How Much Time Should a Solo Founder Spend on Marketing vs Building?

There's no universal ratio, but here's an honest starting point. Before launch, spend most of your week building and a fifth to a third of it talking to the people you're building for. From launch onward, move toward something close to half building and half marketing, where marketing means selling, talking to users and working one distribution channel. Then let what happens each week decide the next shift, not your mood. This guide is for solo and two-person founders in their first 18 months, and it gives you a 20-minute time audit, starting splits by stage, a sample founder schedule and a 15-minute weekly review.

The short answer

Most solo founder time management advice skips the part that matters: the right split changes with your stage, and the wrong one usually isn't a bad ratio. It's no plan at all. You build all week because building feels like progress, and marketing becomes something you'll "do properly after the next release". Or you panic after a quiet launch, post everywhere for two weeks and ship nothing.

The best-known version of a fixed answer comes from Gabriel Weinberg and Justin Mares in their book Traction, in the chapter on traction thinking. They call it the 50% rule: "spend 50% of your time on product and 50% on traction." Their point is that the two should run in parallel, and they name the opposite habit The Product Trap, the belief that improving the product is always the best use of your time. I think 50/50 is a good default once people can actually use what you've built. Before that, it's usually too much marketing for too little product.

• Audit first. Find out where the last two weeks really went before you set any target.

• Pick a split for your stage and treat it as a starting point you'll adjust, not a rule.

• Protect the blocks. Long building blocks, a fixed distribution block, support in set windows.

• Review weekly in 15 minutes. Hours against plan, one outcome per bucket, one change for next week.

Why does the building vs marketing split go wrong?

It's rarely laziness. The two kinds of work give you very different feedback, and your brain notices.

• Building feels like progress. You close a ticket, push a commit, see the feature work. The reward is the same afternoon.

• Marketing pays back late and noisy. You send ten messages and get one reply next Tuesday. You publish a guide and it gets three visits. It's hard to tell if anything worked, so it feels like nothing did.

• Switching between them is expensive. An hour of outreach dropped into the middle of a coding morning costs more than an hour. You lose the warm-up on both sides.

• Support eats the week quietly. Once a few people use the product, small questions arrive all day. Each one takes five minutes, and together they take your best hours.

• "Marketing later" has no deadline. The next release always looks like the moment it'll be ready. It never quite is.

The fix isn't willpower. It's knowing your real numbers, deciding a split on purpose and building a week that makes the split easy to keep.

Where did your last two weeks actually go?

Don't trust memory here. Most founders think they spend more time on marketing than they do, because marketing is the work they worry about most. Give it 20 minutes and use three sources you already have.

1. Your calendar. Count calls, meetings and any blocks you actually kept for the last 14 days.

2. Your git log. Run git log --since="2 weeks ago" --author="you" and look at which days and roughly which hours you were committing. It's a rough proxy for building time, and good enough.

3. Your sent mail and DMs. Outreach, replies to users and support answers all leave a trail. Skim it and sort by type.

Put everything into five buckets: build (code, design, product decisions), sell and talk to users (calls, demos, outreach, follow-ups), distribution and content (your channel, writing, listings), support (answering questions, fixing customer-reported problems) and admin (accounts, invoices, tools, email that isn't any of the above). Rough is fine. You want the shape, not a timesheet.

Illustrative example (made up, used through the rest of this post): Tackroom is a small web app for horse-boarding barns. It keeps feed schedules, turnout notes and messages to boarders in one place instead of a whiteboard and a group chat. Its founder works on it full time, launched seven weeks ago and has a handful of barns using it. The product, the founder and every number below are fiction.

The Tackroom founder would have guessed "about a third on marketing". The real number is 12.5%, and support takes more time than selling and distribution put together. That's a very normal result seven weeks after launch, and it's the kind of thing you only see when you count.

How much time should you spend on marketing at each stage?

Here's the table I'd start from. It comes from experience and from watching other small founders, not from a study, so treat every number as a first guess. "Marketing" here means the sell and talk to users bucket plus the distribution bucket. Percentages are of your working hours, whatever those are.

