What the MongoDB Acquisition Actually Felt Like: Notes from the Founder Seat
Most acquisition stories get told in one of two ways. The first is the press release version: "We're thrilled to join the MongoDB family. This acquisition will accelerate our mission to..." The second is the tell-all memoir, written a decade later when everyone has moved on. Almost nothing gets written in between — the honest, specific, here's-what-actually-happened account from someone who was recently in the seat.
I'm going to try to write that second thing.
Realm was acquired by MongoDB in 2019 after eight years as an independent company. We had raised $40M across three rounds, grown to dozens of engineers, and built a database platform used in over 2 billion device installations. I was one of the co-founders. I don't want to write a press release about it.
How It Started
Acquisitions don't start with a meeting. They start with relationships that accumulate over years before anyone is formally interested in buying anything.
MongoDB and Realm had been in the same ecosystem for years. Our mobile database complemented their server-side database — developers building apps often used both. We had conversations with their business development team at various points. Nothing serious. They were watching what we were building, and we knew it.
By mid-2018, we had a functioning sync product: Realm Sync, which let mobile clients synchronize data with a MongoDB-backed server. We had built this deliberately. A local database is a nice developer tool. A local database with built-in sync to a major cloud platform is strategic infrastructure. We were making ourselves more attractive to a specific kind of acquirer.
The formal process began with a call that wasn't supposed to be a formal process. It was positioned as a business development conversation about potential partnership. The partnership conversation was real — MongoDB was genuinely interested in how Realm might integrate with MongoDB Atlas. But both sides knew what it might turn into.
The Process Itself
I am not going to describe the specific financial terms of our acquisition. But I can describe the structure.
Acquisition processes are exhausting in a way that's difficult to convey. You are simultaneously:
- Running your company (users don't care about your M&A process)
- Participating in due diligence (extensive, intrusive, and time-consuming)
- Negotiating deal terms (complex, stressful, requires specialized lawyers)
- Managing information within your team (not everyone can know what's happening, which creates guilt and anxiety)
You do all of this while pretending everything is normal. When your head of engineering asks why you've been in so many closed-door meetings, you say it's an investor conversation. When your sales team asks why the pipeline has slowed down, you tell them you're being selective about deal quality. You lie, essentially, by omission.
I don't love this part of the process. It's a structural feature of how acquisitions work — the information asymmetry is necessary to protect everyone — but it's genuinely uncomfortable to operate this way with people you respect.
Due diligence took longer than I expected. A large public company acquiring a venture-backed startup has to verify a lot of things. Legal: your cap table needs to be perfectly clean, every investor and option-holder identified, every equity agreement reviewed. Technical: they'll want to understand your codebase, your infrastructure, your dependencies, your security posture. Commercial: your customer contracts need to be reviewed, your key customer relationships assessed. Financial: your historical financials, your runway, your burn rate.
We had good lawyers and a relatively clean cap table, which helped. But we still spent months in this process.
The hardest negotiation was not about price. It was about what happened to our people. MongoDB was acquiring a team, not just software. They wanted our engineers to stay and continue building the platform. We wanted to make sure our engineers had good outcomes — both financially (the acquisition needed to be good for them, not just for preferred shareholders) and professionally (they shouldn't be acquired and then told to do something completely different).
Getting alignment on this required more back-and-forth than anything else. I'm glad we pushed on it.
What Surprised Me
How much the outcome depended on people, not documents. Our deal ultimately came together because there were specific people on both sides who wanted it to happen and pushed through the difficult moments. There were several points where it could have collapsed — a due diligence finding that required explanation, a negotiation where both sides dug in, a timeline slip that tested everyone's patience. In each case, it held together because of relationships that predated the formal process.
This suggests: cultivate relationships with potential strategic acquirers years before you're ready to sell. Not in a manipulative way — genuinely engage with the people building in adjacent spaces. The relationship you build at a conference in 2016 might matter a lot in 2019.
How disruptive it was to keep running the company. I had underestimated how much cognitive load the acquisition process would consume. There were weeks where I spent more time on acquisition-related work than on actually running Realm. This creates a real risk: your company gets worse because you're distracted, which makes the deal harder to close. We managed it, but there were stretches where I was not doing my job well.
How much I cared about the post-acquisition narrative. I knew intellectually that once the acquisition closed, the story of what Realm was would be told by MongoDB. Our brand, our product roadmap, our community — all of it would be shaped by how MongoDB chose to present it. I cared about this more than I expected to. We negotiated specific commitments about how the Realm brand would be maintained and how the open-source project would continue to be supported. Whether those commitments held is a separate question.
The emotional aftermath. The day the acquisition closed, I felt: relief, pride, grief, and disorientation, roughly simultaneously. Relief because the process was over and it had gone well. Pride because of what we had built. Grief because something I had devoted eight years to was no longer mine. Disorientation because I no longer had the daily structure of running a company.
I went from a year of extreme stress (the acquisition process) to a sudden absence of external pressure, and I had no framework for that transition. I spent the next six months doing advisory work, not because it was the right next thing to do, but because I needed something to fill the structure that running a company had provided. I should have taken more time to understand what I actually wanted before committing to the next thing.
What I'd Tell Another Founder Going Into a Process
Get an M&A lawyer who has done many deals, not just a good generalist corporate lawyer. M&A law is its own specialty. The first time your lawyer sees a particular provision type is not when you want them to encounter it.
Your board will have different interests than you do. This is not a criticism — it's structural. Your preferred shareholders are optimizing for return on investment across a portfolio. You are optimizing for this specific company and this specific team. These interests usually align well enough, but there are specific negotiating points where they diverge. Know where those points are before you get to them.
Due diligence will find things. Every company has things that look worse on close inspection than they look from a distance: a dependency on a library with a complicated license, a customer contract with an unusual clause, a security issue that's on the roadmap to fix. Most of these can be explained and resolved. But "can be explained" requires time and energy. Budget for it.
Think carefully about what happens to your most important people. The engineers and product managers who made your company valuable are likely not perfectly aligned incentive-wise with the acquisition. If key people leave shortly after close, the acquirer has less of what they paid for, and you have failed people who trusted you. Spend time on this.
The end is not the point. I spent so much of my energy in 2018 and 2019 focused on the acquisition that I look back on that period as almost exclusively about the exit, rather than about what we were building. The Realm I'm most proud of was being built in 2014-2016, when the team was smaller, the product was in a high-growth phase, and the only thing we thought about was making the best mobile database in the world. The acquisition validated that work. But the validation wasn't the point. The work was the point.
What Came Next
After Realm, I spent time as an advisor — the six months I mentioned, where I was filling structure rather than doing what I wanted. Then I started paying more attention to the industries where I kept finding the same fundamental software problems.
Pharmaceutical companies run compliance-critical processes on spreadsheets and email. Quality systems that can't talk to each other. The same regulatory problems that drove data handling in mobile — offline access, sync, audit trails — appeared again, but in a context where the stakes were higher and the incumbents were worse.
That's what Cohera is. But that's a different story.
The Realm story ends at MongoDB. The lesson I carry forward: the most valuable companies are built by people who fall in love with a hard problem, not by people who fall in love with the idea of building a valuable company. We didn't set out to build something MongoDB would want to acquire. We set out to fix mobile data. The acquisition was the consequence, not the goal.
If you want the technical side of what we built — how the Realm database engine actually worked, what we got right and wrong architecturally — I wrote about that in How Realm Works: The Database Architecture Behind 2 Billion Installs.