Your personal CRM just got acquired. What happens to your data?
Clay was acquired and became Mesh. Nothing broke — but nobody asked the users. How to check an app's exit risk before you trust it with years of notes.
On the day an acquisition is announced, almost nothing happens to your data — the app opens, your notes are there, the logo is different. Clay’s 2025 sale to Automattic and its rename to Mesh is the current example: no features removed, no data lost. What changes is who decides what happens next, and you are not asked.
What actually happened to Clay
In June 2025 Automattic acquired Clay, the design-forward personal CRM at clay.earth. During 2026 the product was renamed Mesh and moved to me.sh, with the old domain redirecting. By every account, including the reviews written by its competitors, the features carried over intact. If you were a Clay user, the practical consequence of the acquisition was that you learned a new name.
That is worth stating plainly, because the instinct in privacy-minded circles is to treat any acquisition as a betrayal in progress. It usually is not. Automattic has run WordPress.com for two decades and did not buy Clay in order to break it.
What did change is who decides. The people who wrote Clay’s original positioning no longer control its pricing, its roadmap, or its data policy, and the users who chose it partly because of that positioning were not asked and will not be. Nothing about that is unusual — it is how software companies work — but it is the thing to actually notice.
The part that is genuinely fragile
Ask what you would lose if you had to leave tomorrow, and the answer splits cleanly in two.
Contact fields travel. Names, email addresses, phone numbers, companies, tags. These move between tools reliably because vCard and CSV are decades old, dull, and supported by everything. No vendor has an incentive to break them and none of them do.
Computed things do not travel. Relationship-strength scores, warmth rankings, reconnect suggestions, the layout of a network graph, the “who should I contact this week” list. These are not stored records; they are calculations the app performs over your data using logic that belongs to the vendor. There is nothing to export because there was never a file.
Between the two sits the category that costs people the most: interaction history. Some tools store every logged note as a durable record and export all of it. Others infer the history from your inbox and calendar, keep a derived summary, and produce something thin and disappointing when you ask for it back. The difference is invisible from the marketing page and obvious the first time you open the export.
Five questions to ask before you commit
None of this requires you to predict the future. It requires you to check the exit before you need it, which takes about ten minutes.
- Is there an export at all? If the answer is a support-ticket process rather than a button, that is the answer.
- Is it complete? Contacts and notes, or just contacts. Export a real account, not a demo one.
- Is it free? An export locked inside the paid tier is an export you lose at the moment you cancel.
- Is the format readable without the app? Markdown, CSV and vCard are. A proprietary archive is a copy of your data you cannot actually read.
- Is there a written commitment about shutdown or acquisition? Most tools have nothing here. It is not disqualifying, but its absence is information.
The fifth question is the one almost nobody asks before signing up and almost everybody wishes they had asked afterwards. We publish our own answer on the no-acquisition and source-release pledge page, and the honest framing is that a pledge is weaker than open source: it is a public, dated commitment rather than a licence someone else already holds. Monica’s AGPL self-hosted edition is the stronger guarantee, and the Monica versus Mesh comparison is the clearest illustration of what each end of that trade costs.
Where the data lives is only half the question
Local-first storage — the data on your device rather than a vendor’s server — removes one failure mode entirely. Nobody can switch off a server and take your notes with it, and nobody can gate your access behind a billing dispute.
It does not remove the other one. An app that stops being maintained will eventually stop running on a new operating system, and a local file in a proprietary format is a slower version of the same trap. The location protects you from the vendor; the format protects you from time. You want both, which is why plain Markdown or CSV on a disk you control is the combination that actually survives — and why our security page treats the export as part of the threat model rather than a feature.
What to do if you already depend on one
Do not switch on principle. Migration costs you an evening and the entire computed history, and the tool you move to can be acquired next year. The cost people underestimate is not the transfer itself — that part is a file — but the accumulated context that never makes it into the file.
Do this instead. Export today and open the file. If your notes are in there, in a format you can read, you are fine and can stop thinking about it. Put the file somewhere you control and repeat roughly twice a year, or immediately after any announcement about ownership, pricing, or data policy.
Then move only when the new owner does something concrete you object to: a price you will not pay, a policy you did not agree to, a feature you relied on being removed. That is a decision based on what happened, rather than on what might. If it comes to that, the buyer’s guide sorts the alternatives by what you actually need rather than by who is loudest — including the several cases where the answer is not us.
We should say plainly that we sell a competing personal CRM and publish the kind of pledge this article recommends you look for. That is a reason to read this piece sceptically. It is not a reason to skip the export.