A few notes on reading it. Pre-launch marketing is mostly talking to the people you hope will pay, not posting. Those conversations shape what you build, which is why they count. After launch, marketing climbs because the product finally exists and nobody knows about it. And support isn't failure. It's a sign people are using the thing. It only becomes a problem when it's spread across the whole day.

The signals in the last column matter more than the percentages. If you only take one thing from this table, take this: change the split because of something you saw, not because of a bad week's feelings.

How do you build a founder schedule that protects both?

A split on paper means nothing if your calendar doesn't match it. The goal is a week where the right work happens at the right time without a daily negotiation with yourself.

Give building long blocks and stack the rest

Paul Graham's essay Maker's Schedule, Manager's Schedule is still the clearest explanation of why this matters. He writes that people who make things "generally prefer to use time in units of half a day at least", and that "a single meeting can blow a whole afternoon". His fix at Y Combinator was office hours clustered at the end of his working day, so they never interrupted it. When he was building his own startup, he describes programming at night and doing "business stuff" during the day, in effect two workdays in one.

As a solo founder you're both people. So give the maker a long block, ideally your sharpest three or four hours, and push calls, outreach and support to the edges of the day.

Theme days or a daily split?

Two patterns work. Theme days give whole days to one kind of work, for example Monday to Wednesday building, Thursday selling, Friday content and review. A daily split gives every day a morning building block and an afternoon for everything else. Theme days suit deep technical work and founders who struggle to switch. A daily split suits the first months after launch, when users need replies within a day and marketing needs small, frequent effort. Try one for two weeks before you judge it.

Protect one distribution block

Put a fixed block on the calendar, every day or every other day, for your one channel. Sixty minutes is plenty. The point is that it happens when it's scheduled, not when building feels finished. Building never feels finished.

Batch support into windows

Answer support at set times, for example right after the morning block and again late afternoon, and say so on the support page if you like. Urgent problems, like a customer who can't log in, still go first. Everything else waits a few hours, which most people are fine with.

Use low-energy hours for low-stakes work

Everyone has a slump hour. Don't waste your best hours on admin and don't waste your slump on architecture. Low-energy time is good for updating a listing, tidying a doc, sending follow-ups you've already drafted or clearing invoices.

Here's what that looks like for the made-up Tackroom founder, in the first 90 days after launch, on a daily split:

That's roughly 16 hours of building, about 12 of marketing and the rest support, admin and the review, across a week of about 37 hours. It sits just under the first-90-days row, which is fine while support is still heavy. As support settles, the freed hours go to marketing. Nothing heroic.

What counts as marketing time, and what only feels like it?

This is where most founders fool themselves. Real marketing time puts your product in front of a specific person or teaches you something about them.

• Talking to users and prospects. Calls, demos, replying to a trial user with a question. If you want a sense of what this looks like when it works, Aura++'s guide on getting your first 10 paying customers without a launch day is mostly this bucket.

• Outreach. Personal messages to people who match your customer, sent by you.

• One channel, worked properly. If you haven't picked one, SideHunt's piece on choosing one weekly distribution habit that fits covers how. Don't pick three.

• Writing for that channel. A guide, a comparison, an answer to a question your buyers ask.

• Keeping listings current. Launch pages and directory profiles with today's screenshots and description.

Fake marketing looks busy and moves nothing:

• Tweaking the logo, the colours or the landing page font for the third time.

• Reading growth threads and saving them for later.

• Switching email, analytics or scheduling tools.

• Planning a content calendar you won't follow.

• Checking the signup dashboard ten times a day.

Log fake marketing under admin when you do your audit. It's honest, and it's a little embarrassing, which helps.

How do you review the week in 15 minutes?

Once a week, Friday afternoon or Monday morning, compare hours with plan and look at one outcome per bucket. Hours tell you whether you kept the plan. Outcomes tell you whether the plan was any good. You need both, and outcomes win when they disagree.

