Designing for a startup is unlike any other kind of design. Teams are small, and everyone participates in design decisions. Limited resources mean that startups usually don’t have a dedicated designer when they need help.  And the pace of change within most startups requires speed, fluidity, and the understanding of how good design supports your business.

After working with 60+ portfolio companies over the past few years, my colleagues at the Google Ventures Design Studio and I have developed a shorthand to help startups get through the thorniest product design challenges. Done right, good design can have a major impact on growth, engagement, and revenue.

Here are the three core design principles we recommend to startups:

1. Focus on the product

Often when people think about design, they think about the surface visual design. But early in a startup, visual design is probably the least of your worries. You need to find customers, understand what they need, decide what your product offering will be, and figure out a way to communicate that value to your customers. In short: you need to figure out what your product should be, not how it should look.

We focus most of our effort on product design (some people call it design strategy). We spend a lot of time gathering data, considering how design can impact key business metrics, building prototypes, and validating them with customers. If startups can iterate through this process a few times, they’ll be well on their way to building a great product.

2. Work together in sprints

It’s tempting to think that you can outsource product design: hire someone who goes away, does some design work, and returns with a magical solution. But because product design is so complex, it can’t be outsourced and solved in isolation. Great design requires the knowledge and experience of everyone on the team.

Our approach is to run product design sprints focused on solving a single well-scoped problem. Designers from our studio embed themselves with the startup for a few days (or weeks).

There’s a lot to running a design sprint, but one thing we’ve realized is that each person in a startup usually understands a different part of the problem.  Customer support knows where users are getting tripped up, engineers know the technical constraints, and the CEO knows the business goals. The first step is to get all of this latent knowledge out of people’s heads and into a visual representation so everyone can understand the extent of the problem. Then we can dive in and get to work.

By the end of the sprint, we’ve usually found a great design solution. And because we worked through it together, everyone on the team understands the design process and can apply it to their next big challenge.

3. Get out of the building

If you’ve read about customer development or lean methodology, the phrase “get out of the building” should sound familiar. Basically, startups need to talk with customers early and often to validate ideas. But how do you do that exactly? Where do you find customers to talk to? What questions should you ask to get the best data? And how do you do all this quickly?

We took everything we knew from running rigorous user research studies, and found pragmatic ways to apply it at startup speeds. For example, you don’t need a fancy user study lab with one-way glass to talk with your customers. WebEx works just fine, and your team can watch the session from just about anywhere.

We also find that most companies do infrequent large user studies that generate thick reports. This doesn’t work for startups because it takes so long to learn from the findings — people often just make a decision without the data they need. So we like to run fast agile studies every other week. Once you get into this rhythm, it becomes easy to show rough ideas to customers, validate your hypotheses, and move even faster.