A pilot that buys six months of runway is worth taking only when the work tests a capability the company already intends to build. If the automation requires a new product, a new buyer and a new operating model, the extra cash may extend the company’s life while quietly replacing its direction.
Consider Ama, an illustrative Cape Town founder with four employees and a SaaS product built for logistics teams. At 4:17 on Friday afternoon, she was holding a cooling mug of rooibos when an email arrived from a US company proposing a paid pilot.
The contract value covered roughly six months of her current burn. The company wanted an answer by Monday.
There was one condition. Ama’s team would need to automate a document-review process adjacent to their product, then connect it to systems they had never worked with. The buyer called it a pilot. Her technical lead estimated that the first version could consume most of the next eight weeks.
Without the contract, Ama might need to let one person go before the end of the quarter. With it, the roadmap promised to existing customers would stop moving.
She had one weekend to decide which risk she preferred.
Six months can hide the cost of saying yes
Runway changes how a founder sees an opportunity. When cash is short, a contract can begin to look strategic simply because it is large enough to matter.
I have made versions of this decision while building products across African, European and US markets. Overseas contracts carry another layer of attraction for an African startup. The currency may travel further at home. The customer’s name may help with future conversations. A successful pilot can appear to validate expansion before the company has earned it.
Those benefits are real. So is the distortion they create.
Ama opened a spreadsheet and added the obvious costs: engineering time, cloud usage and support. The total still made the pilot look attractive. Then she added what the team would stop doing.
Two customer interviews would be postponed. A release requested by three current accounts would move back. Her technical lead would spend eight weeks learning the buyer’s internal systems instead of improving the core product. If the pilot ended, much of that knowledge would leave no reusable asset behind.
The contract bought six months on the bank statement. The diversion could cost two months of product learning and weaken the relationships already keeping the company alive.
Runway should be measured in decisions preserved, not only salaries paid.
Separate reusable learning from paid distraction
The useful question was simple: if the US buyer disappeared after the pilot, what would Ama still own?
She wrote down four possible assets. A document classifier could become part of the existing product. The integration work had little value outside this customer. The compliance requirements might help with future US deals, although no other prospect had requested them. The relationship itself could lead somewhere, but that remained a hope rather than an asset.
Only one part clearly strengthened the company she was already building.
This is where founders often confuse technical possibility with product direction. A capable team can build the requested automation. That fact says nothing about whether it should.
Ama needed a boundary that the buyer could understand. She divided the proposal into two parts: a narrow pilot using the document classifier inside her existing workflow, and customer-specific integration work that would require separate funding, milestones and ownership terms.
The change reduced the immediate contract value. It also revealed what the buyer actually valued.
If the buyer wanted access to Ama’s product capability, the narrower pilot could survive. If it wanted a small Cape Town team to absorb custom development risk, the deal would probably disappear.
That possibility sat on the table through Sunday evening. Losing the pilot could still mean losing a colleague. Accepting the original scope could turn the company into an outsourced automation team before its own product had found stable demand.
The same tension appears when capital comes with a direction attached. I explored that problem in What If Follow-On Funding Requires Building a Company You Never Chose?. Money always buys time. Sometimes the buyer also expects control over what that time produces.
Make the pilot prove a product decision
On Monday morning, Ama sent the revised scope.
Her proposal gave the pilot one job: determine whether the classifier could reduce manual document review inside the workflow her product already served. It defined the data the buyer needed to provide, the result both sides would evaluate and the work excluded from the first phase.
She also placed a decision point before any custom integration began. If the core workflow proved useful, they could price the next phase with better information. If it failed, her team would stop without carrying months of customer-specific code.
The buyer pushed back. Its operations lead wanted the full automation included. Ama held the boundary and explained what could be tested first.
By late afternoon, the company agreed to the narrower pilot.
That outcome is illustrative, and a real negotiation could end differently. The important turn was Ama’s decision before the reply arrived. She had converted a frightening yes-or-no choice into a test of strategic fit.
A good pilot should answer a question your company already needs answered. It might test whether buyers will pay, whether an automation works on real data or whether a workflow survives outside your home market. Each week of delivery should create evidence you can use after the contract ends.
This resembles the discipline required after a successful demonstration. A working product can still leave the company exposed if the next milestone proves nothing important, as I wrote in The Six Weeks Left After the Demo Works, and What Monday Must Prove.
On Tuesday, Ama moved the delayed release back onto the team calendar. The pilot had become smaller, and the runway extension was no longer six months. Her four-person team was still building the same company.
Comments
No comments yet.