Design for early stage founders: what to spend on, what to skip, and how to buy it
Founders rarely ask how much design costs. They ask what it is for, and get an answer from someone selling it. Here is the version from the inside: what actually matters this early, and what you can safely leave ugly.
What design is for at this stage
Before product-market fit, design has exactly two jobs: help you decide what to build, and make the thing credible enough that a stranger will try it.
That is a much narrower brief than most engagements assume. It does not include a full design system, a brand book, an illustration library, or a pattern library for screens you have not validated. Those are real, later. Buying them now converts money into confidence rather than learning.
The hardest part of software is not building it. It is deciding what to build and getting someone to care. Design earns its budget against those two problems first.
The three decisions that compound
- 01
The model: what the nouns are
What objects exist, what they are called, how they relate. This leaks into your database, your URLs, your pricing, your sales calls and your support docs. Changing it in year two is a rewrite. Getting it right takes an afternoon with the right person.
- 02
The core flow: the one thing done well
One journey, from arrival to the moment of value, designed properly. Not five flows at seventy percent. The core flow is what people judge you on and what you demo.
- 03
The system: the smallest set of rules
Type scale, a handful of colors with jobs, spacing, and a dozen components. This is a few days of work that makes every future screen faster and keeps the product coherent as you add people.
Everything else can be fixed later for roughly what it costs today. These three get more expensive every month you postpone them.
What to do cheaply on purpose
- The marketing site. One good page with clear words beats a nine-page site you will rewrite when the positioning changes.
- The logo. Set good type, pick a color, move on. Come back when you know who you are.
- Internal and admin screens. They serve people who are paid to learn them. Make them clear, not beautiful.
- Illustration and custom imagery. Expensive, slow, and usually replaced within a year.
- Anything for a customer segment you have not sold to yet.
- Comprehensive documentation of a system nobody but you is using.
How to buy design at each stage
| Stage | Best option | What you are buying |
|---|---|---|
| Idea, no users | Do it yourself with AI tools, plus one paid review | Speed and learning. A few hours of expert critique before customers see it |
| First prototype, early conversations | A short project with one senior person | The model, the core flow and a minimal system. Weeks, not months |
| Early customers, a real roadmap | An ongoing part-time partnership | Continuity and judgment as the product grows, without a full-time hire |
| Funded, shipping weekly | A founding designer, hired carefully | Someone who owns quality full time and builds the practice |
Most early founders overshoot this table by one row. A large team at the prototype stage produces beautiful artifacts of the wrong product, quickly.
Red flags when hiring
- A portfolio of concepts rather than things in use. Shipping is a separate skill from presenting.
- Invented metrics. Any case study claiming a 340 percent lift with no methodology is a design of a case study.
- A process deck that arrives before questions about your business.
- Unwillingness to discuss what they would not build. People who never say no cost the most.
- A proposal with no first milestone inside three weeks. Early-stage engagements need visible value fast.
- Every project in the portfolio looking identical. That is a template, and templates do not know your customers.
A realistic budget picture
Numbers vary by market, so treat these as proportions rather than quotes. A focused early engagement covering the model, the core flow and a minimal system is typically a few weeks of one senior person's time. An ongoing partnership is usually a fraction of a full-time salary and buys continuity rather than volume. A founding designer is a full salary plus equity, and is worth it once design decisions arrive faster than a part-time partner can absorb.
- Ask for a scoped first phase with a real deliverable, not a retainer with no endpoint.
- Prefer weeks over months for the first engagement. Both sides learn whether it works.
- Insist on working in the real medium early. Prototypes in a browser beat static screens.
- Make sure one person is accountable for the outcome, not for their stage of a relay.