From agency to startup as a Product Designer

Making the move from shiny design solutions and award-winning agencies to dashboards and data-driven design is both challenging and rewarding.

The story summarizes 7 universal challenges of a product designer in a fast scaling environment and highlights why prioritization matters, how to balance speed and quality, and why a hyper-growth environment is a great skill booster, but no career booster.

The time I worked in agencies was full of fun and learnings both personally and professionally. I had the blast to work with talented and creative people and designed for various clients in the areas of mobility, industry and lifestyle. After some years, the idea of working on a product long-term, iterating and improving based on real data and user needs, sounded like the type of work I wished to do. I wanted to be free from relating purely on client input and working on something having a real impact on people.

Freedom and responsibility

Being a designer in a startup is different from being a designer in an agency or a large established company. In big design teams, you have the chance to take less responsibility, spend a lot of time on crafting mocks, and work in an area you like the most. This could be concept and UX, visual design or motion. In a startup, however, you have to take a lot of responsibilities, work end to end, do all kind of work and move with extra speed.

People will barely tell you exactly what to do or assign you a perfectly defined ticket. So do not wait for people to tell you what you are supposed to work on. If you see something that can be improved, it meets the business goals and improves the users’ experience, start the initiative.

01: Priorities vs. Brand Identity

First of all, things like outstanding brand identities or very unique interfaces are not top priority. No product will become successful and a unicorn just because of great design. When the company is in a hyper-growth state you have to prioritize your customers and find new markets and product priorities. Things are constantly changing, meaning that for example requirements or even the whole product, on which you have all been working at some point, may change its direction.

A good approach is to solve interaction and usability problems first. The interface should be clean and accurate, while brand identity shows the direction and the meaning of the product. Once you succeed the company will hire some legendary teams to develop a very high-quality brand, like e.g. Dropbox or Spotify.

02: Design Critiques: Work as transparent as possible

A startup is a small team of people with the same mission. Be open for objective criticism and accept good ideas suggested by anyone. An effective format in terms of time and speed is a design critique session with stakeholders and developers during your process as well as common scrum formats such as sprint reviews with a dedicated design slot.

Show your work at an early stage and gather all the feedback — make sure to do it in a smart manner so that you get something valuable. Ask developers to do the same when developing frontend. This can be tough but it enables you to be not only a pixel crafting designer but a facilitator of all design-related processes. Guide everyone to generate as many useful ideas as possible. So you can twist them into a robust and solid solution. Design sprints will help you to make this happen.

03: Speed matters: Use Design Sprints

Scale does not mean hustle and grind. Design sprints can be used to solve big problems, for example creating a new feature which may make you outstanding in the market and get a lot of new users and investors.

A design sprint will help you to create low or high fidelity prototypes, which can be tested and evaluated within 4-5 days. That means that you can show stakeholders what the new feature is gonna look like, share the user test results and iterate within one sprint. Document decisions, processes, ideas and findings should be made accessible for everyone, e.g. in Confluence or Notion and start with a TL,DR (everybody is always in a rush). Almost everything is useful at some point and it may prevent confusion later.

Afterwards you can spend the time you need to publish a great working feature with confidence. So design sprints help you to move fast without losing quality.

One more note here. Some are of the opinion that design sprints work only for bigger teams. According to the lean startup methodology, the MVP goes first. From my experience, I can say that it is either no time planned for a MVP or the so called MVP will be the final version for a very long time. What is usually done is to decrease the quality to meet the deadline. As a product designer go one step back, help focus on functional testing and persist on design iterations later.

04: Component library vs. design system

In early-stage startups there will be no resource and no time for a design system. But the good news is, a well developed comprehensive component library is not only the single source of truth but also a solid starting point for a design system later, when the day has come.

Your component library does not have to be outstanding with a trendy visual design but it should be accessible, clean and accurate. Put some effort to build it great and think about the bigger picture when you design it. Try to see how it all adds up together.

Focus on a couple of essentials first, such as colors, typography, icons, buttons and basic elements, like input fields or other form elements. Start for example with default states and add others at a later state once the direction or feature is aligned. Create components as soon as possible and use tiny little helpers such as the Figma plugins that convert all regular elements into one component.

This makes not only your design process faster but enables you and your team to onboard freelancers, new design and developer colleagues fast without any lacks or infinite back and forth in their first weeks.

