A production-data residency requirement can stop a SaaS contract even after the buyer, budget and product fit are settled. Founders selling across African markets need to discover that requirement before the final procurement call, then decide whether the revenue justifies a separate hosting path.
At 4:12 on a Friday afternoon, Kojo joined a video call from his small office in Accra with a mug of coffee he had already reheated twice. He had spent weeks answering product questions from a South African enterprise buyer. The commercial lead liked the software. The operating team wanted the rollout. Kojo expected the call to settle the start date.
This is an invented composite, but the decision is familiar. Twelve minutes into the call, someone from procurement asked where production data would be stored. Kojo named the region his platform currently used. The answer came back plainly: production data had to remain in South Africa before the contract could be signed.
The deal had reached a different kind of no
Kojo could support the buyer’s workflow. His team had demonstrated the product, discussed onboarding and worked through the commercial terms. None of that answered the new question.
Could he operate the product with production data hosted in-country?
For a founder with a small engineering team and limited runway, that question reaches far beyond changing a cloud setting. A separate hosting path may affect deployment, monitoring, backups, incident response, access controls and every third-party service that touches customer data. It also creates another environment the team must maintain after the excitement of closing the deal has passed.
The bad ending was now visible. If Kojo said yes too quickly, he could sign an obligation his team had not costed and could not reliably support. If he hesitated, the buyer could move to a supplier whose hosting arrangement already met the requirement.
The contract was close enough to feel won. It was still unsigned.
A hosting answer is also a product decision
Kojo’s first instinct was to ask his engineer how quickly they could create a South African environment. That was the wrong opening question.
Speed mattered only after scope.
He needed to know which data had to remain in-country, whether backups fell under the same condition, which vendors processed that data, who could access the production environment and what evidence procurement expected before approval. A confident deployment estimate without those answers would be theatre.
This is where founders often discover that a sales request has become a product architecture decision. Supporting one buyer may introduce a permanent branch in the system. Every future release must work there. Every incident must be diagnosable there. Every new analytics, support or AI provider must be checked against the same boundary.
Kojo had seen a version of this problem before: approval arrived late, while the cost of waiting continued. The harder lesson in what happens when compliance approval arrives after your cash runway ends is that a commercially attractive deal can still be mistimed for the company carrying it.
So he did not promise a date on the call. He repeated the requirement in concrete terms, separated confirmed facts from open questions and asked for a technical session with the buyer’s security and data teams.
That request kept the deal alive without turning uncertainty into a commitment.
Price the operating burden before the contract value
On Monday morning, Kojo put two columns on a whiteboard.
The first held the work needed to launch: infrastructure changes, data-flow review, vendor checks, testing and buyer evidence. The second held the work that would remain: deployment support, monitoring, backups, access reviews and future engineering changes across two hosting paths.
The second column changed the decision.
A large contract can cover an initial build and still leave a small team carrying an expensive operating burden. The relevant comparison is not contract value against setup cost. It is expected margin against the full cost of supporting the promise for the life of the agreement.
Kojo also had to consider concentration risk. If the buyer represented most of the revenue attached to the new environment, losing that account later could leave him maintaining infrastructure with no commercial reason to keep it. The contract needed either enough committed value to cover that exposure or terms that limited how far he built before approval.
This resembles the decision in the local hire three prospects asked for. Repeated demand can signal a real market requirement. One promising conversation can also tempt a founder into building a permanent cost base for revenue that has not arrived.
Move the question to the first serious call
By the time Kojo returned to the buyer, he had a conditional plan rather than an eager yes. He could describe what his team needed to confirm, which changes appeared necessary and which commercial commitments would make the work defensible.
The buyer still had a decision to make. So did he.
That Friday changed his qualification process. For prospects in a new country or regulated sector, he began asking where production data could be stored, which systems could process it and what approval evidence would be required. He asked before tailoring the demo and before letting a forecast treat the opportunity as nearly closed.
The next time a buyer raised data hosting, Kojo was not holding a reheated coffee at the end of procurement. The question appeared near the top of his first serious call, while both sides still had room to shape the deal.
Comments
No comments yet.