Google announced earlier this month that it's buying Mandiant, the security outfit, for a sum with a lot of zeroes on it. I don't run Mandiant's stuff myself, so my first reaction was idle curiosity rather than alarm. But it landed in the same week I'd been quietly worrying about a different acquisition, one closer to home, involving a tool I actually depend on, and the two together set off a feeling I know well: that low-grade dread when something you build on gets a new owner.
It's a strange anxiety because nothing has technically changed. The tool I use still works the same on the day after the announcement as the day before. The binaries are the same. The docs are the same. And yet the ground feels different, because the incentives behind the thing have just shifted, and incentives are what decide where a tool goes next.
I've been burned enough times to have a rough mental checklist now. The first question is: who bought it, and why? A big infrastructure company buying a tool to fold it into a platform is one story. A private equity firm buying it to extract value is a very different one, and historically the second story ends with the free tier shrinking, the prices climbing, and the bits I cared about being quietly deprecated in favour of the bits that bill monthly. I'm not being cynical, I'm reading the patterns.
The second question is the one that actually matters: how stuck am I? This is the useful exercise, and it's worth doing before any acquisition forces it. For each tool I lean on, I try to know honestly what my exit looks like. Is it open source, so the worst case is a fork and a community that picks it up? Is my data in a format I can export and read elsewhere, or is it locked in a proprietary store I can only reach through their client? Could I swap it out in an afternoon, a week, or never? The answer for the tool worrying me this month was "a week, painfully", which is uncomfortable but not catastrophic. Knowing that number is what lets me stop catastrophising.
What I've learned is that the acquisition itself is rarely the moment to act. It's the moment to take stock. Most acquired tools coast for a year or two before the new owner's priorities really show up in the product, and a lot of the doom I've felt over the years never materialised. The danger isn't the announcement, it's complacency afterwards: forgetting you ever felt that flicker of unease, and then being genuinely stranded two years later when the thing finally gets sunset or strip-mined.
So my actual response to this month's news is unglamorous. I'm not migrating anything. I'm writing down, for the tool that matters to me, exactly what I'd do if the worst happened: where the data lives, what the nearest alternative is, and how long the switch would take. Then I'll get on with my work and check in on the thing in six months. An acquisition is a useful prompt to ask whether you've built on rock or on someone else's balance sheet. Most of the time you can't change the answer. You can at least stop pretending you don't know it.