Can You Actually Leave?

A client of ours was on a legacy tier two CRM. Decent enough on paper. The problem was that the CRM side and the service and ticketing side didn't talk to each other, despite being sold as one system. Same vendor, same login, two separate worlds. So the people selling to a customer and the people supporting that customer were looking at different records, and nobody could answer a question that spanned both.

We moved them to HubSpot. We're a HubSpot partner, so discount that accordingly. Fine, that bit is not interesting.

Here is the bit that is. The old system had no export feature. None. You could not get your own data out of it in any sensible way. What they ended up with was 5,000 files sitting on an SFTP server, which is not an export, it's a shrug.

So I wrote a custom Python app to pull those files down and push them into HubSpot through the API. Four calls per file, because the dependencies all had to exist in the right order before anything would stick. The migration itself ran in hours. Not days, not weeks. First time, no drama. We did have to run it again a week or so later when the client found another pile of files somewhere else, which is exactly the sort of thing that happens and you should budget for.

What they got on the other side was everything in one place under one data model. Reportable KPIs across the whole client lifecycle, from first touch as a prospect through to a long term supported customer. That mattered enormously, because they were moving into a new stage of growth and bringing on additional revenue channels, and you cannot run that blind.

Five Point Six Six Billion

On 20 October 2026, Salesforce announced it was buying Braze for $5.66 billion in cash. $47.50 a share, a 58% premium on the previous close. Braze has somewhere around 2,770 customers. The deal is expected to close in Salesforce's fiscal 2028, with Braze folding into Marketing Cloud, Data 360 and Agentforce.

Braze went public in 2021 at $65 a share, for whatever that's worth as a comparison.

Now, the commentary on this has mostly been about the price and what it means for Salesforce's agentic strategy. Fair enough, that's a story. It's just not the story that matters if you're a founder with 2,770 of you sitting in that customer base wondering what happens next.

And I want to be careful here, because the obvious take is "acquisitions are bad, avoid acquired vendors", and that's nonsense. Salesforce buying Braze might well be great for Braze customers. Better integration, more investment, a bigger roadmap. I genuinely don't know, and neither does anyone else writing about it this week. The deal doesn't close until fiscal 2028.

So what's the actual lesson? Not "don't use Braze". It's narrower than that, and more useful.

The Question Nobody Asks At Procurement

When you buy a revenue system, you ask about features, price, integrations, support, maybe the roadmap. All sensible.

Almost nobody asks the one that bites: if I want to leave in three years, can I take my data with me, in a usable shape, without writing software?

That client of ours answered that question the hard way. By the time you're discovering there's no export, you're already trapped, and the cost of leaving is now a bespoke development project. We happened to be able to do it in hours because I can write the code. Plenty of businesses in that position would have been quoted weeks of agency time, or more likely would have just stayed put and kept paying for a system that didn't work, because moving looked too expensive.

That is what vendor lock-in actually is. It isn't a contract clause. It's a technical fact about whether your data will come out.

Three Things Worth Checking This Week

Not a framework, just three things I'd want to know if I were running a business on a stack I didn't own.

First, for every system holding customer data, find out what the export actually produces. Not what the sales page says, not what your account manager says. Ask for a sample. If the answer involves a support ticket, an SFTP server or a professional services quote, you have your answer.

Second, work out which of your tools is a feature rather than a product. Braze does real time cross channel engagement, which is a genuinely good thing to be excellent at and also a thing every large platform wants to own. If a tool does one job brilliantly and that job sits inside somebody bigger's roadmap, assume it gets bought eventually. That's not a reason to avoid it. It's a reason to know where the data lives.

Third, be honest about whether your systems talk to each other at the data model level or just at the logo level. Our client's two halves were sold as one product and shared nothing. You find that out when you try to report across the lifecycle and the numbers don't reconcile.

The Landing

The $5.66 billion isn't the thing to pay attention to. The thing to pay attention to is whether you could leave any of your revenue systems next quarter if you needed to.

Most founders have never checked. The ones who have usually checked at the worst possible moment.

Common questions

What is vendor lock-in? Vendor lock-in is not a contract clause. It's a technical fact about whether your data will come out of a system in a usable shape. A contract you can negotiate or wait out. A platform with no working export leaves you with a choice between paying for software that doesn't fit and funding a bespoke development project to escape it, and that second option is what most businesses discover only once they've decided to go. The question is not what the agreement says. It's what the export produces.

How do you check whether you can export your CRM data? Ask for a sample export, and ask before you sign rather than after. Not a description of the export on a sales page, not an assurance from your account manager, an actual file you can open. What you want is every object you care about, with the relationships between them intact, in a format something else can read. If getting it requires raising a support ticket, if it lands on an SFTP server as thousands of loose files, or if the answer comes back as a professional services quote, that is your answer and you have it while you still have a choice.

What happens to your data when a software vendor is acquired? Usually nothing bad, and anyone telling you otherwise in the first week is guessing. Acquisitions often mean more investment, better integration and a bigger roadmap, and large deals frequently take more than a year to close before anything changes at all. The risk isn't the acquisition itself. It's discovering, at the point you decide the new owner's direction doesn't suit you, that leaving was never actually an option. Know where your data lives and what it takes to get it out, and the question of who owns the vendor matters a great deal less.