Buying software isn't a vibe

By Damon Anderson, 13 October 2026. 6 minute read.

I've reviewed 426 software companies for the A to Z of AI Accounting Software. A number that is meaningless to anyone buying software. Because in reality, you're only looking for one: the right one for what comes next for you.

And finding the one means kissing a lot of frogs before you meet your prince or princess.

And there are plenty of frogs. Every vendor has vibe-coded a slick website, a demo and an impressive-looking bit of kit. Showroom polish that, in many cases, hides a bag of spanners under the hood.

Finance leaders and accounting professionals are hitting AI fatigue. They're skipping events, switching to silent mode and sticking their fingers in their ears waiting for the droning AI twaddle factories to stick a sock in it.

And fair-a-bloody-nuff too.

Remember the good old days?

It wasn't long ago that working out your company's finance stack was easy. Xero or QuickBooks as the ledger. Bolt on Dext or Hubdoc for data and receipt capture. Something useful as a practice management tool and you're pretty much good to go. Your ledger's ecosystem did most of the curating for you. If an app was in the marketplace, had a stand at Xerocon and enough other firms swore by it, that was good enough.

Well, the times they are a-changin', son. Every player is now scrambling for relevance as the LLMs (the large language models behind tools like Claude) are working their way through the adoption curve to become the daily workspace. Software vendors have turned up to the gold rush, armed with a Claude subscription and a business case to capture their share of your hard-earned cash before Claude eats the world. One accounting practice at a time.

Most new players are cooked before they even start

So AI has made software cheap to build. But building an effective strategy is still hard. We're now seeing a flood of products with near-identical propositions, all competing for your attention, largely in the same way.

While building a great product is still hard, it's a lot easier than it was. Building a company that will still be around, supporting you and protecting your clients' data in two years' time is something else. I don't believe all 426 software companies will exist by then. Many will simply stop, and the investment data backs this up. Early-stage punts are becoming few and far between. The money has concentrated. In the first half of 2026, 87.5% of US venture dollars went into rounds of $100m or more, and the mid-market's share roughly halved in a year. Meanwhile the floodgates are open. New App Store submissions jumped 84% in the first quarter of 2026, while downloads barely moved.

As a revenue operator, I've watched this play out before in the sales and marketing tech space. The revenue tech stack is awash with software and as a result, choice paralysis. In 2011 there were 150 marketing technology products. By 2024 there were 14,106. The market hasn't collapsed. It has stratified. A few core platforms like HubSpot and Salesforce became the foundation everyone connects to, and thousands of smaller tools grew around them.

Accounting is heading the same way. The big platforms like Xero and Intuit will stay big, as their networks, built through years of hard graft and personal relationships, remain as strong as ever. Around them will live a long tail of specialist tools, some on life support, others carving out a niche that is hard to replicate. The challenging part for finance professionals is to figure out which new solutions are here to stay.

How to separate the wheat from the chaff

You don't need to be technical to ask good questions. Four worth asking of any vendor:

Does it solve a problem I actually have? Not the problem the demo was designed around, but does it help with your biggest business priority.
Who's behind it? A great product with no support, no funding, no roadmap and no people is a risk to your clients.
Where is my data held, and who can see it? Look for data residency you're comfortable with, SOC 2 or ISO 27001, and evidence of regular penetration testing. If a vendor sidesteps that conversation, that tells you something.
How does it connect to the tools I use today? In 2026, a product that isn't connected through APIs and MCPs is increasingly a product nearing end of life. Can it safely read and write back to the tools where you hold your client and financial data? Accounting data has to be right. A tool that writes into your ledger needs to be trustworthy and auditable.

The A to Z already scores every vendor on how well they connect through MCP. Trust and security scoring is coming soon, and will feed into the World Rankings from November, so firms can see the difference without needing to dig for it.

A word for the software vendors

If you're building in this market, I hate to break the news to you, but the software you ship and your "first-mover advantage" are no longer your moat. They're fast becoming the easy bit. Distribution is harder, because buyers are tuning out the outbound calls, your social ads, your pay-to-play awards and the exhibition stands. What can't be copied over a weekend is the relationships you build, the communities you serve and how much they trust you. Build that.

Where to start

Somewhere in those 426 is the one that counts for you, whether you're a business, an accounting practice or an investor. Finding it doesn't need to be overwhelming, and it doesn't need to wait until the next conference. It does need a starting point, and that starting point isn't necessarily searching for technology solutions in ChatGPT.

To start with, it's about being honest about what you want from your business. Do you want to grow? Buy? Sell? Create an AI-first accounting firm? Optimise what you've got? Cash in your chips and head to the Maldives? Every business is different, and understanding your own goals is the first job.

Then, and only then, start understanding the technology choices. What the hell AI actually does. When it should be used and when (which is often in accounting) it shouldn't. What an MCP is. And why it's important to know the difference between read-only access and write access with an audit log. What AI-native really means and whether the software you use today will survive the next five years. Decide which tools make sense for where you want to go, not where a vendor wants to take you.

Get the order right and the 426 quickly narrows down to a manageable shortlist. Get it wrong and you're buying software for software's sake.

The goal is choosing with confidence and feeling good about the choice you made. Because building a company is still a people business. Thankfully. And that can't be vibe coded on the weekend.