References
-
Article What happened to Clay (clay.earth): Clay is now Mesh
Wend, 2026
-
Article Clay personal CRM review 2026: now called Mesh
Use Apify, 2026
-
Article Mesh CRM review 2026
Dex, 2026
FAQ
What actually happened to Clay?
Clay (clay.earth) was acquired by Automattic in June 2025 and rebranded to Mesh at me.sh during 2026; the old domain redirects to the new one. Reporting and third-party reviews agree that no features were removed in the transition — this was a rename and an owner change, not a shutdown. Worth noting because the name causes confusion: this is unrelated to clay.com, a B2B sales-prospecting platform that happens to share the word.
Is an acquisition bad news for users?
Not by itself, and treating it as automatic bad news will make you move for no reason. An acquirer with money can fix the bugs a small team could not, and Mesh is the current example of a transition that did not break anything. What changes is governance: the roadmap, the pricing, and the data policy are now set by people who did not make the original promises, and you were not consulted and will not be.
What should I do the day my app gets acquired?
Take an export, and then actually open it. Not because you should leave — most of the time you should not — but because the export you have never tested is the one that turns out to be a partial CSV at the exact moment you need it. Check three things: are your notes in there and not just the contact fields, can you read it without the app, and is exporting behind a paywall you would lose on cancelling.
What gets lost when you migrate between personal CRMs?
Contact fields travel reliably — names, emails, phone numbers, companies, tags — because vCard and CSV are old, boring, and universally supported. What does not travel is everything the app computed rather than stored: relationship-strength scores, warmth rankings, reconnect suggestions, the layout of a network graph. Also fragile are interaction timelines, which often export as a thin summary. Assume anything that felt clever is gone and anything that felt boring survives.
How do I check an app's exit risk before I commit?
Five questions, all answerable in about ten minutes before you sign up. Is there an export? Is it complete — notes as well as contacts? Is it free, or does it sit behind the tier you would cancel? Is the format readable without the app — Markdown, CSV and vCard are; a proprietary blob is not? And is there a written commitment about what happens on shutdown or acquisition? Most tools fail the last one. Very few fail all five.
Does open source actually protect me?
It protects you more than any promise does, because it does not depend on anyone keeping their word. Monica is self-hostable under AGPL, which means an acquisition of the company cannot take away the copy running on your own server. The cost is real: you maintain it, patch it, and back it up. Open source removes the governance risk and hands you an operations job in exchange — a good trade for some people and a bad one for most.
Is local-first the same as being safe from this?
It solves a different half of the problem. Local-first means the data sits on your device, so nobody can switch off a server and take your notes with it — but if the app stops being updated, an unmaintained binary eventually stops running on a new OS. That is why the format matters as much as the location: a local file in a proprietary format is a slow-motion version of the same trap. Plain Markdown or CSV on your own disk is the combination that actually survives.
Should I switch tools just because of an acquisition?
No, not on its own. Migration has a real cost — you lose the computed history, you spend an evening on it, and the destination might get acquired next. Switch when the new owner does something concrete you object to: a price change you will not pay, a data policy you did not agree to, or a feature you depended on being removed. Until then, take the export, keep it, and carry on using the app.
What does a shutdown pledge actually commit a company to?
It varies, so read the wording rather than the headline. The strong version names a trigger (acquisition or shutdown), a specific action (source released under a named open licence), and a timeframe. The weak version says the company values your data and intends to give notice. A pledge is not legally identical to open source — the company still controls the code today — but a written, dated, public commitment is much harder to walk away from quietly than an unstated intention.
How often should I export my data?
Once when you set the tool up, so you know what the export contains, and then roughly twice a year — or immediately after any ownership, pricing, or policy announcement. The point of the first export is not backup, it is inspection: you are finding out today what you would be left holding. Storing the file somewhere you control, in a format you can open, is the whole exercise.
What if the export is behind a paid tier?
Treat it as a serious mark against the tool, and export before you cancel rather than after. An export locked behind the subscription you are about to end is a hostage arrangement — legally you can usually still get the data through a GDPR access request in the EU, but that runs on a multi-week clock and produces whatever the vendor decides to send. The button, while your account is active, is faster and more complete.
Does any of this apply to spreadsheets and Notion?
Partly, and in your favour. A spreadsheet of contacts is about as acquisition-proof as it gets — the format outlives every vendor, and the file is already yours. Notion sits in between: the export works and is complete, but the structure you built with relations and formulas comes out flattened, so what you lose is the system rather than the data. If you are running a DIY setup, the risk you are carrying is upkeep, not ownership.
New for you
Not read yet
You are caught up here.