How do I decide which side project to focus on?

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

Decide by evidence, not enthusiasm: focus on the side project with the strongest growth trend and the best effort-to-impact ratio, not the one you're most excited about this month or the one you've already sunk the most hours into. A project that's flat at a bigger number is a worse bet than a smaller one that's accelerating — the trend matters more than the snapshot. Score your candidates honestly on paper, then commit to the winner for a real block of time before you re-open the question.

Disclosure: SoleOS, which publishes this blog, is a portfolio dashboard for solo founders running more than one product — so read this as an informed view from inside the space, not a neutral one.

Excitement and sunk cost aren't signals

The two forces that most often decide which project a founder works on are the two worst ones: whichever idea is newest and shiniest this week, or whichever one has already absorbed the most time and money and therefore feels too far along to abandon. Both are emotional accounting, not business signal — excitement tells you what you'd enjoy building next, and sunk cost tells you what you'd hate to admit was wasted. Neither tells you whether real users are showing up, sticking around, or paying.

Set both aside and look at what the projects are actually doing, without narrating a story over the top of it.

Look at the trend, not today's number

A project sitting at a respectable number that hasn't moved in two months is a worse bet than a smaller project whose signups, activation, or revenue are climbing week over week. The absolute figure tells you where a project has been; the trend tells you where it's going, and where it's going is what an extra week of your time will multiply.

Concretely: for each candidate, look at the direction of its core metric — new users, trial starts, revenue, whatever fits the product — over the trailing four to eight weeks, not the lifetime total. Is it up, flat, or down over that window? A project can have the smallest total in your portfolio and still be the correct one to focus on this quarter, because it's the one actually moving.

Effort-to-impact: where does an extra week compound

Once you know the direction, ask a sharper question: if you had one focused week to give any of these projects, which one would that week move the most? That depends on whether growth is being held back by something fixable — a confusing onboarding step, a missing integration, an unclear pricing page — or by something structural, like a market that's genuinely small or a channel that's already saturated no matter how much polish you add.

A week spent unblocking a single fixable lever compounds; a week spent polishing a project that's structurally capped mostly just feels like progress. The honest test is whether you can name the specific lever an extra week would pull, and whether pulling it plausibly changes the trend line. If you can't name it, that's information too.

Market size and your own energy for it

Traction is necessary but not sufficient. Two more things belong in the decision. First, an honest read on how big the opportunity actually is — not a market-size figure you can't verify, but a look at how many people are visibly searching for the problem you solve, how many competitors exist, and roughly how large the ones that are working have gotten. A project with great early momentum in a genuinely tiny niche has a ceiling; know where it is before you commit a quarter to hitting it.

Second, your own energy for the work. Focus is sustained attention, and attention is hard to sustain on something you dread opening, even when its numbers look fine. This isn't permission to chase excitement over evidence — it's the opposite check: among projects with comparable traction, the one you have real appetite for is the one you'll actually show up for every week. Between two projects with similar numbers, energy is a legitimate tiebreaker. It just doesn't get to override the numbers.

The real cost of spreading thin

Every live project carries a fixed tax regardless of its size: support tickets, bug triage, dependency updates, marketing upkeep. Split your attention across four projects and each gets roughly a quarter of you — including the one with the most room to grow. That's the real cost of "staying balanced" across too many things: not that any single project suffers dramatically, but that the one with the best shot never gets the concentrated push it needs to break out.

That's a separate problem from the focus decision itself — see how to manage more than one side project without any of them stalling for the operational side of keeping the others alive while you focus on one.

A simple way to score candidates honestly

Put every candidate project in a table, one row each, and score four columns on a plain 1–5 scale:

• Growth trend — direction of the core metric over the trailing 4–8 weeks
• Effort-to-unlock — how much of the current ceiling is one fixable lever vs. something structural
• Market signal — how large and active the space around the problem genuinely looks, based on what you can observe
• Your energy — how much you actually want to work on it right now

Add the columns up. Don't dress this up with weights or precision you don't have — the value is in writing it down and scoring honestly, not the arithmetic. It forces you to separate "I like this one" from "this one is scoring well," which is usually where the real answer surfaces. For two or three projects, a spreadsheet with those four columns, revisited every few weeks, is genuinely enough — you don't need software for this.

Decide on evidence, then commit

Once a project wins the comparison, commit to it for a real stretch of time — a defined number of weeks decided in advance, not "until I feel like checking again." Re-scoring every few days isn't focus, it's still shopping. Pick the window, write down what you expect to see by the end of it, and don't reopen the decision until you get there. If a project you didn't pick keeps nagging at you, revisit the scoring at the next real checkpoint — not mid-week.

Where SoleOS fits — and where it doesn't

This is the portfolio-triage problem SoleOS is built around: once you're running enough products that keeping the trend, revenue, and traffic numbers for each one current becomes its own chore, a unified view turns the comparison above into minutes instead of an afternoon of logging into separate dashboards. The point isn't to pick a winner for you — it's to keep the inputs honest and current so your own scoring stays grounded in what's actually happening. See how the layout works via the live demo if you want the mechanics first.

Honestly, if you're comparing two or three projects, stick with the spreadsheet above — a dedicated dashboard is solving a problem you don't have yet. It starts earning its keep once the number of live products makes "just check each dashboard" the actual bottleneck to making this decision at all.

Frequently asked questions

How long should I give a project before deciding to focus elsewhere?

Long enough to see a real trend, not a blip — a trailing four-to-eight-week window is usually the minimum for signups, activation, or revenue to tell you anything meaningful. Deciding off a single good or bad week is how sunk cost and excitement sneak back in disguised as data.

What if the project I'm most excited about has the weakest numbers?

Keep working on it in whatever spare time you'd spend on a hobby, but don't call it your focus. Excitement is a legitimate reason to keep a side project alive at low intensity; it's not a reason to give it the concentrated push that traction should earn.

Should I shut down the projects I'm not focusing on?

Not automatically — focusing on one doesn't mean the others need to die. Some are worth leaving on low-maintenance autopilot. If a project has been flat or declining for a while with no plausible unlock, that's a separate and harder call — see how to know when it's actually time to kill a side project for that decision specifically.

My projects have completely different revenue models — how do I compare them fairly?

Compare direction, not units. A subscription product's MRR growth rate and an ad-supported app's traffic growth rate aren't the same number, but "percent change over the trailing weeks" is comparable across both. Normalize to trend before you try to normalize to dollars, and if you're trying to grow the winner afterward, the mechanics of taking a small product from its first traction to real recurring revenue covers what to measure next.

None of my projects are growing right now — what does that mean?

It means the decision this month is less about which one to focus on and more about which one has the clearest, cheapest experiment to run to find out if growth is possible at all. Pick the one where you can name a specific, fixable lever to test, give it a defined window, and use that result — not enthusiasm — to decide what happens next.