Services meets software: why buyers of data businesses want both

Why do services groups buy European data consultancies?

International services groups buy local data consultancies for two things: certified depth on a major data platform, and a team in a new country. Most of these buyers are backed by private equity or listed, and they grow country by country and sector by sector.

Infinite Services is a clear case. Adastra's announcement pointed to its Snowflake Premier Partner status and its award as Snowflake's AI Data Cloud Partner of the Year for Central and Eastern Europe. It also said the deal strengthens Adastra in Poland, one of its key growth markets in Europe.

Infinite Services is not an isolated case, as these announced deals show.

Where a headcount was published, the consultancies bought had between 90 and 240 people. In almost every announcement, the buyer gives a platform, a partner tier or a country as a reason for the deal.

When I first speak to a potential buyer, the opening questions are always the same: which platforms the firm works on, where it's based, and what kinds of clients it serves, in which countries. Consolidators grow by deepening or widening their platform portfolio, so those answers decide whether the conversation goes any further. Only once they match do buyers ask about skill levels and partner status.

Pro tip from Audrius Sankalas: Before your first conversation with a buyer, put the answers to its opening questions on a single page: the platforms your team is certified on, where your people work, and which sectors and countries your clients are in. A clear page lets the buyer match you to its plans quickly and moves the discussion on to your people and partner status.

Why are services groups buying accelerators and software tools?

Services groups buy software tools, or businesses that combine them with consulting, mainly because a consultancy wins work on how quickly and how well its people deliver. Owning the IP is part of the reason, but competitive edge is the larger part.

Most of the software in question is accelerators, which are tools that do repeatable data work, such as migrating or checking data, faster than a team can by hand. Consultancies used to build them for one client and one job, and nobody treated them as products. Firms now package their accelerators, and some combine them into a suite or a platform. These tools make a delivery team quicker and are rarely sold to end users on their own.

Syniti is the clearest example of this route. Capgemini completed its purchase of Syniti on 3 December 2024, and tied the deal to client demand for complex data migrations. Syniti sells enterprise data management software alongside services for data quality, migration and governance. Capgemini said Syniti brought a global team of more than 1,200 data experts.

In Germany, SNP, which sells both software and consulting for SAP data transformation, bought Datavard, a software and consulting company, and said the two fitted together on both counts. Carlyle completed its takeover of SNP in April 2026.

InfiniteData sits further along the same line. AutomateNOW! is a product in its own right, and Enterprise Management Associates named it a technology leader in its 2023 Radar Report for Workload Automation. Beta Systems said the deal expanded its market position in enterprise automation. It kept the InfiniteData engineering team and has since relaunched the platform as the ANOW! Suite.

Mark Sapsford, who advised on both deals with me, puts it this way:

'Software and consultancy have sat side by side for years, so it's no surprise the deals are following. A buyer looks at a consultancy's tools in one of two ways. Either they help the team sell and deliver faster, or they keep earning after the project ends, because the client carries on running what was installed. The test of whether they're real is simple. Does someone other than the founder use them, on clients the founder doesn't touch, and does the company own them? If the answer lives on one person's laptop, it's a slide in the deck.'

Are software and AI companies buying data consultancies?

Software and AI companies have started buying data consultancies in 2026, with the AI companies moving first. Until this year, software vendors buying consultancies was the least common of the three routes in data.

There are only a few earlier examples in data. Teradata bought UK consultancy Big Data Partnership in 2016 to build its services business across the UK, France, Germany and the Nordics. In 2024, Datavant, a health-data company, bought UK privacy consultancy Trace Data as part of its push into Europe.

The pattern changed in May 2026, when OpenAI launched the OpenAI Deployment Company, backed by more than $4bn from 19 investment firms, consultancies and systems integrators. To staff it from day one, OpenAI agreed to buy Tomoro, an applied AI consultancy founded in 2023 and based in Edinburgh and London. OpenAI said the deal would bring about 150 forward-deployed engineers and deployment specialists into the new company.

In July 2026, the new company agreed to buy Northslope, an applied AI firm founded by former Palantir staff. Ode with Anthropic, the enterprise AI services firm Anthropic set up with Blackstone and Hellman & Friedman, bought Fractional AI in May 2026 and Casper Studios in August 2026. Among data platform vendors, MongoDB said in May 2026 that it had bought Clarity Business Solutions, a services firm that had been one of its partners since 2021.