1. Total the hours by bucket from your calendar and git log. Five minutes.

2. Write one outcome per bucket. Build: what shipped. Sell: how many real conversations. Distribution: signups or replies you can trace to the channel. Support: tickets, and whether any repeated.

3. Change one thing for next week. One. Move an hour, move a block, or drop something.

Notice what happened in the made-up week 2. Building time spent on the cause of the repeated tickets gave the founder hours back in support, and those hours went straight into conversations. That's the review doing its job. If you'd like a product-numbers ritual to sit next to this one, the DEV post on a 20-minute weekly metrics review for under 100 users pairs well with it. Keep this review about time against outcomes.

What if you only have 8 to 12 hours a week?

Same logic, smaller numbers, and less room for waste. With 10 hours a week after launch, something like 5 building, 4 marketing and 1 for support and admin is a fair start. The difference is in the shape of the week.

• Keep building in one or two long blocks, usually on the weekend, because an hour after work rarely gets you into hard problems.

• Put marketing on weekday evenings. Replies, outreach and a short distribution block fit into 45 minutes far better than code does.

• Answer support once a day at a fixed time and say so.

• Drop the admin bucket to near zero by picking simple tools and not switching them.

The trap at this size is spending all ten hours building because it's the only time you get to build. Two weekday evenings of marketing will usually do more for you than two more hours of features.

When should you break your own split?

A split is for normal weeks. Some weeks aren't normal, and pretending they are is its own mistake.

• Launch week. Flip it. Most of your hours go to replying, talking to new users and fixing what they hit. Building is limited to fixes. Go back to the plan the week after.

• A serious customer bug. Drop everything until it's fixed and the affected people have heard from you. Then do the review and see what it cost.

• A fundraising sprint. Treat it as its own temporary bucket with an end date.

The rule for breaking the rule: write down when you'll go back to the normal split, and actually go back.

What traps should you watch for?

• Marketing only on launch day. One big push followed by six weeks of silence teaches you nothing about what works.

• Building features as procrastination. If a feature nobody asked for appears the week outreach was due, be honest about why.

• Handing marketing off entirely too early. You can't brief someone well on work you've never done yourself.

• Measuring hours instead of outcomes. Six hours in the channel with zero replies for a month is a signal, not a success.

• Guilt spirals. Missing the plan one week isn't a character flaw. Adjust and move on. The review exists so you don't need to feel bad.

• Changing the split every few days. Give each split two weeks before judging it.

A 14-day plan to balance building and marketing

FAQ

Is 50/50 building and marketing right?

It's a sensible default once people can use the product. The 50% rule in Traction is about running product and traction work in parallel instead of one after the other. Before anyone can use what you've built, I'd lean further toward building, with steady conversations alongside. After launch, start near half and half and let the weekly review move you.

Should a technical founder learn marketing or hire someone?

Learn enough to do it yourself for the first few months. You'll find out which channel fits and which messages land, and that's what you'd need to brief anyone later. Hiring help makes more sense once you know what you'd be asking them to repeat.

How much marketing should you do before launch?

Enough that you're talking to people who have the problem every week. Pre-launch marketing is mostly conversations and a few honest posts about the problem, not a campaign. If you reach launch day without ten people you could message personally, you did too little.

What if marketing isn't working after a month?

Check the outcomes before you change the split. Did you actually work one channel on schedule? Do the people who arrive sign up, or bounce? A month is short for most channels, but if you have zero replies and zero conversations, change one thing about who you're reaching or what you're saying, not the amount of time.

Where to start this week

Block 20 minutes today, open your calendar, your git log and your sent mail, and count the last two weeks into five buckets. Then pick the row in the stage table that fits you, put a distribution block and two support windows on next week's calendar, and review it on Friday. That's the whole system. It's small on purpose, because a schedule you keep beats a perfect split you abandon by Wednesday.

And when you're ready to put what you've built in front of more founders and early adopters, Aura++ is a launch and discovery platform built for that.