Test price with a step size and a cadence

1. Why test price

Raising price means charging more for the same thing. Price is said to move gross profit more than any other of the eight ways to make a customer worth more, which is why it comes first among them (SOP 72, section 1, has all eight). In one practice here, pricing tests are run more than anything else in the business.

The arithmetic given: at a 10 percent profit, a 20 percent price rise with sales held level triples the business rather than growing it by 20 percent. Put it on an invented base of 100: revenue of 100 with costs of 90 leave a profit of 10. Raise prices 20 percent and the same sales bring in 120 against the same costs of 90: a profit of 30, three times what it was. The tripling assumes sales stay the same, as the example says, so costs stay at 90. All the extra goes straight to the bottom line.

Testing is called important, and, from experience, the right price is usually higher than you think. The evidence offered:

• A research study suggested a tight link between how profitable a company is and how often it tested its prices. Figures credited to the same research company show that companies which tested pricing more frequently made more profit and, as a result, grew faster and faster. Both findings are shown in graphs. This page gives no figure for how much more profit frequent testers make. SOP 77 — Run the ten pricing plays that add profit now, section 1, carries the study's figures on pricing against retention and acquisition.
• In one company that was bought, everything was reviewed and the only change was to double the price. The business tripled, with nothing else done. In this view, making more money does not have to be complicated.

Testing has its own cost: it is expensive for a few months. The only thing more expensive is never testing at all and running a business that stays under-monetized for life.

2. What to measure

Find the sweet spot by multiplying the sales conversion rate by lifetime gross profit. The price that gets the most people to buy at the highest gross profit is the price that makes the most profit. The aim is the most money, not the most units sold; the money is what keeps you in business and lets you help more people for longer. For how to work out lifetime gross profit, see SOP 55 — Compute lifetime gross profit and the ratio that gates spending.

Before adjusting a price, look at:

• how many people buy at the price you charge now: conversion rate;
• how many repeat purchases each buyer makes, or how long they remain, at that price: churn.

For goods that are not luxury goods, a higher price means people buy less often and come back to buy less often. With these metrics you can start to see the best price for your product. Unless you have a different short-term strategy, the best price is the one that makes you the most money, not the one that wins you the most customers.

3. The comparison table

In one practice here, findings go in a small table. The worked example, with 100 clicks at each price:

Measure	At $10	At $20	At $100
Clicks	100	100	100
Conversion rate	5%	4%	2%
Sales	5	4	2
Churn	10%	10%	33%
Lifetime value	$100	$200	$300
Total return	$500	$800	$600
Difference	—	60%	20%

How the rows fit, checked: sales are clicks times conversion rate. Lifetime value is the price divided by churn: $10 ÷ 10% = $100, $20 ÷ 10% = $200, and $100 ÷ 33% is about $303, printed as $300. Total return is sales times lifetime value. The difference is each price's total return against the $10 price: $800 is 60 percent more than $500, and $600 is 20 percent more. The table takes no delivery cost off, so its lifetime value is price over churn, while the rule below names lifetime gross profit.

The rule: choose the price that converts the most people at the highest total lifetime value, specifically lifetime gross profit. In this table that is the total return column, where the highest figure is at $20: $800.

The same table, with a further reading of it (where the next test would go, and testing on new customers first), is on SOP 83 — Write the price-raise letter with a vanishing discount, section 3.

4. Run a price test

• Set a new price. Tell your sales team, or change it on your site.
• Work out how much the gross profit per unit changes between the old price and the new one.
• Work out what conversion rate would let the new, higher price only break even with the old one.
• Track your conversion rate at the new price.
• If it comes in above your break-even rate, you have a winner.
• Then test again, until conversion rate times lifetime gross profit drops.