The reason is the gap between a pilot and production. Gartner research found that by the end of 2025, at least half of generative AI projects had been abandoned after proof of concept. Gartner's analysts point to unclear business value and rising costs among the causes, and closing that gap takes people working inside the client's business as well as better models.

Neither OpenAI nor Anthropic bought a traditional consultancy. Both bought firms built around engineers who sit inside the client's business, work out where AI will pay, and then build it. Others are building their own teams. In June 2026, Databricks formalised a forward deployed engineering organisation, and Amazon Web Services (AWS) committed $1bn to a unit that puts its engineers inside customer teams.

Most software vendors in European data still deliver through partner networks. That is one reason partner status turns up so often in consolidators' announcements, because the partner network is where the delivery capacity sits. Some vendors have now started buying that capacity directly, and MongoDB's purchase of a partner it had worked with since 2021 is the plainest example.

I haven't yet come across a deal in the Nordics, the Baltics or Central and Eastern Europe where an enterprise software company has bought a professional or managed services business. The cases so far are further west, with Tomoro in the UK among them. I'm watching whether they reach those markets, and whether the data platform vendors follow the AI companies.

Mark sets out what this means for UK founders:

'Tomoro is the deal UK founders should notice. It was a young Edinburgh firm, bought for its people: engineers who sit inside a client's business and get AI into production. For a software or AI company, a consultancy like that is a ready-made way to sell and to deliver. What they'll look for in a UK firm is what any buyer of a consultancy looks for: a team that delivers without the founder in the room, and clients who'd stay if the founder didn't.'

What should a data business founder do before talking to buyers?

A data business founder should build for clients first, keep partner status current and evidenced, make accelerators provable and company-owned, and reduce the firm's reliance on the founder. Those steps help a founder reach the buyer who values the firm most, on terms that keep the team and the client base together.

Four things follow, whichever side of the line your business sits on.

• Build for your clients, and the buyers will follow. The things that win you work are the things consolidators ask about, namely certified depth on a major platform and tools that make your team faster. A product business can draw software groups and services groups alike, as InfiniteData went to a software group and Syniti to a services group.
• Treat your partner status as evidence of what the firm can deliver. It's rarely a buyer's first question, but buyers check it and quote it in their own announcements, so keep it current, documented and backed by named, certified people.
• If you've built accelerators, make them provable. A buyer will want to see who uses them, what they save, and whether they work without your founding team in the room.
• If your team takes AI from pilot into production, there's a new kind of buyer. The AI companies have shown they'll buy firms whose engineers work inside client teams, and Tomoro shows that includes UK firms.

If you want your accelerators to count when you sell, make them assets of the company. That starts with clear IP and clear ownership, and the more productised the tools are, the better, which applies equally to proprietary frameworks and automation tools.

We would rather tell a founder to wait than take a firm to market before its tools and client relationships can be evidenced, because a buyer will find the gaps in diligence and price them. Our guide to how CapEQ runs a business sale explains the process once the groundwork is done.

Mark's advice to UK founders is about timing:

'Start 18 months to two years out, and get yourself out of the client relationships first. In a data consultancy the platform badges are easy to check. What a buyer can't see on paper is whether the clients belong to the firm or to you. Put a second senior person on every major client and let the team win work without you. Where a client still runs software you put in, get the support onto a contract. Buyers can tell the difference between a relationship that's been handed over and one that's only been described as handed over.'

What do buyers check in a data business before they buy?

Buyers check ten things in a data business, grouped by who is asking: consolidators, services groups, AI companies and every buyer. Start evidencing them 18 months to two years before a sale, because several take that long to show.

1. Certified depth on at least one major data platform.
2. Current partner status, backed by named, certified people.
3. Clear answers on platforms, location, clients and countries.
4. The company, not the founder, owns every tool and its IP.
5. Evidence of who uses each accelerator and what it saves.
6. Tools that work without the founding team in the room.
7. Engineers inside client teams who get AI into production.
8. A team that wins and delivers work without the founder.
9. A second senior person on every major client.
10. Support contracts for software you installed at clients.

Items one to three are what consolidators test first, items four to six matter most to services groups buying accelerators, and item seven is what the AI companies are buying. Items eight to ten apply to every buyer, and they take longest to put in place.