An offer that pays for customer relationships while assigning little value to the technology is telling you where the buyer believes the durable advantage sits. In an African AI startup, that often means the defensibility is the trust, workflow knowledge, and route into a customer account, while the model layer can be rebuilt or replaced.
In April 2012, Facebook agreed to acquire Instagram for about $1 billion. Instagram had a small team and around 30 million registered users. Facebook’s announcement focused on helping Instagram’s community grow and bringing its experience to more people. The transaction still needed approval, and the outcome was not complete when it was announced. Facebook closed the acquisition later that year.
The public announcement did not separate a price for Instagram’s code from a price for its audience. It did show what Facebook wanted to preserve: a product people already used, a habit already formed, and a relationship with a growing community.
That distinction matters when an acquisition conversation lands in Accra, Lagos, Cape Town, Berlin, or London.
The code may be good, but the account knowledge may be harder to replace
An AI startup can build a capable extraction tool, support assistant, credit workflow, or internal automation in a relatively short period. A buyer with engineers, cloud credits, and access to the same model providers may believe it can reproduce much of that technical capability.
It cannot reproduce your place inside a customer’s daily work from a repository.
That place is usually built through less visible work: learning why a finance team rejects certain records, knowing which manager can approve a pilot, adapting to the documents a customer actually receives, and staying useful after the demo has stopped being interesting.
A startup that serves businesses across African markets may also have earned practical knowledge that is easy to overlook in a product roadmap. Customers may operate across currencies, countries, informal handoffs, legacy tools, and teams split between local operations and overseas leadership. The defensible asset can be the accumulated understanding of how work really moves through that organisation.
The buyer is often pricing that understanding, even if the offer never uses those words.
Read the offer as a map of your bargaining position
When a buyer assigns a high value to contracts, renewals, customer introductions, or the founders staying through a transition, pay attention. It may be evidence that your commercial position is stronger than your technical position.
That is neither flattering nor insulting. It is useful.
A low valuation for the technology can mean the buyer sees it as replaceable. It can also mean your technology has not yet been documented as a repeatable asset. If the system depends on one engineer’s judgment, private prompt experiments, or customer-specific workarounds, the buyer will see execution risk rather than a transferable product.
This is close to the problem in Samira’s workflow rebuild. A workflow that lives mainly in one person’s head has less value than one a team can operate, inspect, and improve.
Before negotiating, separate what you have built into three files:
- A customer map: active accounts, renewal risk, champions, decision-makers, integration points, and evidence of daily use.
- A product map: what is configurable, what is reusable, what depends on a particular customer, and what can be transferred without the founding team.
- A dependency map: model providers, data access, key employees, security commitments, and the customer permissions required to keep the product working.
This makes it harder for a buyer to treat every asset as interchangeable.
Build defensibility where customers would feel the loss
The right response is not to pretend the model is a moat. Models change quickly. Providers change terms. A capability that looked rare six months ago can become a checkbox in a larger platform.
Build the parts a customer would struggle to replace.
That can mean a decision record that explains why an AI system took an action. It can mean approval paths shaped around a customer’s real controls. It can mean a reliable onboarding process, data mappings that survive staff changes, or a product that has become part of a weekly operating rhythm.
The stronger question for a founder is: if a buyer rebuilt our application, what would customers lose in the first month?
If the answer is “our interface,” you have work to do. If the answer includes trusted people, embedded workflows, decision history, and a clear record of what the system has learned in use, you have something worth defending.
Instagram’s acquisition is a reminder that buyers can pay heavily for an existing relationship with users. For an African AI startup, the equivalent may be far smaller and far more specific: a procurement lead who trusts your team, an operations manager who knows the workflow works, and a system that has earned its place before the buyer arrives.
Comments
No comments yet.