The Upgrade Prompt
An in-product ask for money, placed at the moment the ceiling is felt.
The problem it solves
Free users rarely go looking for the pricing page. Something has to raise the subject, and most attempts raise it at a moment that has nothing to do with what the person is trying to do.
The difference between a good upgrade prompt and an advert is timing. If the prompt arrives while someone is reaching for a capability they do not have, it is information. If it arrives on a Tuesday because a scheduler fired, it is noise, and it trains people to dismiss the next one without reading.
Three placements, ranked
- At the blocked action
- The person clicked something the plan does not cover. Highest intent, lowest annoyance. The prompt should name the specific action, not the plan tier.
- At the approaching ceiling
- Ninety percent of a limit, or the fourth teammate on a three-seat plan. Good intent, and it gives them time to decide without pressure.
- Ambient
- A persistent badge or a line in the account area. Low intent, near-zero cost, and the only one that works for people who never hit a limit.
Write the prompt about the outcome that is currently out of reach. Plan names mean nothing to someone who has not read the pricing page. Say what will happen, then say what it costs.
Prompt copy that gets dismissed
- Upgrade to Pro to unlock more features
- You have hit a limit on the Free plan
- Go Premium today
Prompt copy that gets read
- Version history goes back 30 days on Standard. Your plan keeps 7. See plans
- You are inviting a fourth editor. Seats 4 and up are 8 dollars each per month. Add seat
- Exports over 5,000 rows need Standard, at 12 dollars a month. Upgrade and export
Frequency is the part teams get wrong. Cap prompts per user per week, suppress them for anyone who dismissed the same one twice, and never interrupt a task in flight. An upgrade prompt that appears mid-edit costs you the edit and the upgrade.
Worked example
A hypothetical suppression rule
A team caps contextual prompts at two per user per week and drops any prompt a user has dismissed three times. They expect total prompt impressions to fall sharply. Whether upgrades fall with them is the actual question, and it is worth running as a hold-out for a full month, since annoyance shows up in churn: The rate at which customers stop paying or stop using the product. later than it shows up in clicks.
When it fits
- The product has a real ceiling that active users reach through ordinary work.
- The upgrade resolves the exact thing the person was trying to do a second ago.
- You can suppress repeats per user rather than per session.
- The prompt can lead to a checkout that takes under a minute, so intent survives the trip.
When it backfires
- It interrupts a task, which costs you the task and the sale at once.
- It fires on a schedule rather than on behaviour, so it reads as spam and gets pattern-blindness within a week.
- The copy names the plan instead of the outcome, leaving the person to do the translation.
- Dismissal is hard to find or the modal reappears next session, which is sludge: Friction added deliberately to stop someone doing a thing they are entitled to do. and gets remembered as such.
Products using it
Notion
Surfaces plan limits at the point of the blocked action, such as inviting past the guest limit, rather than in a standalone campaign.
Zoom
Shows the meeting countdown on the free tier during the call, so the ceiling is visible while it is being approached.
Canva
Marks premium elements in the editor itself, so the cost is attached to the specific asset a person has just chosen.
The psychology under it
The takeaway
Attach the ask to the blocked action, cap how often it can appear, and write it about the outcome.
Finished the teardown? Bank it and the day counts toward your run.