Pricing strategy covers two decisions, and a team takes them in order. The first is the pricing metric, which is the unit the customer is charged for, such as a seat, a transaction or a gigabyte. The second is the level, which is the number attached to that unit. Choosing a business model settles the shape of the money and leaves both decisions open, so those two decisions are where a product either keeps the value it creates or gives it away. Madhavan Ramanujam and Georg Tacke opened Monetizing Innovation in 2016 with a finding from Simon Kucher's pricing survey. Of the new products launched over the previous five years, 72 per cent had failed to meet their financial targets.
The sections below set out the metric and the two tests it has to pass. They then give three ways of setting the level and explain what packaging and tiers are for. The page ends with why the whole decision belongs in discovery.
The pricing metric
The metric decides more than the level does. It decides how the bill moves as the customer uses the product more, and that movement is the whole of what a customer notices between one invoice and the next. A metric worth keeping passes two tests. It rises with the value the customer receives, and the customer can work it out before the invoice arrives.
A per seat licence on a product that automates work is the standard case of a metric failing the first test. One analyst buys one seat, the product removes the work of forty people, and revenue from that account falls while the product succeeds. If the same product charges for the records it processes, the bill follows the value, so a customer getting more from the product pays more for it.
Predictability is the harder test. If a customer cannot work out next month's bill from this month's activity, the product is an open ended risk, and a finance department will refuse a risk it cannot size. A meter that counts something the customer can reproduce from their own records survives that conversation, and a meter that counts something they cannot see turns the invoice into a support ticket every month, for as long as the account lasts.
Three ways to set the price level
The metric settles the unit, and the number attached to it comes from one of three places.
| Approach | Where the number comes from | What it needs | The failure mode |
|---|---|---|---|
| Value based | The gain to the customer, measured in the customer's own numbers | Evidence of the gain and a study of what buyers will pay | A guess about the gain, written down and then treated as evidence |
| Cost plus | The cost of building and serving, with a margin added | An accurate cost of serving one more customer | Serving one more customer costs almost nothing, so the number drifts away from anything a buyer would pay |
| Competitive | What comparable products charge | A genuinely comparable product and a buyer who compares | The decision moves to a competitor who may have priced badly |
Value based pricing is the approach everybody claims and few practise, because it asks for a number out of the customer's accounts. A scheduling product cuts empty running for a haulage firm by four per cent. The price is then set against the fuel that firm no longer buys, and the figure sits in the firm's own ledger where either side can check it. Without that figure, the value based label describes an opinion.
Packaging and price tiers
One price serves one kind of buyer, and most segments hold more than one kind. Packaging groups what the product does into things a customer can buy, and tiers are the levels of that grouping. A tier exists to separate buyers who value the product differently, so the line between two tiers has to sit on something a buyer cannot easily misreport. Number of users, volume processed and the need for an audit trail all work, because an organisation cannot pretend to be smaller than it is for long. If the line is drawn on a feature nobody wants, nobody moves up a tier. If it is drawn on a feature everybody needs, the tier is a price rise with a new name on it.
Pricing as part of discovery
A team that prices in the week before launch is pricing after every decision that determines the price has already been taken. Ramanujam and Tacke put the conversation about what customers will pay at the start, before the product is built. They name four ways a product fails when that conversation happens late.
- Feature shock. The product carries so much that the price needed to cover it is one no customer will accept.
- Minivation. The product is priced below the value it delivers, so the revenue never arrives.
- Hidden gem. A product with real willingness to pay behind it never ships, because it sits outside the core business.
- Undead. Customers never wanted the product, and enthusiasm inside the company kept it alive anyway.
Only the minivation looks like a pricing mistake while it is happening. Feature shock, the hidden gem and the undead all look like product decisions at the time, and that is why the conversation that would have prevented them belongs beside the product work.
The oldest instrument for that conversation is the price sensitivity meter, introduced by the Dutch economist Peter van Westendorp in 1976. The meter asks one customer four questions about the same product. The questions run from the price at which the product would be too expensive to consider down to the price at which it would be so cheap that its quality came into doubt. The answers give a range, and a range taken from customers beats a number argued out in a meeting.
Both decisions produce a number, and a number reaches nobody on its own. A priced promise still has to arrive in front of the person willing to pay for it. The route it travels is the next choice the strategy has to make.