Another important case as mentioned above is when the company changes the direction or decides to ship a new product within only a few weeks. This means that there is no time to invent new styles, components and patterns. You have to re-use what you have, what worked well and develop it further or iterate during the process.

Sometimes you will have downtimes between tasks or projects. Use this time is to improve and polish the components. Believe me, the time you will spend on this will pay off during the creation of new features or assembling another prototype.

05: It’s all about the numbers

Collecting data is a tool in the design process, but not the solution. The most specific quantitative data is not enough to tell you what needs to be iterated or improved in your product.

In general, data driven design uses two types of data: qualitative and quantitative. Any kind of data is valuable, a combination of qualitative and quantitative data will help you to complete the understanding of the context and to measure whether you’re failing or succeeding with your feature or product.

Quantitative data is objective data that can be measured through concrete numbers or values. You can get this data from the following sources:

• Analytics data, from sources like Google Analytics or Mixpanel, shows page views, clicks and bounce rates.
• Heat maps which show exactly where users click on your website or product.
• A/B testing which tests single variants to see which performs better.
• Usability test results that measure how long it takes for users to complete a task or find a specific piece of information.
• Multivariate testing, which uses multiple variants to test a hypotheses.
• Usage analytics which comes from actual user interactions with a product or service. This data can give information about how users navigate through a website, what features they use the most, and how long they stay on a page.

Qualitative data offers deeper insights into the why behind users behavior and can be gathered via methods like:

• User Interviews, where users share responses to a predetermined set of questions.
• Focus groups, where a group of users talk about a product or experience in a moderated conversational setting.
• Surveys, where users leave written feedback on their product experience and opinions.
• User feedback, which is gathered from surveys, interviews, focus groups, and other research methods. This type of data tells us about users’ needs and preferences.

The next step is to analyze and present the results from your research to uncover the themes or insights that will inform your design decisions moving forward. Qualitative data can easily be an analyzed e.g. with an Affinity Diagramm, while quantitative data can be summarized more straightforward because specific patterns will often arise on their own e.g. in Google Analytics.

It’s very helpful to create a well-rounded presentation or documentation that shows not only the data results, but also connects your research to other areas of the business. Use visualization tools like graphs, user quotes ideally with photos and draw correlations with stakeholder requirements to delight your audience and realign their goals with user goals. The more compelling your research and presentation, the more likely it is to achieve the results you’re hoping for.

06: Define metrics for Design

Performance in agencies is mostly evaluated in yearly reviews. Abstract metrics such as idea excellence, creative collaboration or the willingness to go the extra mile, are valued in skill matrixes. The responsibility of measurable key indicators is mostly on clients side and as a designer you neither have relevant touch points nor have to follow up on them later.

The situation in startups is different. Regular performance reviews on a quarterly or half-yearly basis are a standard for everyone and individual performance is evaluated by achieved key indicators, such as KPIs or OKRs. These indicators are defined on a company-level as well as on team and individual contributor level.

It is easy to track all events and properties and the first intention is to measure all the things. But if we report on everything, everything will become noise.

So one can say that KPIs are a smaller data set that help us to make decisions and take actions. Selecting product metrics is a journey, not a destination and just as the product itself they need to be iterated. My favorite framework to think about the right key things to track is HEART by Google, which reflects the quality of user experience and addresses how users interact with the product and how satisfied they are.

07: Career growth traffic jam

At a certain point in the hyper-growth state many new leaders with job titles you probably have never heard before are going to join the company. Roles and responsibilities will shift constantly while time for mentoring is lacking and developing individual growth paths is no priority at all.

One of the main reasons for this is that all the company effort goes into funding completely new teams and initiatives. So it might feel scary at some point and you start wondering if this is stunting your growth and when your investment in your career will come. These are very real challenges especially scaling companies struggle with while frustration in the teams is growing and colleagues which became friends are leaving.

To succeed and grow further, despite the limitations your team might face, try to identify areas for you and your peers to stretch. Align with supervisors on opportunity roadmaps instead of career paths and use the time as a skill booster. At some point it will equalize and then you have a really good group of designers and a solid leadership pipeline.

Some more eye opening insights on scaling design teams shared David Hoang, Design Director at WebFlow at Figma Config 2021.

✌ Good reads and talks

→ The Design Sprint Book: How to Solve Big Problems and Test New Ideas in Just Five Days

→ Design Critiques: Encourage a Positive Culture to Improve Products, NN Group

→ The universal challenges of every scaling design team, Figma Config 2021