From Vinson · 25 September 2026 · for Austin and Gavin
Austin, this is new. We haven't discussed a trial win-back before. It came out of checking why so many profiles were suppressed. Nothing has been built or sent, and nothing in Klaviyo changes until you approve it. The proposal is in the "Win-back proposal" section, and the decisions it needs from you are at the end.
| Month | Trial users | Ever paid | Share who paid |
|---|---|---|---|
| Nov 2025 | 2,047 | 240 | 12% |
| Dec 2025 | 5,147 | 1,386 | 27% |
| Jan 2026 | 3,006 | 954 | 32% |
| Feb 2026 | 3,990 | 1,171 | 29% |
| Mar 2026 | 3,227 | 963 | 30% |
| Apr 2026 | 2,717 | 836 | 31% |
| May 2026 | 4,294 | 1,494 | 35% |
| Jun 2026 | 4,110 | 1,403 | 34% |
| Jul 2026 | 3,215 | 801 | 25% |
| Aug 2026 | 673 | 154 | 23% |
| Sep 2026 | 287 | 23 | too early |
Trials dropped in August because the trial left the default checkout on 5 August.
| Status | People |
|---|---|
| Paying now | 4,758 |
| On a trial now | 368 |
| Paid, then left | 4,701 |
| Never paid, trial over | 22,949 |
The tables on this page are counted in slightly different ways, so their totals differ by under 0.3%.
| Cancelled during the trial | Trial ran out, first charge failed | |
|---|---|---|
| People | 15,894 | 7,055 |
| Ran the optimizer at least once | 11,899 | 4,515 |
| Ran it twice or more | 5,443 | 1,519 |
| Had picked the yearly plan | 9,150 | 3,810 |
| Active in the app or portal in the last 30 days | 374 | 193 |
| Last trialed since June | 3,528 | 2,309 |
The 7,055 whose first charge failed matter most. They did not explicitly cancel: their card didn't go through, and the only email they got pointed them to Discord, not to a place to update the card. The flow rebuild fixes this for new trials.
| Trialed | Never paid | Of those, ran the optimizer | Paid, then left | Of those, ran the optimizer |
|---|---|---|---|---|
| May to Sep 2026 | 8,655 | 4,805 | 1,771 | 1,282 |
| Jan to Apr 2026 | 8,863 | 6,888 | 1,982 | 1,754 |
| Nov to Dec 2025 | 5,431 | 4,721 | 948 | 925 |
| Total | 22,949 | 16,414 | 4,701 | 3,961 |
Of the May to September group who never paid, 3,775 were active in the app or portal in the last 90 days. This is a re-engagement signal, not proof that the app is still installed.
| Group | Emails received |
|---|---|
| Everyone, on starting the trial | The trial welcome. Until 23 July also a day-2 email and a "trial ends soon" email; since then, the welcome only |
| Cancelled during the trial | "Quick question..." (tells them they cancelled a subscription, asks why, gives code GAMER, has no link), then a generic "hear about your experience" email on day 3, then a free month on day 13 |
| Trial ran out, charge failed | The failed-payment emails: "try a different card" with only Discord links |
| Campaigns | "Beta over" on 28 June; its second email went only to people who opened the first. The 8 September campaign sent to 6,601 profiles from two cancellation segments, but one segment also admitted current payers |
There has never been a dedicated win-back for trial users. Since 26 June and 6 August most of them have been suppressed and receive nothing.
Suppression is not an unsubscribe and does not itself mean these people opted out. Gavin still must confirm that the original marketing consent remains valid. The other costs are the Klaviyo plan and deliverability risk, since these addresses have gone months without email.
| People | |
|---|---|
| Trial users who are not customers | about 27,650 |
| Unsubscribed or bounced | about 1,000 |
| Trial users who haven't unsubscribed | about 26,500 |
| Of those, can be emailed today | about 5,500 (mostly people who paid, then left) |
| Of those, suppressed by the team | about 21,000 (mostly people who never paid) |
The split between "can be emailed" and "suppressed" comes from random samples of 50 and 100 profiles looked up in Klaviyo, so read it as about ±1,500.
The exact list comes from the segments or Kamil's export in the proposal below.
Can the suppressed 21,000 be emailed in a new flow? Operationally, a manual suppression can be reversed. That is different from overriding an unsubscribe, bounce or spam complaint, which this proposal never does. Gavin confirms that the original marketing consent is still valid before anyone is unsuppressed. A manually suppressed profile receives no marketing flow or campaign email, so eligible profiles must be unsuppressed before they enter.
