Open a typical recruitment desktop on a Monday morning and you'll usually find more than Bullhorn (or Salesforce) open. There's a document generation tool for contracts. A separate app for call recording and transcription. Maybe a data cleansing tool running in the background, a workflow automation platform, an compliance portal. Five logins, five vendors, five support lines, just to get through a single placement.
None of this happened by design. Nobody sat down and planned a five-vendor stack. It happened one point solution at a time: a problem came up, a tool got bought to solve it, and eighteen months later nobody remembers why there are three different systems that all touch candidate data.
This is stack sprawl, and it's quietly expensive in ways that don't show up on a single invoice.
Tech Stack spend has crept up everywhere, and recruitment is no exception. Recent cross-industry benchmarking puts average UK company software spend at roughly $209,000 a year once every subscription across the business is totalled, and that's before factoring in headcount growth (Cledara, 2026).
The uncomfortable part is how much of that spend is dead weight. On average, organisations waste over $135,000 annually on software licenses nobody is actively using (JumpCloud, 2025) — seats and subscriptions still being invoiced long after anyone logged in.
And the subscription count rarely goes down on its own. The median company today runs around 25 active software subscriptions (Cledara, 2026), most added individually over time rather than reviewed as a set. That's exactly how a stack tends to grow around your ATS: not through one bad decision, but through many small reasonable ones that were never revisited.
None of the numbers above are unique to any one agency, they're general benchmarks that illustrate the scale of a problem most recruitment businesses are living with in some form. The real question worth asking isn't "is this happening to other people," but "how many logins does my own team need to get through a single placement, and how many of them are we actually paying twice for."
That's usually the more useful audit to run.