The headline figure on the HMRC contact center deal is hard to ignore.
A $670 million contract, awarded to Capgemini with NiCE CXone as the CCaaS engine, covering eight years of service delivery for one of the UK’s most politically exposed public sector organizations.
When you include VAT, that number climbs closer to $814 million.
Beneath the headline, though, is a more interesting story; one about how enterprise CCaaS procurement actually works at the highest level, who gets in the room, and why the same names keep coming out on top.
The context that made this deal necessary is about as stark as procurement context gets.
A 2024 National Audit Office report found that HMRC’s customer service had deteriorated badly, with taxpayers facing average telephone wait times of nearly 23 minutes in the first 11 months of the 2023-24 tax year.
More damaging still, an estimated 72% of those calls were driven by failure demand: people calling because a process had broken down, not because they needed genuine assistance.
The NAO’s conclusion was that taxpayers were being “let down.”
For HMRC, in many ways, this is as much about repairing the company’s reputation, as it is about implementing fresh technology.
A Field of Thirteen, Down to Two
According to the UK government’s official tender notice, the procurement process drew significant interest before narrowing sharply, with the final decision coming down to two vendors: NiCE and Genesys.
That outcome is, by now, fairly familiar in mega-enterprise CCaaS procurement. Despite the entry of AWS, Microsoft, and Salesforce into the contact center market – and the genuine momentum each has built – deals of this complexity tend to resolve down to the same two names.
Blair Pleasant, President and Principal Analyst at COMMfusion, is careful not to overstate what that means for the wider market:
“I think there’s still room for others. HMRC had some very specific requirements for what it was looking for, and I think NiCE and Genesys were the only real options.”
“But in general, I think there’s still more players in the market who can serve large enterprises.”
Pleasant is right that the CCaaS market isn’t a closed shop. Five9, Talkdesk, and Content Guru are all landing larger deals than they were a few years ago.
However, at the tier HMRC was operating in, the procurement criteria alone did a lot of the filtering.
According to the contract notice, bidders were required to carry at least $190 million in annual CCaaS revenues, and the contract couldn’t account for more than 10% of a supplier’s total CCaaS business.
Those conditions eliminated most of the market before a presentation was made.
The analyst validation that underpins both vendors’ positions at that tier is well documented.
Both NiCE and Genesys were named Leaders in the 2025 Gartner Magic Quadrant for Contact Center as a Service, for the 11th consecutive year each.
The Forrester Wave for CCaaS Platforms Q2 2025 tells a similar story, with NiCE ranked first in Strategy (4.14/5), and Genesys first in Current Offering (3.98/5).
At procurement committees where getting the decision wrong carries real political consequences, that kind of sustained, third-party validation carries more weight than it might in a lower-stakes technology purchase.
Where the Hyperscalers Stand
The HMRC process is also a useful reference point for where AWS, Microsoft, and Salesforce actually sit in the enterprise CCaaS market right now, which is somewhat short of where their overall scale might suggest.
Despite these hyperscalers making big CCaaS plays in recent times, there are still obvious shortcomings
Salesforce only launched native voice for its contact center offering at Enterprise Connect 2026, while Microsoft failed to appear in the 2025 Gartner Magic Quadrant evaluation at all.
AWS, however, is seemingly breaking the mold. The vendor is named a Leader in both the Gartner MQ and the Forrester Wave, recognized for its GenAI capabilities and consumption-based pricing.
Pleasant provided her own views on the reasons behind the varying fortunes of these would-be CCaaS powerhouses:
“Microsoft doesn’t have the same level of capabilities [as NiCE and Genesys], and it certainly doesn’t have the experience… HMRC was looking for the company to have a certain amount of revenue.
“Both Microsoft and Salesforce have the required revenue that HMRC was looking for. But if you’re looking just at the CCaaS piece, then they certainly don’t.”
What’s missing for most of the hyperscalers is reference deployments.
When a contract covers a government department of HMRC’s scale, where failure will end up in front of a parliamentary committee, procurement teams want evidence that a vendor has operated at that level before.
