Which product deserves your next hour?
A checklist for makers with three or more products: which one is earning more than it gets, which is leaking, which is quietly dying, and which one to let go.
By Torstein · · 4 min read
On this page
Once you run more than two products, the question every week is the same: which one do I work on? Most of us answer it by mood. We work on the one that's fun, the one with the newest bug report, or the one we launched most recently. The products that need attention most are often the ones we haven't opened in a month.
I run 14 products, most of them small. This is the checklist I use, with examples from my own portfolio. Real numbers are given as rounded ranges or ratios.
1. Is anything silent?#
Before any strategy: did a product, or a data source, stop? Revenue that drops to zero after a normal run, or traffic that disappears overnight, is either a real emergency or a broken connection. Both beat everything else on this list. The full checklist is in is my app dead, or is my tracking broken?.
2. Is something earning more than it's getting?#
Compare each product's share of your revenue with its share of your time and traffic. A product that brings in a large share of the money but hasn't had a release in months is the cheapest win in your portfolio. It already works; it just isn't getting your attention.
The reverse matters too. BibleWise, my Bible trivia app, gets more downloads from Brazil than from any other country, and close to none of its revenue. About a fifth of its downloads in one 90-day window came from ChatGPT referrals. That is attention without money, and it changes what the next hour should be: monetising the traffic it already has, not buying more.
3. Is a funnel leaking?#
Traffic up, conversions flat or down, means the problem is between the visit and the purchase: the store page, onboarding, the paywall or the trial. For GraceLock, trials started within minutes of install, from the onboarding paywall, and auto-renew was switched off within hours while the person was still using the app. More installs wouldn't fix that. A newer version moves the trial offer to after the first prayer, once the app has shown what it does.
A leak is worth fixing before you spend on ads. Every new visitor goes through the same hole.
4. Is something falling slowly?#
The dangerous declines are the gradual ones. A product losing a few percent a week never triggers a "something broke" alarm, and after two months it's a third smaller. Look at 8 weeks of revenue or trials, not last week vs the week before. If the trend is clearly down, decide: fix it, or accept it and stop expecting it to recover.
5. Is it time to let one go?#
Some products cost more than they make, month after month. Take a small puzzle game with a few downloads a week and no paying players (an illustrative example, not a real app). That doesn't automatically mean shut it down: if it costs little to keep running, it can stay on autopilot. But it shouldn't get your best hours, and once it starts costing real money, three loss-making months in a row is a fair point to make the call.
Be honest about what "costs" means here. Hosting and fees are visible. The hours you spend on a product that isn't going anywhere are the bigger cost, and no dashboard shows them.
Putting it together#
Each week, go down the list in order and stop at the first two or three things you find:
- Anything silent: fix it today.
- A leaking funnel on a product with traffic: fix it before buying more traffic.
- An earner that's being neglected: give it a release.
- A slow decline: decide, don't drift.
- A long-running loss: keep it on autopilot or let it go.
Three things a week is plenty. If a week has nothing, that is a good week, not a failed check.
How Plask automates the list#
Plask runs this checklist for you every Monday, on Pro, as fixed rules over your own numbers. Each rule needs enough data before it can fire, so a product with a handful of sales won't trigger anything:
| Rule | Fires when |
|---|---|
| Silence | Revenue or traffic stops, or a source stops syncing |
| Revenue vs traffic | Revenue falls 15%+ over 28 days while traffic holds |
| Leaking funnel | Sessions up 20%+ in a week, trials flat or down |
| Attention mismatch | A product earns 3x its traffic share, no release in 45+ days |
| Falling trend | A statistically clear decline over 8 weeks |
| Sunset candidate | Three loss months in a row, traffic not growing |
You get at most three cards a week, at most one per product, ranked by the money at stake where your data shows it. Each card shows the rule that fired and the numbers behind it. A short AI-written explanation sits on top, but the AI never decides what fires. You can mark a card Done, Not now (it comes back in 7 days only if the rule still fires) or Wrong, which makes that rule less eager for your products. Details are on the features page.