As businesses increasingly prioritize good design and as we start demanding it in our daily lives, more and more technology companies will face tensions over how to incorporate design principles without compromising on the engineering goals that can also make or break a product or service. It’s a challenge that every company handles differently, and in many cases, one group wins out at the expense of another’s ideals.
But at Pinterest, the social media darling that has seen incredible user growth over the past few year and whose design has copied widely, the designers and engineers have one seemingly-improbable message: We can do both. At least for now, anyway: one of Pinterest’s greatest challenges in turning from web sensation into a viable business will be its ability to adhere to these principles as revenue and profit pressures sink in.
I talked with Pinterest software engineer Joshua Inkenbrandt and lead product designer Jason Wilson this week, and we discussed the recent Pinterest re-design: specifically, how both visual and engineering elements played into the overhaul, and how those tensions play out at Pinterest. The two were obviously passionate about the issue, re-iterating the unique qualities of Pinterest’s internal culture and the crossover between their roles. At one point, Inkenbrandt referred to a feature that Wilson had designed, and Wilson automatically jumped in to correct him:
“It’s the designs that I visualize. We’re all designers.”
But at an awful lot of companies, designers and engineers aren’t finishing each other’s sentences. Kleiner Perkins partner and former Twitter VP of engineering Michael Abbott explained the tensions best at our Structure:Data conference this week: “The tension on the design side is that it’s never good enough,” he said, “and on the engineering side he or she wants to ship… How do you get that balance? Because you still need to ship.”
Building a culture of mutual respect
It’s not that Pinterest is somehow immune from the tensions that other companies face. In fact, Wilson and Inkenbrandt noted that they think the media has probably underplayed just how much the site has changed over the years and the compromises that are inherent in every tweak:
“With every feature, there’s that tension between it being beautiful and then figuring out how hard it is to actually implement it. So everything you see on the new site has actually gone through that balance,” Inkenbrandt said. For instance, just looking at the main Pinterest grid — which doesn’t feel like it’s changed all that much since the site launched in limited beta in March 2010 — you would never know the number of iterations it has gone through.
“We’ve iterated on that 70+ times,” said Inkenbrandt, who just joined the company slightly more than a year ago. Of course, rapid iteration on the web wasn’t invented at Pinterest: Google’s web development strategy in its early years, overseen in large part by current Yahoo CEO Marissa Mayer, blazed this trail for a generation of Silicon Valley engineers and designers.
In the relatively short time Inkenbrandt has been with Pinterest, the company has grown from about 17 employees to more than 100. Pinterest does not release official user stats, but recent ComScore numbers put users of the popular social content pinning site at more than 48 million unique visitors globally, and a February Pew report put the percentage of online adults using Pinterest at 15 percent, which is about the same as the number who use Twitter. The company recently raised $200 million in a new round of financing, putting its valuation at $2.5 billion.
So the fact that the company has managed to weather this type of rapid growth without any notable or constant outages (just compare to the number of fail whales in Twitter’s early years), while also remaining one of the most-copied designs on the web at the moment, certainly speaks to Pinterest’s ability to balance both priorities.
It comes down to the company’s culture, the two said, in that the company values open communication among employees and prioritizes the goals of both designers and engineers. Of course, whether this can scale along with the company is always in question — it’s easier to support both when you’re just a 20 person operation. But the two said that heavy involvement from co-founder Evan Sharp plays a key role in fostering that balance.
“Evan himself designed and coded the original site,” Wilson said. “So we know that at the end of the day he understands what it takes to get it done.”
Wilson, who’s designed for companies like Apple, Lytro, and Adobe, said he thinks the company’s efforts to make sure new hires understand the importance of collaborative culture has helped keep the company on that track:
“I’m basically a vagabond in the industry. I’ve worked at a hell of a lot of places. And every place has its way of doing things. When I worked at Apple, it was very design-centric, and things came from the direction of the executive and it’s thrown over the wall at the engineers. And I’ve worked at places like Adobe, where they develop the features and then say, make it look good. And I don’t believe personally — and this is just a personal opinion — that either of those really are ideal. It’s not bullshit to say that here, it’s a total team effort. I can point to a design element, like actual pixel-level design elements, that Josh gave me creative direction on.”
Design and engineering building the new look
Pinterest rolled out its new design to the general public earlier this week, and to the un-trained (or un-Pinterested) eye, the new look wouldn’t seem too different. But two key changes to the look — larger pins, or photos, that come with the content suggestions along the side, and a new way to remain in-stream while scrolling — actually started at engineering problems that were solved by the design team.
The engineering team had long wanted to fix some of the structural issues that had come with the site’s rapid growth, Inkenbrandt said, and one of them was how to suggest more content for users in a way that was both natural and useful. But in the end, it was Wilson’s design team that found an answer to the recommendation engine quest, by surfacing similar pins from boards users had created:
“One of our challenges was how to recommend content,” Inkenbrandt said. “So we’re coming up with like, secret sauce algorithms, where we’re trying to figure out based on content what people might like. And then it turns out that based off the design, just having the actual board where the pin was from, well, that content is obviously highly relevant, usually in the same theme, and can help people find great stuff.”
By showing related content next to individual pins, and then allowing users to navigate to that content without leaving their spot in the stream, also solved another engineering issue with navigation. Wilson said Sharp gave him and the designer some high-level goals for the re-design, and one was to make the overall feel much more lightweight.
“Before, you would look at one close-up and then go back to the grid and then back to the close-up. So we wanted to figure out a way so you wouldn’t have to do that jump. It seems like a simple thing, but if you’re on there for an hour let’s say, and you do that repeatedly, you might do that a few hundred times.” he said. “To me anyway the holy grail to navigation is to get to the user to navigate via content.”
But while Pinterest has been able to balance the priorities of design and functionality fairly remarkably so far, the rubber might be about to hit the road. The company has been nudging its way toward a business model recently, adding features like analytics for businesses and encouraging marketers to the site, which makes sense if the company is going to live up to that $2.5 billion valuation.
But just ask Twitter and Facebook: it’s a lot easier balancing design and engineering when salespeople aren’t breathing down your neck every day.