Your software vendor just got acquired. Five questions to ask

On 2 September 2026, Accrual announced it was acquiring Puzzle's accounting-firm technology and business, including the AI-native ledger and the month-end close products. Puzzle's founder Sasha Orloff and members of his team are moving across. Terms were not disclosed, the transaction is expected to close within weeks, and Accrual expects broader availability of the combined product by the end of 2026.

If you have never heard of either company, read on anyway. The deal is a clean worked example of something that will happen to a system your firm depends on, probably within the next two years, and almost every firm handles it badly for the same reason: the questions that matter have to be answered before the announcement, not after it.

What actually happened, because the shape matters

This is not a straightforward "company bought, product continues" deal, and the details are where the lesson is.

Puzzle, founded in 2019 and serving more than 7,000 startups, small businesses and accounting firms, has been split. Accrual bought the firm-facing side: the technology, the business, the team and the close product that Accrual co-founder Cosmin Nicolaescu singled out as the reason for the deal. The startup and small business side was not acquired. It continues independently under a new chief executive, Ozgur Uzuner, and will stop marketing a standalone product to accounting firms, moving instead to working through firm partners.

Accrual itself is nine months old. It launched in February 2026 with 75 million dollars in funding led by General Catalyst, sells to large firms including Armanino, Aprio, H&R Block, Creative Planning, BMSS and Stephano Slack, and had until now been a tax and practice technology company rather than a client accounting one.

So if you are a firm using Puzzle's firm product today, four things changed at once. Your vendor changed owner. Your product changed roadmap. The team you knew changed employer. And the company you originally bought from is now, in this segment, a partner channel rather than your supplier. Any one of those is manageable. All four in one press release is a migration, whatever the announcement calls it.

The five questions, and when to ask them

Ask these of every core system in your stack now, while nothing is happening. The answers take an afternoon to collect when there is no deadline, and they are close to impossible to get calmly in the four weeks after a migration email arrives.

1. What exactly was bought, and is my product inside it?

Acquisitions of software companies are frequently acquisitions of a part: a product line, a customer segment, a team, sometimes just the code. The Puzzle deal splits a company along a customer boundary, which is common and rarely obvious from the headline. Find out which side of the line you are on. "We acquired Puzzle" and "we acquired Puzzle's firm business" are very different sentences for a firm holding a subscription.

2. Where does my data live, and can I get it out in a form I could actually use?

This is the one firms consistently assume is fine. An export exists, therefore the data is portable. Test it. Export a real client file and look at what comes out. A CSV of transactions without the mapping, the history, the attachments or the reconciliation state is not your ledger, it is a receipt for your ledger.

Do this once a year regardless of any acquisition. It is the cheapest insurance in your stack and it takes an hour.

3. Is my contract assignable, and what happens at renewal?

Most software agreements permit assignment to an acquirer without your consent, which means the counterparty can change while your obligations do not. Worth knowing before it happens rather than reading it for the first time in a forwarded email. The renewal after an acquisition is where terms move, so find your notice period now and diarise it.

4. What is the migration, and who is doing the work?

Vendors describe migrations in terms of data. Firms experience them in terms of hours: retraining, rebuilding templates and workflows, re-checking a period, explaining a new interface to clients. That labour is real, it always lands on your team, and it is almost never in the vendor's project plan. Estimate it in staff days before you agree to a timeline, the same way you would scope any other project that lands on your own people.

5. What is my walk-away, and when do I have to decide?

The point of asking is not to leave. It is to know what leaving would take, so you are negotiating from a position rather than a hope. A firm that knows it could move in ninety days is a different customer from one that could not move at all, and vendors can tell the difference.

Why this keeps happening now

The pattern behind the Puzzle deal is worth naming, carefully, because it predicts the next few.

A well-funded platform company is trying to cover tax, audit, client accounting and advisory in one system, and buying a capability is faster than building it. Accrual said as much: tax and client accounting still live in separate systems with different vendors and different logins, and every engagement that crosses between them rebuilds context that already exists. That is a real problem in a lot of practices and it is a reasonable thing to go after.

What it means for buyers is that the point-solution part of your stack is now, in principle, acquirable. Not doomed, and not necessarily worse off afterwards. Some acquisitions genuinely improve a product. But the assumption that the tool you chose in 2024 will still be sold by the same people, on the same terms, to the same customers, is not a safe basis for a five-year plan. Firms already carrying too many disconnected tools feel this hardest, because each one is an independent chance of the same email arriving.

None of this argues for buying only from large vendors. Large vendors discontinue products too, and they do it with less warning because you are a smaller share of their revenue.

What to do this week

1. List your core systems: ledger, practice management, tax, portal, payroll. Five to eight names, not thirty.
2. Against each, write the owner, roughly how long they have been independent, and whether they have taken outside investment. Public information, fifteen minutes.
3. Run the export test on the two that hold the most client data.
4. Find your notice period for each and put the dates in a calendar with a real reminder.
5. Pick the one system whose disappearance would hurt most, and write a single page on what you would do. That page is the whole exercise. Everything else is preparation for writing it.

Firms that keep this current treat it as part of practice management rather than as a technology chore, and it usually lives alongside the rest of the system documentation in Xero Practice Manager or Kloud Connect. The firms that get caught out are not the ones with the wrong tools. They are the ones who had never written down what they would do, and then had four weeks to work it out during a busy period.

For more on how the technology decisions in a firm are shifting, our AI for accounting firms hub collects the rest, including when building something yourself is the better answer.

Frequently asked questions

My firm uses Puzzle. What happens to us?

It depends which product you are on. Accrual is acquiring the firm-facing technology and business, and expects broader availability of the combined offering by the end of 2026. The startup and small business side continues independently under a new chief executive and is moving to work through accounting firm partners rather than selling a standalone firm product. Ask your account contact in writing which entity holds your agreement after the close and what the migration path is, and keep the answer.

Can a software vendor transfer my contract without asking me?

In many software agreements, yes. Assignment clauses commonly permit transfer to an acquirer of the business without customer consent. Your protection is usually the notice period and the renewal date rather than a veto, which is why knowing both matters more than the clause itself. Check your own agreement, and take advice if a system is genuinely critical to you.

Is consolidation in accounting technology good or bad for firms?

Both, and which one depends on your position rather than on the trend. Fewer systems that share context removes real friction from engagements that cross between tax and client accounting. It also concentrates your dependency and reduces the number of places you can go if the product moves away from what you need. The useful question is not whether consolidation is good, it is whether you would still have options.

How often should we run this review?

Annually for the full list, and immediately whenever a vendor announces funding, an acquisition or a leadership change. The trigger events are public and the review takes an afternoon, so there is no good reason to be reacting to a migration email instead.

Should we delay buying new software because of this?

No. Waiting for the market to settle means waiting indefinitely and working with worse tools in the meantime. The adjustment is at the point of purchase: check the export before you sign rather than after, read the assignment and notice terms, and prefer vendors who can show you your data leaving in a usable state. That is a five-minute change to how you buy, not a reason to stop.