Option 1 · Klaviyo segments. They count people and send nothing. Not built yet; they need your go-ahead.
| Segment | Rules |
|---|---|
| WB · Trialed, never paid, not opted out | A Stripe event of type "Stripe First Free Trial Activated" at least once, at any time · Successfully Paid zero times, at any time · no trial started in the last 8 days · never unsubscribed, bounced or marked as spam. Do not require "can receive email", because that would exclude the manually suppressed people being counted |
| WB · Trialed, never paid, manually suppressed | The same, plus current manual-suppression status. Gavin confirms the exact condition before relying on this count; otherwise use Kamil's export |
| WB · wave by trial date | The same, with the trial date limited to May to Sep 2026, Jan to Apr 2026, or Dec 2025 |
What the segments can't do:
Option 2 · An export from Kamil. Trial users who are not paying, with trial month, how the trial ended and whether they ran the optimizer. Slower, but it is the only way to build the waves below, which lead with people who ran it.
Recommendation: build the segments first for Klaviyo-visible counts, then use Kamil's export for the exact waves.
| Wave | Who | People |
|---|---|---|
| 0 | Already emailable, not customers (the campaign audience in the plan) | about 8,000 |
| 1 | Last trialed May to Sep, never paid, ran the optimizer | 4,805 |
| 2 | Last trialed Jan to Apr, never paid, ran the optimizer | 6,888 |
| 3 | Last trialed Nov to Dec 2025, never paid, ran the optimizer | 4,721 |
| Separate | Trialed, never paid, never ran the optimizer | 6,535 |
| Separate | Paid, then left (most are already emailable, so they sit in wave 0) | 4,701 |
Each wave runs for about two weeks. The next wave starts only if the last one stayed inside the guardrails below.
Sender "Austin from Hypertune", plain founder note, one button. {{ offer }} is whatever Austin picks.
Email 1 · Day 0
> Hi there,
>
> You tried Hypertune a while ago and didn't continue.
>
> The most common reason trial users give for leaving is that they didn't see enough difference. If that was you, this gives you another way to test Hypertune on your PC.
>
> If you'd like another go, {{ offer }}.
>
> Austin
Email 2 · Day 4, didn't click email 1
> Hi there,
>
> Hypertune reads your hardware, streamlines background processes and tunes Windows to your PC. It creates an automatic restore point, and one click reverts its changes.
>
> {{ offer }} is there if you want to try it again.
>
> Austin
Email 3 · Day 9, didn't click either
> Hi there,
>
> This is the last time I'll email about coming back. If you ever want to give Hypertune another go, {{ offer }}.
>
> Austin
Anyone who still is not paying or trialing and hasn't clicked by day 14 is suppressed again.
| Option | For | Against |
|---|---|---|
| 50% off with code GAMER | The code already exists; price was the second reason trial users gave for leaving (27%) | Discounts a product most of them left over performance. Kamil must confirm the coupon's duration and eligible plans before copy is final |
| Another 7-day free trial | Easy yes; they already know the app | Their first trial didn't convert; a second may go the same way |
| No offer, troubleshooting help instead | Answers the top reason for leaving (43%, "didn't give me performance") | Austin and Kamil must confirm the support route and capacity first |
No dedicated trial-user win-back has run before, so there is no baseline for it. The 8 September campaign had 10 clickers. Treat wave 1 as the test: if it produces fewer paying customers than the Klaviyo step up costs, stop there.
| Number | Source | Window |
|---|---|---|
| Trial users, months, statuses, how trials ended | Stripe subscriptions and customers | All time to 25 Sep |
| Optimizer runs, yearly choice, recent activity | The app and account sign-ins | All time to 25 Sep |
| Emailable, suppressed, unsubscribed shares | Random samples: 50 never-paid trial users, 100 paid-then-left, looked up in Klaviyo | 25 Sep |
| Suppression dates and rules | Klaviyo suppression events, bulk job and segment definition | 26 Jun and 6 Aug 2026 |
| Reasons for leaving | In-app cancel survey, 140 trial answers and 718 paid answers | 21 Aug to 25 Sep |
| Emails received | Klaviyo flow and campaign history | Read 25 Sep |
| Stripe events in Klaviyo from 9 December 2025 | Klaviyo event history | Sep 2025 to Sep 2026 |