There is a difference between a growth experiment that fails and one that teaches nothing.
A failed experiment may still show the business that its target market is wrong, its message is weak, its data is unreliable or its chosen channel cannot reach the right people economically. An experiment that teaches nothing leaves the company with only an expense.
Over recent weeks, we have been looking more closely at the economics of early customer-acquisition tests inside SaaS and other digital businesses.
When a company tries a new growth channel, the cost is usually easy to identify. There may be list-building fees, software subscriptions, advertising spend, agency charges or the cost of an employee’s time.
The expected return is also familiar: qualified meetings, new customers and additional recurring revenue. But there is another return that receives less attention.
Information.
Suppose a SaaS business runs an outbound campaign and does not generate the expected number of meetings.
The immediate conclusion may be that outbound does not work. But what has actually been established?
Were the contact details accurate? Did the campaign reach the intended decision-makers? Were conversations taking place but failing to create interest? Was there interest but no clear reason to book a meeting? Were meetings being booked with companies that were never suitable customers?
Each possibility describes a different failure. Each requires a different response.
Poor data should lead to changes in sourcing and verification. Low contact rates may point to timing, channel or execution. Conversations without interest may reveal a problem with the market, message or proposition. Interest without action may mean the next step requires too much commitment. Unqualified meetings may indicate that the targeting criteria are too broad.
When all of those stages are compressed into a single result, management cannot tell which decision needs to change.
The company knows that money was spent. It does not know what the money discovered.
This is why we have come to think that a growth pilot should be designed as a controlled financial experiment rather than simply a smaller version of a full campaign.
Before the pilot begins, the company should be clear about whom it is trying to reach, what response it is testing, what outcome would justify further investment and how much capital or time it is prepared to risk.
During the pilot, the movement through each stage should remain visible.
How many records were usable? How many decision-makers were reached? How many meaningful conversations occurred? How many prospects matched the intended customer profile? How many agreed to the next step?
The purpose is not to create reporting for its own sake. It is to protect the next decision. This becomes particularly important when comparing different acquisition channels.
A direct conversation can produce useful feedback quickly. Even when the answer is no, the reason may become visible. The prospect may already have the problem solved, may not consider it urgent, may not understand the proposition or may simply fall outside the intended market.
Silence through a less direct channel is more difficult to interpret.
It may mean the prospect was not interested. It may also mean the message was never seen, the contact information was poor, the account was inactive, the communication was filtered or the request did not feel important enough to answer.
The result may be the same—no meeting—but the information produced is very different. This does not mean one channel is automatically superior to another. It means that the value of a channel during an early experiment depends partly on how clearly it allows the company to observe what is happening.
Financial reporting often becomes interested in customer-acquisition cost, pipeline value and payback periods. Those measures matter, particularly once an acquisition process becomes repeatable.
During a small pilot, however, they can create false confidence. One early customer may make the economics look exceptional. No early customers may make the entire channel look worthless. Neither conclusion is necessarily supported by a limited amount of activity.
At that stage, the more useful questions are whether the underlying process is becoming clearer and whether the company now knows what should be tested next.
A properly structured growth experiment can therefore produce three valuable outcomes.
It may generate revenue. It may produce evidence that justifies scaling. Or it may show the company what must change before more money is committed.
The weakest outcome is not simply a campaign that produces no customers. It is a campaign that produces activity without allowing anyone to explain the result.
That distinction affects how growth budgets should be governed.
A business should not continue spending merely because activity is taking place. Nor should it abandon an entire channel because one poorly observed attempt produced nothing.
The decision should depend on what the evidence says about the audience, the message, the execution and the economics of reaching the next customer.
At N.P. Ummer Financial, we have been spending time considering growth expenditure in this way: not only as money committed in pursuit of revenue, but as capital used to improve the quality of the next decision.
A controlled experiment does not need to succeed commercially to justify its existence. But it should leave the business knowing more than it knew before.
Revenue may be the hoped-for return. Information should be the minimum.