When a venture runs a twenty percent discount, the obvious cost is twenty percent of the revenue on every transaction the discount applied to. Most founders calculate this cost, weigh it against the volume the discount produced, and conclude that the discount paid for itself if the volume was high enough. This calculation, while arithmetically correct, misses the larger cost. The larger cost is structural rather than transactional, and it does not show up on the income statement of the month the discount ran. It shows up over the following twelve months in revenue that did not arrive, customers who waited for a discount that came later than they wanted, and a price point that has been quietly relocated below where the venture originally operated.
I want to argue in this piece that the predictable discount is one of the most expensive marketing instruments a venture can use, and that the expense is invisible at the time of running and obvious only in retrospect. The mechanism by which it does its damage is straightforward. It teaches your customers to wait, and waiting customers behave very differently from buying customers, in ways that compound across quarters.
The first loss and the second loss
The first loss from a discount is the one founders see. You sold a product or service for less than its full price, and the difference between the full price and the discounted price is the immediate revenue you did not receive. This loss is bounded, calculable, and acceptable if the volume justifies it.
The second loss is the one founders rarely notice and almost never measure. When a discount is predictable, customers learn its pattern. They learn that you discount in November for Black Friday, in December for the holidays, in January for the new year clearance, in March for the back-to-school spend, in June for mid-year, and so on. The learning is implicit; no customer sits down and writes out your discount calendar. The learning is also durable; once a customer has bought from you at a discount once, the discounted price becomes the price they remember as the venture’s actual price, and the full price becomes the price they perceive as a markup they are now wise to.
The behavioural consequence of this learning is that customers who would otherwise have bought at full price defer their purchase to the next discount. They do not write you and tell you they are deferring; they simply do not buy when they would have, and they do buy when the discount arrives. From your perspective, the discount produced a sale. From a more honest accounting, the discount cancelled a full-price sale that would have happened anyway and replaced it with a discounted sale that produced less revenue. The “increased volume” the discount appeared to generate was, in part, the same customers who would have bought at full price, choosing the discount path instead.
The second loss is not bounded. It compounds. Every discount cycle reinforces the customer’s expectation that another discount is coming. Over time, a venture that runs predictable discounts trains its entire customer base to operate on a deferred-purchase model. The full-price revenue stream shrinks because the customers have learned to wait. The discount-period revenue stream grows because that is when the customers buy. The total revenue across the year may look stable, but the average revenue per transaction has dropped, the unit economics have eroded, and the venture is now structurally dependent on running the discount cycles to produce any revenue at all.
The competitor effect
There is a third loss that operates at the level of the customer’s competitive comparison set, and it is the most insidious of the three.
When a customer has learned that your full price is “really” the discounted price plus a markup, your competitors’ full prices look comparatively reasonable. A competitor who does not discount, who has held a stable price for years, looks expensive next to your discount price and reasonable next to your full price. The customer who is shopping in the moment between your discount cycles sees your full price, sees the competitor’s price, and makes a comparison that goes against you. They might buy from the competitor at the moment they are ready to purchase, even if your discount price would have been lower than the competitor’s price, because the discount is not currently available and they do not want to wait.
This is the deepest cut. The discount cycle has not just deferred your own customers; it has lost some of them entirely to competitors who happened to be reasonably priced at a moment when you were between discounts. Founders rarely notice this because they cannot easily measure it; the lost customer never appeared in your CRM as a lost opportunity, because the customer never engaged with your venture at full price. They simply went elsewhere.
What this looks like in African markets specifically
The dynamic I am describing is universal, but it has a particular shape in African markets that is worth naming.
In economies where consumer purchasing power is constrained and seasonal, customers are more sensitive to discount cycles than they are in stable markets. They learn discount patterns faster, defer purchases more aggressively when a discount is anticipated, and spend more time in shopping mode looking for the discount window. This means African ventures that run predictable discounts train their customers more efficiently than equivalent ventures in stable markets, and the second loss accumulates faster.
It also means that the founder, watching the financial picture, may attribute the soft full-price sales to the macroeconomic environment rather than to the discount cycle they have been running. The macro is real, but the discount cycle is amplifying its effect. A venture that has held its prices stable in the same market is less affected by macro tightening because customers do not have a discount to wait for, so they buy when they need the product. A venture that has trained its customers to wait will see disproportionate softness in the months between discounts, regardless of the macro.
When discounting is actually correct
I want to be specific about when discounting is the right move, because the argument so far might suggest it never is. There are situations where the discount makes economic sense, and they are worth naming so the rule has the right exceptions.
The first is time-sensitive inventory. If you have stock that will go out of date, lose value, or become obsolete, discounting is correct. The alternative is zero revenue from the stock, and any positive price beats zero. This is the classic clearance scenario and it is a legitimate use of discount as inventory management.
The second is structurally non-recurring purchases. If you sell something a customer buys once and rarely returns to (a wedding photographer, a one-off renovation), the customer is not in your discount-training population, and discounting at the point of decision can convert a customer who would otherwise have walked away. There is no second loss because there is no second purchase to defer.
The third is deliberate first-purchase incentives to acquire a customer into a recurring relationship. If you can convert a discounted first purchase into a full-price ongoing relationship, the discount is acquisition cost rather than discount in the structural sense. The discipline is to make sure the conversion to full price actually happens; if the customer remains on discount terms forever, you have permanently relocated your price point downward for that segment.
The fourth is strategic competitive response in a defined window. If a competitor has launched a price war and you need to defend market position for a defined period, a strategic price reduction can be correct, but it should be framed as a war response rather than as a recurring discount. The framing matters because the war response can end when the war ends, while a recurring discount cannot.
In every other case, discounting is almost always the wrong move. The fix for soft sales is to address the underlying issue producing the softness, which is usually one of the structural causes I have written about elsewhere. The discount is the easy lever, and it is easy because it transfers the cost from this quarter to the next several quarters in ways the founder can ignore for a while.
The diagnostic
If your venture has been running predictable discount cycles, the most useful diagnostic this week is to look at your full-price revenue trend over the last four quarters. If it has been declining or flat while your discount-period revenue has been growing, you are seeing the second loss. The fix is not to discount harder. The fix is to break the cycle, accept a soft quarter or two as customers re-learn the new pricing reality, and rebuild the venture on stable pricing. This is uncomfortable and it works, and the founders who do it consistently arrive at year three with healthier unit economics than the founders who chose the easier discount path and are still running the cycles a decade later.
The discount looks like a marketing instrument. It is actually a customer-training instrument. The customers it trains are not customers who buy more; they are customers who buy later, for less, and only when the cycle says they should. That is not the customer base you want. It is also not the customer base you have to build.
For why “your prices are too high” is rarely actually a price problem, see When Customers Say Your Prices Are Too High. For the structural risk of drifting into a position no segment is willing to pay for, see The Stuck Middle.
