What review cadence keeps side projects honest?

From the SoleOS answers series — written about our own product space; grounded in published definitions and documented behavior, never invented numbers.

A review cadence that keeps side projects honest has three tiers: a 10-minute weekly glance at trend direction, a 45-minute monthly review per product against last month, and a quarterly gate where you decide to keep, fix, or sunset each app. Anything more frequent than weekly mostly measures noise, not signal, and anything less frequent than quarterly lets dying products coast unnoticed for too long.

The mistake most solo founders make isn't skipping reviews — it's reviewing too often at the wrong depth. Checking Stripe every morning feels like diligence. It's actually a habit that trains you to react to random day-to-day variance instead of real trend changes.

Why daily checking backfires

Daily revenue and install numbers are dominated by noise: weekday/weekend cycles, one refund, a payout timing quirk, an App Store report that lagged by a day. If you're running more than one product, daily checking multiplies fast — five apps checked daily is 35 checks a week, most of them telling you nothing you didn't already know.

Worse, daily checking creates a specific behavioral trap: you start making decisions off single-day blips. A bad Tuesday triggers a panicked pricing change. A good Thursday triggers overconfidence in a growth strategy that was actually a fluke. The metrics dictionary is useful here precisely because it forces you to think in defined formulas and windows rather than raw daily numbers — MRR, not "what Stripe showed me this morning."

If you want a rule of thumb: don't look at a metric more often than the interval over which it can meaningfully change. Daily active users can shift daily. MRR for a subscription product usually can't — most of it is locked in by existing subscribers who renew on their own schedule.

Tier 1: weekly (10 minutes, all products)

Once a week, glance at trend direction across your whole portfolio, not depth on any one product. The question is simple: is anything moving in a direction that needs attention before the month ends?

What to look at:

• MRR or revenue trend line, not the raw number
• Any product that dropped off a cliff (crash, outage, App Store rejection)
• New signups vs. the prior week, roughly

This is a triage pass. You're not analyzing — you're scanning for fires. If everything looks like it did last week, you're done in ten minutes. This is also where a single-screen view across products earns its keep, because opening five separate dashboards for a ten-minute scan turns into thirty minutes fast.

Tier 2: monthly (45 minutes, per product)

Once a month, go one level deeper on each product. Compare this month to last month and to the same month last year if you have the history. This is where you actually decide things — pricing tweaks, whether to keep investing marketing time, whether churn crept up.

What to check monthly:

• MRR growth rate (and whether it's growth or just avoided churn)
• Churn rate, ideally split from involuntary payment failures
• Where signups are actually coming from — attribution across products gets messy fast when you're running several apps with overlapping audiences
• Whether Stripe and RevenueCat or App Store numbers still reconcile — see tracking Stripe and RevenueCat together if they've drifted apart

Monthly is also the right cadence for looking at a forecast, if you use one. Any projection needs enough history to mean something — with SoleOS, for example, the underlying model wants at least 21 days of data before it will even generate one, and it reports its own confidence (R²) rather than pretending certainty it doesn't have. A forecast checked daily just shows you the same noisy curve wiggling; checked monthly, it shows you whether the trend itself is bending.

Tier 3: quarterly (a real decision, per product)

Quarterly is the gate. This is where you ask the harder question: does this product still deserve the hours I'm giving it? Weekly and monthly reviews catch problems; quarterly reviews catch products that are quietly dying at a pace too slow to notice week to week.

A useful quarterly checklist:

• Compare growth rate over the quarter, not the month — a single bad month is noise, three bad months is a trend
• Check revenue concentration across your portfolio — if one app is carrying the rest, know that explicitly rather than assuming
• Decide: invest more time, hold steady, or start a sunset plan
• If sunsetting, do it as a process, not a shutdown — a checklist beats an abrupt kill

This is also the cadence at which the portfolio view matters more than any single product's dashboard. Looking at ten apps individually every quarter takes a full day; looking at them side by side takes an hour and shows you which ones are actually worth that day. If you're weighing whether a dedicated tool is worth it for this versus a spreadsheet you update by hand, SoleOS vs a spreadsheet walks through where each breaks down as the number of products grows.

What this cadence doesn't replace

None of this replaces paying attention to genuine anomalies outside the schedule — a payment processor outage, a sudden spike in refunds, an app getting rejected from the store. Cadence is for routine review, not incident response. If something looks broken, look at it immediately regardless of what day it is.

It also doesn't mean you need a dashboard tool to do this well. If you're running one or two products, a recurring calendar reminder and a spreadsheet is a completely reasonable version of this cadence — see the free spreadsheet template for a starting point. The cadence matters more than the tooling. Where a portfolio tool like SoleOS starts to earn its cost is specifically at the weekly and quarterly tiers once you're past four or five products, because that's where manually opening Stripe, RevenueCat, App Store Connect, and GA4 separately for each app turns a 10-minute scan into an hour.

Disclosure: SoleOS is portfolio intelligence for multi-product founders, and this post was written by the team building it — so take the recommendation with that in mind. If your portfolio is small enough that a spreadsheet and a calendar reminder cover this cadence fine, you don't need us yet.

Frequently asked questions

Isn't checking daily just more responsible?

Not really — it's more responsive to noise, not more responsible. Responsibility comes from acting on real trends, and most single-day movements aren't trends. If daily checking helps you sleep better, that's a legitimate reason to do it, but separate the "checking for peace of mind" habit from the "making decisions" habit, and only let the latter run on a weekly-or-slower cadence.

What if a product is genuinely volatile week to week?

Some categories — ad-supported apps, seasonal tools, anything tied to a platform algorithm — really do move fast. For those, tighten the weekly tier into a twice-weekly glance, but keep the monthly deep-review and quarterly decision gate at the same cadence. Volatility changes how often you scan, not how often you decide.

How many products can I realistically review well?

It's less about a hard number and more about time budget. Ten minutes weekly and 45 minutes monthly per product scales roughly linearly — five products is manageable by hand; fifteen starts to demand either serious discipline or a tool that aggregates the scan step. The live demo shows what that aggregated weekly view looks like with sample data, no sign-up needed.

Should projections change my cadence?

Only at the monthly tier. Projections need enough data history to say anything — under three weeks of data, treat any forecast as a placeholder, not a plan. Checking a forecast daily just shows the same underlying uncertainty redrawn; checking monthly shows whether the trend and its confidence are actually improving.

What's the first sign a product needs a cadence change, not just attention?

If you find yourself checking a product's numbers outside its scheduled review because you're anxious about it, that's a signal — either something's actually wrong and deserves an off-cycle look, or the product has become a source of low-grade stress disproportionate to its revenue, which is itself a quarterly-review question worth asking honestly.