Revenue

Growth used to get reported in users. How many signups this month, how many monthly actives — that was the first slide. While the number climbed it was easy to believe the company was working.

The trouble is that the two numbers often don't climb together. More free users means more server cost and more support tickets without more money coming in. A product with 100,000 signups and 40 paying teams looks like a fast-growing company on the user chart and like something else entirely in the bank account.

Pagedeck ran into the gap in a meeting. Grace mentioned 100,000 users to an investor, who asked how many of those were paying teams. The answer was 48. The distance between the expectation the first number set and the second number made the rest of the conversation hard. That is why revenue takes a box of its own at the end of AARRR: all four earlier boxes can be turning while this one sits at zero.

Three numbers get counted here.

1. The share that converted to paid. If 48 of 1,000 teams that started for free paid, that's 4.8 percent.
2. What one team pays. Forty-eight teams paying about $4,300 a month between them works out to roughly $90 each.
3. The share that moves up a plan later. Of the teams that paid once, how many moved to a higher tier within three months.

Keeping the three separate matters because identical revenue gets built in very different shapes. Forty-eight teams at $90 and 480 teams at $9 add up the same, but they don't wobble the same amount when one customer leaves, and they don't need the same number of people answering their questions.

Three things set this box apart from the earlier ones. First, the unit of counting changes. From acquisition through referral you count people; here you count whatever the contract attaches to, usually a team. A team going from one user to nine is a ninefold change in headcount and no change at all in paying teams, and mixing the two units breaks every conversion rate you calculate.

Second, this box inherits the numbers in front of it. The surest way to double the absolute figure here usually sits outside the box. Lift the share of teams that create a first document in the activation stage from 45 percent to 60 percent and the count of paying teams follows without anyone touching the checkout screen.

Third, the lag is the longest of the five. With a 14-day trial, whether a team that arrived today pays is settled two weeks from now, and if it lands inside a company's budget cycle it slips to the next quarter. Group teams by when they arrived and follow them for at least eight weeks, or the rate reads low.

There are ways to lift this box alone. Shrink what the free plan covers, or raise the price, and next month's number goes up. Those moves usually cut into an earlier box, though. Dropping free sharing from five people to two raises conversion and lowers return visits in the retention stage at the same time.

This is the last of the five boxes, but stopping here means stopping at one payment. Three things have to hold for the box to turn into company growth.

1. The first payment has to reach a second month. If 48 teams pay this month and 9 of them cancel next month, the real gain is 39. Counting conversions without counting cancellations produces a table that only looks like it's growing.
2. What a team pays has to grow over time. A team that started with 3 seats and runs 12 six months later pays four times as much. Widening an existing team costs far less than winning a new one, so a company where this happens grows revenue faster on the same flow of new teams.
3. What a team pays has to cover what it cost to win. Spend about $270 in ads to land a team that pays $90 a month and you break even in the third month. Stretch that past a year and the company has to spend more to earn more, which puts a cash ceiling on how fast it can grow.

When all three hold, the box feeds back into the first one: the money left over buys more ads or more people, which lifts acquisition again. The five boxes are a loop rather than a line.

"Conversion is 4.8 percent and teams pay about $90, so we won't hit the quarterly target."

Multiplying the two numbers lands short of the revenue the plan needs.

"Don't shrink the free plan just to move the revenue box. Activation will fall over first."

A warning that the lever available inside this box cuts an earlier one.

"Nine of the 48 teams that paid this month look like they'll leave. Re-count it as a net number."

A request not to report a figure that hasn't had cancellations subtracted.

Pagedeck's last quarter ran from 10,000 visits to 48 paying teams. When Mia split the box in two, more than half the teams that started a trial reached the checkout screen, and half of those stalled at the card entry field. Teams that needed a tax invoice were turning back at a screen that only accepted cards. Adding bank transfer and invoicing took the same flow from 48 paying teams to 61. Sofia now gets a weekly list of teams that turned back at checkout and calls them herself.

The most common reading is that a low conversion rate is this box's problem to fix. The number lives here, so the reaction is natural. In practice the cause often sits upstream. A team that hasn't yet seen what the product is for won't pay however good the pricing page is. Left uncorrected, the team reworks the checkout screen and the price list over and over while the teams that leave in week one without creating anything keep leaving.

Assuming a price increase lifts the box is the second misreading. Hold team count constant and raise the rate, and on paper revenue rises. But price moves together with conversion. Take $90 to $135 and watch conversion fall from 4.8 percent to 2.9 percent, and the total barely changes — while the team count drops, which shrinks both next quarter's expansion candidates and the pool of people who could send an invite.