In multi-site rollups, SaaS waste rarely comes from one bad contract. It comes from license levels set site by site, seat counts nobody revisits, and decentralized purchasing that leaks money one user at a time.
Somewhere around the fourth or fifth bolt-on, the operating partner asks a question the CFO can’t quite answer: why does software spend per site keep going up? Headcount per location hasn’t changed much. No one signed a big new enterprise contract. The major vendors were renegotiated during diligence. Yet every quarter, the per-site SaaS number drifts a little higher, and nobody can point to where it’s coming from.
The answer is rarely in the contracts. It’s in the licenses. For platform companies running a buy-and-build strategy, decentralized purchasing is the default state after every acquisition, and it leaks money one seat at a time. Vendor consolidation and procurement centralization are usually pitched as contract exercises. In a multi-site rollup, the bigger win sits one level down.
How every site ends up buying its own seats
No one decides to run twenty separate software budgets. It happens because each acquisition arrives with its own stack, its own admin, and its own habits, and the integration plan has bigger priorities in the first 100 days.
The acquired practice keeps its productivity suite tenant because migrating email in month two feels risky. Its office manager keeps buying CRM and scheduling seats on a company card because that’s how it worked before the deal closed. HR adds payroll and recruiting licenses for new hires locally, and nobody removes the ones belonging to people who left during the transition. A few clinicians need design or document tools, so someone upgrades them to a premium tier and never looks back.
Each of those decisions is reasonable in isolation. Stacked across a dozen or more locations, they produce a specific kind of SaaS sprawl: the same vendors, bought many times over, at the wrong license level, with no single owner watching seat counts.
That last part matters. The platform company may already have a master agreement with the vendor. The bolt-on just isn’t using it, or it’s using it with a seat count and tier mix that made sense for someone else’s org chart. Contract terms get negotiated once, at the center. License levels get set everywhere, by everyone, and they drift.
The proof: 11 of 26 findings, none of them dramatic
Consider one multi-site specialty healthcare operator in the Saasrooms customer base. When Saasrooms mapped its software estate, the vendors under review carried roughly $2.1 million in annual recurring spend. For a group operating across many clinic locations, that’s a modest number. Nobody would call it a runaway SaaS budget.
The review surfaced 26 savings opportunities. Eleven of them were license-level findings: seat counts that no longer matched headcount, tiers above what users actually needed, and duplicate license pools for the same tool.
License levels accounted for $43,752 of the identified savings, about 62% of the total. Contract terms came next, with ten findings worth $17,529. Three supplier switches added $7,381, and two payment-term changes brought in another $2,344.
So license optimization produced more findings than any other category and close to two-thirds of the money. The more telling detail is how small most of those findings were.
The median license-level finding was worth about $1,300 a year. Eight of the eleven came in under $2,500. They spread across payroll, recruiting, design software, cloud infrastructure, clinical research tools, identity management, and a specialist healthcare IT vendor. The CRM platform showed up twice, as separate license findings on two separate products. Only the productivity suite ($16,000) and one CRM product ($10,000) cleared five figures, and both adjustments have already been completed.
Meanwhile, the single largest contract in the estate, the electronic health record platform at over $800,000 a year, has so far produced no quantified savings at all. It’s still in scoping.
That’s the signature of this pattern. Many spend reviews are driven by one oversized contract where a single renegotiation carries the whole business case. This is the opposite. There’s no headline number here, just a long tail of small leaks, each too minor to trigger an escalation on its own, together worth more than every contract-term finding combined.
For a CFO, the obvious question is whether $43,752 is worth anyone’s time. Look at it the way a buyer will. These are recurring savings, so they show up in run-rate EBITDA every year, not once. At a 10x to 12x exit multiple, roughly $44,000 of recurring license savings adds somewhere between $440,000 and $525,000 of enterprise value. And that comes from a single operator with modest spend. Repeat the same review at every bolt-on in a buy-and-build plan, and the long tail becomes a line item worth defending in front of the board.
