[Editor’s Note: Guest commentator David Lifson of General Assembly is one of the speakers at the upcoming HIVE technology design conference, June 20 and 21 in Seattle, presented by AIGA Seattle, the professional association for design. Tickets available here.]

HIVE_eventbriteHeaderIn today’s fast-moving world of lean startups, minimum viable products, and an Internet full of competitors and alternatives, rapid product development is required. For designers, this means getting comfortable with a concept I call “iterating in public.”

Most designers are taught to think like artists, with the goal being to produce a design that is as close to perfect as possible before revealing it to the public. The traditional (and mostly abandoned) waterfall product development process highlights this: design is a step that is discrete from development. With industrial or print design, this makes a lot of sense — once the design is manufactured and delivered to stores, that’s it! Yet, who says the same process should apply to digital product design?

When it comes to digital product design, the medium is code and the delivery is the internet (as opposed to software sold on disk or CD-ROM), so the constraints of manufacture and distribution no longer exist. With that comes the freedom to get to market as quickly as possible and iterate based on qualitative and quantitative feedback.

hivepullHere are four steps to successfully iterating in public:

  1. Identify your goal. Are you looking to improve sales by increasing conversion rate? Increase the number of registered users?
  2. Don’t get hung up on the details. You can fix it later. At this stage, focus on the design aspects that directly impact achieving your goal, which is typically content and layout.
  3. Build it and ship it. Find a developer, sit next to them, and get to work. Or, code it yourself! If you have an existing style guide or template, use it. If not, keep it simple. The focus is on testing if your design achieves the goal you identified.
  4. Look at the data. Talk to users. Iterate. Consider both qualitative and quantitative data. What does Kissmetrics say about your conversion rates? Do users accurately identify the goal of the page? What information is missing that would get them to convert?

Obviously, this approach isn’t appropriate for every situation. Consider the risk of launching with an imperfect design. Do the downsides outweigh the upside? Here are some scenarios where the more traditional approach would be better:

  1. Company re-branding. You probably want to do your time to run a full research process, as the statement your brand makes to customers is something you don’t want to change frequently.
  2. Regulation-heavy contexts. There are some situations where, before changes can be published, products need to go through time-consuming rounds of regulatory sign off. Obviously, the cost of iteration outweighs the benefits.
  3. High security requirements. The software that runs the stock market, or bank ATMs, or air traffic control towers are not safe for rapid iteration. The consequences that would come with accidental introduction of software bugs is a cost too high to bear.

Iterating in public can be an uncomfortable process for many designers, but in the right context, it can be a very powerful approach that leads to better products in less time. All it takes is thoughtfully identifying your goal, designing and implementing that core experience, and iterating on that experience as often as possible based on metrics and user testing.

David Lifson is General Manager of Academics at General Assembly. Formerly, he was GM of Digital Product and Engineering, responsible for all technology at GA, from e-commerce website to internal tools to online education delivery. Prior to joining General Assembly, he co-founded Postling, a social media management tool for small businesses that he sold to LocalVox Media in 2012. He has also been VP of Product at Etsy, Technical Product Manager for Amazon’s Personalized Recommendations team, and started his career as a software engineer on Amazon’s Community team. He also volunteered for 3 years as the official AngelList Scout in NYC, connecting over 450 investors to high-potential startups.

Comments

Job Listings on GeekWork