Designing at the speed of a startup

I work at a fast-paced company. We ship often. We measure twice before and twice after. We constantly push ourselves to deliver a better customer experience. The technology we’ve created is reinventing the way our users approach their jobs, and there is no blueprint for what we are building. We’re actively redefining an industry and I have fifty important things to do at any given moment.

Sound familiar? If so, you most likely work at a startup.

Startup environments create an interesting blend of challenge and opportunity for designers. Having worked as a designer at companies like Microsoft, Redfin, and Amazon, startup life, by comparison, prompts continuous lessons in the art of compromise. You don’t have time for lengthy process or round after round of opinionated design reviews, practices I’ve found to be commonplace at large organizations. You must move fast and keep your focus on shipping.

At Convoy one of our company values is:

Always have a sense of urgency: Don’t put off until tomorrow what can be done today. We think and act in days instead of weeks. Fast is better than perfect.

Executing in this way means we get features in the hands of customers as quickly as possible. This also means we capture valuable data and feedback from experiments much faster than most companies. This allows us to learn from real-world user interaction at an incredibly fast rate. However, designing at this speed can create some interesting tradeoffs –

1. Multi-tasking can spread one thin and make attention to detail difficult.
2. Time-consuming meetings are created by constantly shifting priorities.
3. Estimating aggressively means deadlines are always too soon.
4. Research has to be rapidly planned, recruited, executed, and synthesized.
5. “Good enough” design runs the risk of becoming the norm.

So what does a designer do when in this midst of all of this? There are some simple methods you can use to help balance the many inputs and monitor the quality of your output. Staying vigilant when making design decisions will show that it is possible to maintain a high bar of deliverable while balancing quality with speed.

1. Challenge yourself to design in less time

Set boundaries and see what you’re capable of. This will not work for all tasks as priorities and direction can move under your feet at any given moment, but you can test yourself to see how fast you can deliver a solution. The confines of a problem will help focus a design, and similarly, a time-boxed approach can help rapidly move it along. Be aggressive with your personal estimates. Think a design task will take 8 hours? Try to finish it in 4 and see what you come up with. You will most likely surprise yourself.

2. Accept that a design is never finished

Every design is constantly in flux. There is never a static state where an experience needs no consideration. Every flow is always in need of a refinement, even if it continues to provide positive feedback from customers and is a cornerstone of your offering. Designers need to completely ditch the concept that something is ever finished, polished, or “final.” This mentality will hold up decision making and drag out timelines by over-attaching you to a solution.

3. Communicate directly and frequently with your partners

Over-the-shoulder conversations have saved me countless hours in meetings. Regular in-person huddles are lifesavers when something is being rapidly adjusted in that ‘can we just tweak this a bit’ phase. Embrace the impromptu chats and use them to unblock anyone collaborating on your designs. Get to be as friendly as possible with the folks building your work. It will help create more frequent check-ins, and those discussions will help move the process along at a much faster pace than big over-the-fence handoffs or impersonal monologues at stand-up.

4. Choose your documentation wisely

Make them short, sweet, and limit documents only to the necessities. Stay lean on paperwork and heavy on problem statements, goals, and metrics. You don’t need to spend a lot of time spelling out every scenario. The mockups and prototypes will do that better than any table or bulleted list. On top of that, everything but objectives and goals will definitely change during a project. Use limited documentation to get the team on the same page, then use your designs to continue telling the story.

5. Test very often

Setting up a regular cadence of testing is necessary. The frequency gets you in practice, the insights continue to provide direction to your ongoing work, the team gets better at engaging with customers, and it becomes easier for all designers to participate in usability sessions. Test plans and scripts are helpful, but it’s more important that you are getting experiments in customers hands and recording what they have to say in an easily shareable fashion. Even just a few quick calls with customers can do wonders in pushing a design toward a better solution. During these sessions, you can also collect continuous feedback for deeper analysis in the future. The long-term goals of research don’t have to be sacrificed, but the short-term goals need to be focused on refining the project at hand.

6. Synthesize just the right amount of research

When handling a mix of research and design, time gets eaten up by busy work. Preparing and synthesizing a research session is usually more costly than running the session itself. Given that this is the case, use techniques to assemble feedback pertinent to the problem at hand and save related insights for later investigation. Focus on feedback that is immediately actionable. As long as there is a large enough sample size of insights for a specific problem, you’ll answer the important questions that keep your project moving. Continuous collection of research should be a steady practice that will always evolve your customer story, but for a quick turnaround project you should focus on the part of the story that helps you deliver immediate results.

7. No to wireframes. Yes to prototypes.

Forget wireframes. Wireframes can be a difficult concept for some users to understand. If they are not familiar with technology, or testing out new UI on a device, wireframes are going to confuse them. There is no reason to do wires unless you have a lot of lead time and a tech-savvy customer base. Convoy has neither of those and in all honesty, I don’t miss anything that wires could or should deliver. Every conversation during our fast-moving projects has been better served with sketches and prototypes. With all of the tools we have available today, the prototyping process should be quick and easy. If prototyping methods are not a fast process for your team, take steps to make it part of your weekly (even daily) practice until it is.

8. Compromise at speed

Get over yourself and learn to compromise for the sake of shipping and learning. You are not perfect and I’m pretty certain you are surrounded by a bunch of people who are equally as brilliant and talented as you are. Learn to listen and react with new designs quickly. Stand your ground if it’s a researched or mission-driven design consideration, but don’t become stubborn when it’s really your artistic pride at stake. If you can’t learn to compromise quickly on aspects of your design then conversations WILL get out of control. It will just lead to wasted time in reviews when you could simply timebox the decision, delegate quick consensus with the team, and get right back to refining deliverables.

9. Empower the devs to contribute to your design

If you are a designer+coder and your company is set up such that a designer can ship code, then awesome! Get into the code as early as you can and prep your work for code review as part of your design process. If not, then you should keep the coding to the keepers of the codebase. Deliverables that give an indication of interaction and style but don’t walk through every detail are sometimes sufficient when under tight deadlines. Be sure to get on the same page with your devs in person, and quickly talk through any gaps in interaction. This will help keep UX discussion to only the necessary tweaks. I’ve also found tools like Zeplin helpful in eliminating redlining, so make sure all of your devs can easily use any tools in the handoff process. Letting devs handle more of the design is also dependent on a style guide and component library of some sort — one that devs can easily pull from to quickly implement. That is a much bigger discussion, and I’ll get into that more in a future post.

Now don’t get me wrong on all of these points… if you or your team has more time for extensive fidelity polish, detailed microinteractions, and robust research cycles, then by all means spend more time refining. In a startup world of multitasking and quick turnarounds, that ability is often a luxury. It is okay if some deliverables tell just enough of a story about the design in order to save time. This requires putting your ego aside, and trusting in your team’s ability to deliver.

As a designer, you want your company to succeed because of the experiences you create. With a few practices like these, you can meet your deadlines and accomplish that goal while maintaining a high bar for your deliverables. Doing something of quality in a short period of time is always challenging, but no matter where you work this is a skill you must master to succeed.

— Jason Kopec