Why license waste stays invisible in a consolidated view
Finance teams at platform companies usually track software spend at the vendor level. That’s the view the general ledger produces, and it’s the view the board sees. It answers one question well: who are we paying, and how much?
It doesn’t answer the question that matters in a rollup, which is whether the seats behind each payment are still earning their keep.
A $1,300 overage on a payroll platform doesn’t move a consolidated P&L. Neither does an extra tier on a recruiting tool, or a cloud subscription sized for a site that has since merged with another. Rolled up, those lines look like normal vendor spend. Split by location, they look like exactly what they are: licenses bought by one site, for one moment, that nobody revisited.
A few things keep it hidden longer than it should be. Ownership is split: central IT or procurement holds the vendor relationship, while site managers decide who gets a seat, so neither side sees the whole picture. Most procurement reviews only kick in above a spend threshold, and individual license findings sit comfortably below it.
Then there’s the data problem. Contract values live in finance systems, while logins and activity live in SSO logs, admin consoles and browser data. Until someone joins the two, a paid seat and a used seat look exactly the same.
This is where site-by-site usage visibility changes the conversation. Saasrooms ties each license back to the location and the person holding it, then flags the seat when activity drops off, which is how a finding worth $900 surfaces at all. Without that join, it simply renews.
The cost of not looking compounds. Every new bolt-on inherits the same blind spot, and the per-site creep the operating partner noticed is what that blind spot adds up to.
A procurement centralization playbook for bolt-on integration
Fixing this doesn’t require a heavy-handed ERP rollout or a freeze on local purchasing. It requires treating license management as a standing part of the integration plan, not a clean-up project someone gets to eventually.
- Inventory and map every bolt-on in the first 30 days. Before any vendor consolidation talks start, list what each acquired site pays for, including card spend and expense-reimbursed tools, and tie every license to a named user and a location. Shadow IT tends to be heaviest in newly acquired entities, where old approval rules no longer apply and new ones haven’t landed yet.
- Fold local subscriptions into the platform’s master agreements. Once you can see three sites buying the same tool at list price, the vendor consolidation case writes itself. Right-size the combined seat count before the move, not after.
- Set license levels by role, centrally. Decide once which roles get which tier and apply it at every site. This is where most of the long-tail savings live, and it’s the step decentralized teams almost never take on their own.
- Review small findings together, on the renewal calendar. A $1,000 finding doesn’t justify a meeting. Eleven of them do. Batch license adjustments into a quarterly review and flag every renewal 90 days out, so seat counts get trued up before auto-renewal locks in last year’s number. Saasrooms puts each renewal date next to current utilization to make that conversation easy.
- Build it into the onboarding template for the next deal. The playbook should run the same way at acquisition 14 as it did at acquisition 4. Standardize it, and each bolt-on arrives into a system instead of a scramble.
The EBITDA impact isn’t dramatic in any single quarter. It is durable, though, and it scales with the number of sites, which is exactly the variable a buy-and-build strategy is designed to grow.
One site’s fix doesn’t scale itself
Most platform companies have already solved this problem once. Somewhere in the portfolio, a sharp site manager cleaned up their seat counts, cancelled the unused tier, and saved a few thousand dollars a year. It worked.
It didn’t spread, though, because nothing made it spread. The next site never heard about it, and the site after that was acquired six months later with a fresh set of habits.
That’s the real lesson from a portfolio where 11 of 26 findings sat at the license level. The fix itself is simple. Keeping it applied across every location, through every hire, every departure, every renewal and every new bolt-on, is not. Without someone owning SaaS spend management centrally, with the data to see it site by site, the savings found at one location quietly leak back out everywhere else.
If you’re integrating bolt-ons this year, start with one exercise before the next renewal cycle or the next close: pull license counts site by site and compare them to who is actually logging in. That single view usually tells you whether your per-site creep is a contract problem or a license problem, and who needs to own the fix.
