The idea and why it works
Most small products ask for money on a calendar: a trial that ends on day seven, a paywall on day one, a discount email on day 30. The calendar knows nothing about the user. A need-shift upsell asks at a different moment: when the user's situation changes and they now need something the free product does not do.
Flo, the women's health app, is the clearest current example. In a RevenueCat Sub Club interview published on 11 November 2026, Mike Charniatsou, who works on growth and web onboarding at Flo, said that "40% of all subscriptions take place after six months" after install. Flo's users, he said, "may easily fall in and out of different needs". The app adds value tied to trying to get pregnant, following a pregnancy or understanding perimenopause, and that new value is what converts freemium users three, six and 12 months later.
The reason it works is simple. A person who has used your free product for months already trusts it. When their life or business changes, they have a new problem and a product they already know. If you meet that problem with a specific answer at that moment, the offer reads as help, because it is help.
Flo's results also show that the moment alone is not enough. When Flo added a perimenopause step to onboarding, the first test came back with "almost 10% drop in conversion rate". A second run changed nothing. The team then found that almost 75% of users were not being assigned any perimenopause phase, so the step was confusing most people who saw it. Flo rewrote the question with its medical team and let users who could not be placed move on. That third version "came back with 23% win". To check the lift was real, the team ran a removal test that took the new section out. Removing it lost, which confirmed the result.
Who it's for
This play suits products whose users stay for months and whose needs grow or change: invoicing and bookkeeping tools for freelancers, AI writing or research assistants, scheduling tools, course and community platforms, habit and health apps. It suits you if your free tier already solves a real job and you have enough active free users (a few hundred is enough to start) for change events to happen every week.
It suits you less if you are young and paying for every install. Charniatsou himself calls waiting on monetisation "a luxury, not universal advice": an app that must pay back acquisition from day zero cannot wait six months. If that is you, keep your existing first-week offer and add need-shift offers on top rather than replacing it.
What you need (tools, time, cost)
Event tracking. Any product analytics tool works. As of 13 Nov 2026, PostHog's first 1M events are free every month, with 1,500 free survey responses, which covers most small products.
An in-app message slot. A banner, modal or card you can show to one user based on an event. Most analytics or onboarding tools can do this, or you can build it in an afternoon.
Billing that supports upgrades mid-cycle. As of 13 Nov 2026, Stripe Billing's pay-as-you-go plan costs 0.7% of billing volume.
Time. Roughly a day to map moments and write events, a day to build the first re-onboarding flow and message, then four to six weeks to collect a result.
Step-by-step
Map the need-change moments. List what happens in your users' lives or businesses that creates a new problem: first client abroad, first subcontractor, first team member, crossing 20 invoices a month, launching a second product. Read support tickets and cancellation reasons for the words people use. Keep the two or three moments that happen often and that a paid feature genuinely solves.
Write each moment as a detectable event. Some moments show up in behaviour (a user adds a second currency). Others only the user can tell you, so ask with a one-question check-in ("Has anything changed in how you work?"). Flo's lesson applies here: if most users cannot answer the question, rewrite it, and let anyone who does not fit carry on without friction.
Rebuild a short onboarding for the new stage. Treat the moment as a second first day. Show three to five screens that set up the product for the new situation, using the user's own data. The free part of this setup must be useful by itself.
Make the offer inside that flow. Put the paid option at the step where the free version runs out, and name the specific thing it does for their new situation. One offer, one price, one button.
Run the end-user test before launch. Ask: if this user says no, did they still get something useful today? If the answer is no, the flow is a disguised paywall. Fix it before you ship.
Measure with a holdout. Show the new flow to most users who hit the event and hold out a random 10 to 20%. After four to six weeks, compare conversion and 60-day retention.
Confirm with a removal test. Once a version wins, switch it off for a slice of users, as Flo did. If conversion drops, the lift is real.
Template / copy / example
Here is a designed model for a small AI invoicing tool for freelancers. The core job, creating and sending invoices, stays free. The need-shift moment is the first time a user adds a subcontractor, because they now need to split income, track costs and pay someone else.
Event schema
{
"event": "need_shift_detected",
"properties": {
"shift_type": "first_subcontractor",
"source": "behaviour",
"trigger_event": "contact_added_as_payee",
"account_age_days": 143,
"free_jobs_completed": 38,
"plan": "free",
"eligible_for_offer": true,
"holdout": false
}
}
// follow-ups: stage_onboarding_started, stage_onboarding_completed,
// stage_offer_viewed, stage_offer_accepted, stage_offer_dismissedIn-app message
You just added your first subcontractor. We have set up a free split view for this month so you can see what you earned and what you owe them. If you want this every month, plus automatic payee invoices and a cost report for your accountant, Studio is £12 a month. You can keep using everything you use now for free either way.
[Show my split view] [See Studio]
The user gets the split view for the current month whether they pay or not. That is the value at the moment of the offer.
Illustrative maths (ILLUSTRATIVE, not a result). Say you have 2,000 active free users and 3% hit the first-subcontractor event each month: 60 people. If 10% of them upgrade at £12, you add six subscribers, or £72 in monthly recurring revenue, every month from one moment. Add two more moments at similar rates and you add about 18 subscribers a month from users you already have, at no extra acquisition cost. Your own rates will differ; the holdout tells you the real number.
Metrics to watch
Event rate: the share of active free users who hit each need-shift moment per month. Too low means the moment is rare or your detection misses it.
Assignment rate: the share of users who can actually answer a stage question. Flo's first version failed here, with almost 75% left unplaced.
Offer conversion versus holdout: upgrades within 14 days of the event, against the held-out group.
Share of new subscriptions from users older than 90 days: if the play works, this share rises over time.
Free retention after a dismissed offer: users who say no should stay at the same rate as the holdout.
Pitfalls
Asking a question users cannot answer. A confusing stage question cost Flo almost 10% before it was fixed. Test the wording and always give an exit.
Moving the core job behind the paywall. If free users find the thing they came for has become paid, the trust that makes this play work is gone.
Guessing the moment. Inferring a life or business change from weak signals feels intrusive when wrong. When in doubt, ask.
Calling one win final. A single test can be noise. A removal test is cheap insurance.
Waiting when you cannot afford to. If cash is tight, run this alongside your existing first-week offer.
Variations
Volume shift. Trigger the offer when usage crosses a level where the free plan starts to slow the user down, and pair it with a free one-off fix for today.
Use-case shift. When a user starts using your tool for a new kind of work (a writing assistant used for a first sales proposal), re-onboard them with templates for that work.
Team shift. The first invite to a colleague is a need-shift moment. Give the pair a free shared setup for the first project, then offer team features. Our 30-day upsell window covers what to offer in the month after a purchase, and this play covers the months after that.
Product speaks first. Combine the detection event with a proactive first message so the product opens the conversation only when the moment arrives.
Shift at cancellation. Some cancellations are need shifts in disguise; a reason-matched cancel flow can route "my needs changed" to a plan that fits the new stage.
