The Revenue We Refused
The contract was there for the taking. We stopped pitching it.
June 2, 2026
The remediation contracts were there for the taking.
Large contracts, repeatable, predictable. A large organization discovers its master data has decayed — duplicate parts, orphaned records, descriptions no one can search. They commission a cleanup. A team arrives, works through the catalog, hands back a clean dataset. The invoice clears. Everyone moves on.
Two or three years later, the data has drifted back. The same organization commissions the same cleanup. Often from the same firm. Sometimes against a proposal that is, line for line, the one they signed before.
We stopped pitching that work.
Not because it was hard. We were good at it — it was how the company started. Not because the money was bad. The money was excellent, and recurring, and it arrived on a schedule you could plan a business around.
We stopped because winning it meant selling the disease as the cure.
The most profitable thing in the room
There is a version of this business that is very comfortable.
You build a remediation practice. You get fast at it. You develop tooling, playbooks, a bench of people who know how MRO catalogs fall apart. You sell a cleanup, deliver it well, and the client is genuinely happy — their data is better than it was.
Then you wait.
Because the thing about a cleanup is that it treats the symptom and leaves the mechanism untouched. The records were clean on the day of handover. But the system that let them become dirty in the first place is exactly as it was. New parts still get created the same way, by the same people, under the same pressure, with the same absent guardrails. The drift starts the day after go-live. It is slow, and then it is not, and in two or three years you are back where you began.
For a remediation vendor, this is not a flaw. It is the business model. The decay is the demand. We have seen firms build decades of revenue on exactly this cycle — and we understand the appeal, because for a while it was ours.
The uncomfortable part is what it asks you to want. To grow that business, you have to quietly prefer that the underlying problem never gets solved. You are not sabotaging anything. You are simply selling a cure that, by design, wears off — and counting on it wearing off.
We sat with that for a long time. Then we decided we could not build a company around hoping our customers’ data stayed broken.
What it cost to walk
The decision sounds principled written down. In practice it was a series of specific, expensive nos.
It meant declining engagements we could have won. When an organization came to us describing the symptom — we have fifty thousand suspected duplicates, can you clean them — the easy, lucrative answer was yes. The honest answer was a different conversation entirely: that a cleanup alone would buy them eighteen months, and that the money would be better spent making the next fifty thousand records impossible to create badly in the first place. Some of those conversations turned into better, deeper work. Many of them did not. The client had budget for a cleanup, not for a change in how they operated, and they went to a vendor who would simply take the cleanup money.
It meant a slower revenue curve. Recurring remediation is beautifully predictable. What we chose instead — prevention, built into the point where data is created — does not bill on a two-year decay cycle. When it works, there is no relapse to sell against. You are, in the most literal sense, engineering away your own follow-on revenue. We traded a known, repeating number for a harder, lumpier one.
And it meant building instead of billing. For years, the most valuable thing we did generated no invoice at all. We were putting fifteen years of watching catalogs fail into Ark — a platform built to prevent the failure rather than clean up after it. Dictionaries, rules, the logic that catches a bad record before it ever lands in the ERP. None of that was a line item. It was cost, for a long time, before it was ever product.
Why the long horizon justified it
The bet underneath all of it is simple. The remediation market is large because the problem is permanent. We did not want a share of a permanent problem. We wanted to be the reason it stopped being permanent.
That only pays off on a long horizon. On a short one, the remediation vendor wins every quarter — the demand is reliable, the delivery is known, the money is now. Prevention loses those quarters. It asks you to forgo certain revenue today for an outcome that, when it arrives, is invisible by definition: the duplicate that was never created, the cleanup that never had to be commissioned, the proposal a future CIO never has to read.
Invisible outcomes are hard to sell and harder to invoice. But they are the only ones worth building a twenty-year company around. Organizations tell us, eventually, that the absence of the relapse is the thing they value most — they just could not see it as a line item at the start.
We could not see it as a line item either. We refused the revenue anyway.
The remediation contracts are still there. They will always be there, as long as the mechanism that creates dirty data is left in place. Any firm that wants that revenue can have it.
We decided a long time ago that we would rather build the thing that makes the contract unnecessary than keep selling the contract.
It cost us. It still does, quarter to quarter.
It is also the only reason we are still here, fifteen years in — building Ark deeper into the same conviction instead of selling the same cleanup again.
We Build. We Stay.
About the Author
Raghu Vishwanath is Managing Partner at Bluemind Solutions, a product engineering firm specializing in MRO master data governance. He writes about software engineering, AI, and building platforms that last.