NiCE’s work with the Department for Work and Pensions (40,000 seats, 2025) and a $578 million deployment for Services Australia in 2024 – the largest contact center in the Southern Hemisphere – provided exactly that.
Of the newer entrants, like Gartner and Forrester, Pleasant sees AWS as the most credible challenger to that position:
“AWS has been doing this for a while. They’ve got a lot of customers, so they’re definitely ahead of the game in this area.”
The Part of the Deal People Are Missing
One of the easier details to gloss over in coverage of this contract is that NiCE didn’t technically win it; Capgemini did.
NiCE is the CCaaS platform, while Capgemini is the prime supplier, responsible for implementation, system design, workflow integration, ongoing support, and continuous optimization across the life of the contract.
As Pleasant explains: “It wasn’t NiCE. It was NiCE as part of Capgemini.”
At the scale of a deployment like this – spanning voice, webchat, AI agents, and a full migration away from aging legacy infrastructure – the systems integrator doing the delivery work is carrying as much of the weight as the technology itself.
The complexity of integrating a new CCaaS platform into the workflows and data systems of a government department handling tens of millions of citizen interactions annually is a different challenge from a commercial contact center deployment.
Capgemini also came into this with an existing relationship with HMRC and a deep track record across UK public sector work.
The procurement process wasn’t just evaluating the technology; it was evaluating who would be responsible for making it work.
Route 101, a Bristol-based SME, also secured a subcontractor role, having worked the DWP transformation alongside NiCE.
For smaller vendors looking at deals of this scale, there’s a signal in that. The route to involvement at this tier doesn’t always run through being the prime.
The Eight-Year Question
If there’s a detail in this deal that deserves more attention than it’s received, it’s the contract length: eight years, running from May 2026 to May 2034, with an option to extend by two further years.
As Pleasant quips, “eight years is like 80 years these days. It is a long time with the way technology is changing so quickly.”
And yet the rationale holds up. A deployment of this complexity isn’t something you rip out and replace every few years.
The global CCaaS market is projected to grow from $8.33 billion in 2026 to $30.15 billion by 2034, according to Fortune Business Insights.
If accurate, this period of rapid consolidation and change makes vendor stability a genuine procurement consideration in its own right. There are plenty of CCaaS vendors that look compelling today but whose long-term position is much less certain.
NiCE’s September 2025 acquisition of Cognigy added a conversational AI agent layer that directly addresses the contract’s scope for ongoing innovation and future capability development.
There’s a customer experience angle to the eight-year commitment that may get overlooked when the analysis focuses on vendor strategy.
According to Pleasant, “customers don’t want things changing on them all the time. Going to a new vendor would change things a lot. Stability for both HMRC and its customers is really important.”
For a government agency that spent years being publicly criticized for inconsistent service, that might be the most important part of the specification.
What the Market Should Take from This
Taken on its own, the HMRC deal is a significant contract win for NiCE. Taken alongside the DWP and Services Australia deployments – all secured within the past two years – it looks more like a deliberate accumulation of proof points in the government and regulated enterprise segment, the kind that makes each subsequent win progressively easier to secure.
The Gartner Critical Capabilities for CCaaS report, published September 2025, notes that platforms are now expected to deliver “AI-enhanced, scalable customer engagement across digital and voice channels.”
The HMRC specification maps directly onto that. The vendors that win at this tier aren’t simply the ones with the strongest platform on evaluation day; they’re the ones who have spent years building credibility through analyst recognition, large-scale reference deployments, and the right SI partnerships – the kind that gives procurement committees the confidence to sign an eight-year, nine-figure contract.
For the hyperscalers, every deal NiCE and Genesys close at this level compounds the distance.
A competitive CCaaS platform is one part of the problem. The other part is a track record that can’t be acquired overnight, regardless of the engineering resources behind it.
HMRC started with a large field of candidates and finished with two. The more useful question for the rest of the market is probably less about this specific deal, and more about what it would actually take to be one of those two next time around.
