A number that goes up and down
You check the subscriber count most mornings. It went up after the gift season and the promotion with a big creator, and down in the spring. That is roughly all it tells you. You do not know whether the people who joined in the promotion are still there, whether gift recipients who converted stay longer than people who signed up on their own, or whether the box people received first affects how long they stay.
Your subscription app has a churn figure, but it mixes everyone together, and it counts ended gifts and paused subscribers in ways you are not sure about.
One total hides very different groups
Subscribers who arrive from different places behave differently, and a total count adds them up into a number that does not describe any of them.
- Promotional signups with a big first-box discount often leave sooner.
- Gift recipients who convert may stay longer, or not, and you cannot tell.
- A disappointing first box may lose more new subscribers than long-standing ones.
- Prepaid and monthly subscribers are counted together.
- Pauses and skips look like churn in some reports and not in others.
Without splitting these apart, marketing decisions are made on the total.
What not seeing retention costs
The main cost is spending on the wrong things. A channel that brings many signups who leave after one box looks good on the count and poor on retention. A creator whose audience stays for a year looks modest on signups. First-box discounts may be buying subscribers who never pay full price. Without retention by group, you keep paying for what looks good today.
The owner ends up holding opinions about which channels work, and nobody can check them.
The retention reports we build
- Your subscription app's history is loaded, including every subscription's start, skips, pauses, plan changes and end, with the reason where you have one.
- Each subscriber is tagged with signup month, acquisition channel from their first order or code, first box received and plan type.
- Ended gifts, pauses and skips are treated by clear rules you agree, so they do not distort the churn figures.
- Retention is shown as a grid: for each signup month or channel, the share still subscribed after one, three, six and twelve boxes.
- A second view compares groups side by side, such as promotion signups against full-price ones, or subscribers whose first box was a particular month.
- The reports refresh daily from the subscription app, so each new month joins the grid without anyone building it.
| Grouping | Question it answers |
|---|---|
| Signup month | Did people who joined in a given month stay? |
| Acquisition channel | Which channels bring subscribers who stay? |
| First box received | Did a particular first box lose new subscribers? |
| Plan type | How do prepaid and monthly subscribers compare? |
| Discount at signup | Do discounted signups stay after the discount ends? |
The reports show what happened in your data. What to do about it is your decision, and where groups are too small to read much into, the report says so.
Deciding with retention in front of you
Before the next promotion, you open the grid and compare the last one's signups with full-price signups at the same stage. You can see how each group has done by box three and box six. When a creator proposes a paid campaign, you can check how their previous audience stayed. When a first box seems to have lost more newcomers than usual, you can see it within weeks rather than guessing a year later.
Monthly meetings stop arguing about whether a channel works, because the grid is open on the screen.
Is this the report you are missing?
- You track a total subscriber count and not much else.
- You cannot say how long promotional signups stay.
- Gift conversions are not reported separately.
- Your churn figure mixes pauses, gifts and cancellations.
- Marketing decisions are made on signups rather than who stays.