The figures in this example are invented. Take a one-off sale at $100 that costs $40 to deliver, so lifetime gross profit is the $60 made on that one sale. At a 20 percent conversion rate, each lead is worth $12 of gross profit (20 percent of $60). Raise the price to $120 and each sale makes $80. The new price breaks even at whatever conversion rate still earns $12 a lead: $12 ÷ $80 = 15 percent. Track the rate at $120; above 15 percent, the new price wins. A move from $100 to $120 is a 20 percent step, the step size in section 5.

5. Start low, then step up

Start with a low price. You need sales to make sure people actually want what you sell, and early sales also run water through the pipes, showing you where the leaks are. Once things are humming, nudge the price up.

The recommendation is to nudge the price by 20 percent every 10 sales or so, and to keep going until sales drop dramatically; at that point, return to the sweet spot (section 2).

Flag: the step cadence. Here the step is 20 percent every 10 sales or so. SOP 146, section 5, steps the price about twice as often. Stepping less often gives each price more sales to judge it by; stepping more often finds the ceiling sooner. This page does not settle which reading is right.

If someone bought at a price you later find was a little too high, add some extra bonuses for them. If they are unhappy about it, you can always refund them the difference. Both remedies are said to have been used.

6. How often

• In one practice here, prices are tested every quarter.
• While starting low, the nudges come every 10 sales or so (section 5), counted in sales rather than on a calendar.
• The test (section 4) runs again after each winner, until conversion rate times lifetime gross profit drops. This page gives no figure for the time between repeat tests.
• In the closing checklist of the eight ways, the line for price is to set a date for your first price test.

How far to keep raising, and the signs a price has gone too far, are on SOP 82 — Set price from the pricing rules and the three models, section 4.

7. Where the pricing plays say to test

• A yearly increase in contracts. It was adopted partly on the strength of pricing data: companies that go on optimizing and testing price earn much more than companies that do not (SOP 79 — Write an annual inflation increase into a contract, section 1).
• Billing once a year. Whether a yearly price, at twelve times the monthly rate, cuts conversions five times over is not known; it has to be tested in your own business (SOP 77, section 5).
• Price endings. Test them; sometimes shorter numbers do better (SOP 77, section 6).

8. What this page does not decide for you

• How each step is measured. The 20 percent step is a recommendation for starting out. Whether each step is taken on the starting price or on the latest price is not established on this page.
• Where to stop. The test (section 4) stops when conversion rate times lifetime gross profit drops; the start-low nudge stops at a dramatic drop in sales and goes back to the sweet spot; further signals are on SOP 82, section 4. None is ranked above the others.
• What counts as a dramatic drop. This page gives no figure for a dramatic drop in sales.
• Your conversion at a new price. Not established on this page.
• Which short-term strategies justify a lower price. Not established on this page.

9. The checklist

Question	The answer
Why test	Price moves gross profit more than any of the eight ways; from experience, the right price is usually higher than you think
What to maximize	Conversion rate times lifetime gross profit: the most money, not the most customers, unless you have a different short-term strategy
What to record	Price, clicks, conversion rate, the sales made, churn, lifetime value, total return, and the difference against the lowest price
Which price wins	The highest total return
The test	Set a new price; work out the change in gross profit per unit and the break-even conversion rate; track; above break-even is a winner; test again until conversion rate times lifetime gross profit drops
Starting out	Start low; once things are humming, 20 percent every 10 sales or so, until sales drop dramatically; then back to the sweet spot
Sold too high	Extra bonuses; refund the difference if they are unhappy
How often	Every quarter, in one practice here; nudges every 10 sales or so while starting low
First step	Set a date for your first price test
How far to push	SOP 82, section 4

10. What this page does not cover

Choosing a pricing model, and the rules for how far to push a price, are SOP 82. The ten pricing plays, and the study's figures on price against retention and acquisition, are SOP 77. Working out lifetime gross profit is SOP 55. Cutting what delivery costs, instead of raising the price, is SOP 127 — Cut delivery cost nine ways. Telling existing customers about a raise, and revisiting price at least once a year, are SOP 83.

Consumer-protection rules on changing prices and on refunds are not covered on this page.