We Deleted the Trial Countdown Banner. Trial-to-Paid Followed.

A six-person B2B SaaS. Self-serve 14-day trial. No sales team, no demo calls. Trial-to-paid had been flat for two quarters, and the team was arguing about price.

Price was not the problem. The trigger was.

Every user saw the same thing on day one: a banner at the top of the app reading "Your trial ends in 13 days." It was the only upgrade prompt in the product until the day the account locked.

What the countdown banner was actually doing

It fired on our calendar, not on anything the user did.
It interrupted people mid-task to tell them something they couldn't act on yet.
It was dismissible, so by day three it was furniture. Users had been trained to close it without reading.
On day 14 the app locked. For most users, the first moment they seriously considered paying was also the first moment they lost access.

That last point is the killer. You're asking for a purchase decision at the exact moment you've made someone's day worse. No amount of pricing-page polish fixes a bad moment.

The one change: fire the upgrade prompt at the limit, not on a timer

We deleted the banner. In its place, one upgrade card that appears only when a user hits a hard limit on their current plan.

The trial plan caps projects at three.
A user clicks "New project" for the fourth time.
The card appears inline, in context, replacing the empty state they were about to fill.

No day-7 email. No modal on login. No countdown anywhere in the product.

This is a feedback-and-guidance fix, not a persuasion fix. The card's only job is to explain what just happened and show the single next step — Nielsen-style heuristic work applied to one moment instead of a whole design system.

If your upgrade prompt fires on a timer, you can't tell whether it's working, because the signal is buried in people who were never going to buy. Run a free audit on your upgrade flow and you'll get the trigger points flagged before you touch a single line of copy.

Before and after: the upgrade card

Before

Your trial ends in 3 days. Upgrade now to keep using [Product]. Choose from three plans. [ See Plans ]

Three problems: a deadline with no stated consequence, a button that ships people to a comparison page (a decision tax at the worst possible moment), and zero reassurance about their data.

After

You've hit your 3-project limit. Your three existing projects stay exactly as they are. Upgrade to add more — projects are unlimited on every paid plan. [ Add project 4 — $29/mo ]

What changed:

The headline names the exact thing they just tried to do.
The body answers the fear ("do I lose my work?") before it's asked.
The price is on the button. No surprise at checkout.
One plan, one action. No matrix, no "compare plans" detour.

If you genuinely have three plans with different jobs, keep the button as "Add project 4" and show the comparison on the next screen — but pre-select one plan and say why. We apply the same rule on our own plan page; see /pricing.

P0/P1/P2 for your upgrade flow

P0 — fix these first

Kill timer-based upgrade prompts. If it fires on day 7 regardless of behavior, it's noise you're paying for in goodwill.
Land the wall at the action. Never interrupt a task to sell.
State what stays and what breaks. Data ambiguity is the most common reason people stall at a paywall.

P1 — next sprint

Pre-select one plan with a one-line reason.
Show price before the click, not after.
Button copy that names the next action ("Add project 4"), not the transaction ("Upgrade").

P2 — worth doing

Don't lock accounts at expiry. Drop to read-only so people keep their data and come back.
Turn the expiry email into a summary of what they built, not a countdown.
Log every wall dismissal. A dismissible wall can't tell "not yet" from "no."

How to tell whether it worked

Split trial-to-paid into "hit a limit" and "never hit a limit." The second group was never your buyer — stop counting them as a loss.
Track the wall cohort separately: conversion rate and time from wall to payment.
Read the tickets. "Why can't I add another project?" means your copy isn't explaining the limit.
Watch for the same account dismissing the wall repeatedly. That's a packaging or pricing problem, not a copy problem.

The fix here was one trigger. Not a redesign, not a pricing change, not a new onboarding sequence. If your upgrade prompt is on a schedule instead of tied to behavior, you're guessing — and you can't tell a good guess from a bad one.