you would notice that's changed is that the New York Times knows enough about who you are so that your stats carry over and you can play anywhere.'
Common mistakes
Improving the core loop because you now own it
The reason the acquisition was worth making is the thing users already love. Touching the game board, the loop, or the difficulty destroys the asset — and even the perception that you did (people still believe the Times made Wordle harder) costs you trust.
Assuming a back-end fix reaches everyone mid-migration
During a phased migration the user base is split across two builds. A hotfix on the new back end only reaches the migrated half — so incident plans written on the assumption of a single controllable system will fail at the worst possible moment.
Is it for you?
Best for
A product leader integrating a beloved consumer product acquired from a solo creator or small team, where user attachment is the asset being bought
Not ideal for
Acqui-hires or technology acquisitions where the product itself is being sunset — there is no core magic to protect
From the transcript
“are you ready to actually integrate an acquisition?”
“Wordle players feel such connection to the game and we really wanted to make sure not to interfere with the core magic of the experience.”
“to make sure that the game board experience and the core loop of the game remained unchanged”
“we decided to do was undertake a project to connect Wordle to a New York Times account”
“because we were midstream on the migration process, and some users were on the original Wordle game, others had migrated to the new version, it…”
“this is where like coming out being really transparent about the facts, in some cases just exposing more about the product development process really helps”
From the episode
An inside look at how the New York Times builds product
Alex Hardiman (CPO at The New York Times)