Usage Limit

Any company that lets people try the product first has to decide where free ends. For years the answer was time. A free trial opens every feature for 14 days from signup and locks it afterward, which means the wall arrives by calendar and has nothing to do with what the person did in those two weeks.

Cutting by time puts two opposite people in front of the same wall on the same day. Someone who signed up and never opened the product is locked out on day 15, and so is someone who came in daily and invited ten colleagues. When Pagedeck ran a 14-day trial, most of the incoming requests said the same thing: they hadn't made up their minds yet and wanted the trial extended. Asked what they had done with it, half of them turned out to be accounts that had never created a single document.

A usage limit moves that wall from time to volume. Open three documents and the person who writes one a month stays free for a whole quarter, while the team writing five a week hits the wall within days. The people with the strongest reason to pay run into it first, so the moment of being blocked lands near the moment paying makes sense. When you sort paywalls by what they count, this is one branch and splitting access feature by feature is the other.

Three decisions make the design: what to count, where to put the ceiling, and what to show the person who reaches it.

1. The unit — pick something that grows as the customer grows. Seats go up as a company hires, while document count fills up in a few weeks even for someone working alone. The same number behaves completely differently depending on which one you count, because it blocks a different group of people.
2. The number — don't pick it from a hunch in a meeting room. Take it from the distribution of people already using the product. Line up how many documents free accounts have created and look at where the median falls. If the middle account sits at two and accounts that have turned into teams are usually past five, the line goes at three, between them. Solo users never hit it; accounts that became teams do.
3. The treatment — there are three. Lock only new creation and leave existing work alone, switch everything to read-only, or charge for the overage. Pagedeck uses the first. Locking documents somebody already wrote reads as holding their data hostage rather than asking them to pay.

The hard part is always where the line sits relative to value. It has to come after the moment someone first feels the product is useful, usually called the aha moment. At Pagedeck that moment wasn't a person writing a document alone — it was a colleague leaving the first comment on it. Block earlier than that and all the user takes away is a few minutes with a text editor.

The pricing page is where it appears first. Three columns for free, team and enterprise, each carrying a number like three documents or unlimited documents — those numbers are the usage limits. They're the common way to build a soft paywall, which opens part of the product and blocks only what lies beyond, and they're nearly always present in a freemium model that keeps the basics free and charges for scale.

Inside the product it shows up in two places. One is the remaining-quota indicator. Something as small as "2 of 3 documents" keeps the wall visible so people see it coming instead of colliding with it. The other is the message that appears the instant the line is crossed, and what that message says decides whether anyone pays. A screen that only says the limit is reached gets closed. A screen that also says four teammates are looking at this document right now makes the cost of walking away concrete. Designing that blocked moment as a purchase moment has its own name: usage-based upgrade.

Companies selling an API — the channel programs use to exchange data with each other — price it as ten thousand calls a month for free and a fee per ten thousand above that. Here two different numbers end up on the same page: the usage limit that draws plan boundaries, and the rate limit that caps how many calls per second the servers will accept.

"Should we raise the free document limit from three to five? Way too many accounts are hitting it in their first week."

The line sits ahead of the aha moment, so it's turning people away before they feel any value.

"That belongs in gating, not a limit. You're splitting the feature itself, not counting volume."

Correcting which branch of the paywall the team is reaching for.

"Four percent of people who see the limit message click through to checkout. Let me try new copy."

A proposal to improve the conversion rate at the blocked moment.

At Pagedeck this number belongs to Mia and Ravi together. Mia treats the limit as product experience: how much has to be open for free before a team feels what it's like to work in the product. Ravi reads the same number as signups and revenue. If raising the limit to five brought more signups in a month but left paid conversion flat, the two of them argue about which direction the number should move.

There's one way to settle it. Pull a month of records and compare what accounts blocked at three did next against what accounts blocked at five did next.

The most common misreading is that a tighter limit produces more payment. Block sooner, pay sooner — the arithmetic feels obvious. What actually happens is that the rate goes up and the headcount goes down. Block at one document and the share of remaining users who pay may jump from 3 percent to 12 percent, but a tenth as many people stay, so less money comes in. Squeeze the limit while watching only the rate and the percentage improves as the revenue dries up.

The other frequent mistake is treating a number, once chosen, as a fixed setting. The limit was drawn against how people actually used the product, so when the product changes the meaning of the line changes with it. Add file attachments and one document weighs more than it used to, which makes the old three and the new three different amounts. Leave it alone and the free plan either gets too cramped for solo use, which costs signups, or drifts so generous that nobody ever crosses. Look at the distribution again every six months and move the line.