Oh lord, Seamless. Do you want me to starve to death?
That’s what a lot of people have been asking themselves since the popular food delivery service rolled out a dramatic redesign a few weeks ago. Nothing is where it was before. The new interface’s huge elements feels like a prison of clutter compared to the old design. In a painful throwback to less enlightened days of web design, search results were suddenly paginated. 18 pages? R u serious? Worse, the redesign brought with it bugs where there hadn’t been any before. Users reported simply not being able to order, location detection problems, and other errors.
Yes, Seamless is serious—and as its CTO told me this week, it has evidence to back up the new changes. Is it possible the real problem is our grand, time-honored internet tradition of throwing tantrums when a website or company makes a big design change? Possibly. But it’s so fun.
Negative reactions to new designs is hardly a problem unique to Seamless. Every time products with devoted user bases change anything, the users revolt. Look no further than the recent rebellions at Reddit for evidence that passionate users want things exactly the way they want them. Traditional companies encounter this problem too. Today, the choice to redesign—or even just alter—a logo is a decision depends on just how wide the public’s collective scoffing could go: The Gap. Airbnb. Even Logitech.
Nor is Apple immune. Following the dramatic design shift from Apple’s iOS 6 to iOS 7, Christina Wodtke neatly summarized the problem, often known as the “9X effect:”
So when a big change comes, the end user is focused on what they have lost: productivity, comfort, familiarity. And the users weigh that loss as three times more important than any gain that company professes to offer. The exact same math applies to redesigns: you moved my cheese and I am not happy about it.
The 9X effect, which seems to have first been described in the early 2000s in a Harvard Business Review essay by John Gourville, works like this: A product’s users overestimate how beautiful and useful the original product was by a factor of three. On the other side, the makers of the product overestimate how beautiful and useful their new product is by three, too.
In the end, the redesign ends up being the focus of intense nostalgia and optimism on both sides—a fiery mix that only time and use seem to really dampen. In some cases, it consumes the new product completely.
When I first saw the new Seamless I wanted to seize my iMac and throw it out the window in protest.
I almost always order food using Seamless when I am too busy to even think about how I’m going to put calories in my body. I can get from the front page to burrito almost without thinking. Or at least I could—until the design switched over. All of a sudden I had to use my brain an click my way through a labyrinth of decisions that don’t matter to me. Why the hell do I care what the delivery fee is? Have I ever looked at this before? You sold me without information, why do I need this information now?
My friend, the old Seamless, above, followed by the redesigned Seamless.
According to Brian Lanier, the CTO of Seamless parent GrubHub, the design was rigorously tested and everything you see on the page right now is based on real usability data. The new look might be jarring to existing users, it’s actually across the board better for new users. “The clickcount to an order is consistently lower,” says Lanier, who did have a harder time explaining away the clutter of the design is a little harder to me. According to him, people have “different modalities” for using Seamless, which is reasonable. For me, the clutter could still use some well-placed hierarchy.
One seemingly huge mistake with the redesigned Seamless was that it hid its new features so well. If you look closely, Seamless is actually subtly better in some ways. Search has been improved so that you can search for terms more descriptive of the food you’re trying to eat, like “vegan” or “gluten-free.” If you’ve ever tried to search for “salad” on Seamless, you know searching for food types was a terrible way to find anything. Other improvements: You can now pay for your food using cash and PayPal.
But here’s the rub with those (very useful!) new features: Seamless’ loyal user base was already using the site without them. They’d adapted to the shitty search functionality and the payment hoops. They’d grown used to these design quirks.
This new menu page wastes a lot of space and requires a lot of scrolling. It gets one star, just like the Orion Diner and Grill.
In psychology, this is called “status quo bias,” another term that Gourville used alongside his explanation of the 9X effect. It states that users value something that’s familiar over something that adds unfamiliar functionality, even if it’s useful. In a weird way, the backlash against the new design was proof of Seamless’ health as a product, according to Ian Burns, a Creative Director at the design firm Huge. “For a lot of products that don’t have fanatical user bases, people don’t even notice redesigns,” he pointed out.
We won’t know whether the redesign was a true disaster for a long time—right now, we just know that it looks like one to users’ eyes.
So what can be done? Are users and companies destined to be forever at odds due to the 9X effect? Should companies bow to the pressure of the internet rage machine and never change? Should all websites look like Geocities circa 1997?
Lanier says the redesign was part of a necessary and inevitable update: An effort to modernize Seamless’ backend. GrubHub and Seamless merged in 2013 and the two companies were sitting on entirely different technology stacks. Fair enough. But you wouldn’t be wrong if it sounded to you like your loyalty had been sacrificed to the altar of growth.
Still, as user experience expert Patrick Neeman tells me, a little bit of outrage is OK for a company like Seamless. “There’s always the risk of making somebody mad because you move things around,” he says. They have a built-in safety net: The fact that it they have so many regular users without a better alternative.
Where a redesign can sink a product is when a company doesn’t have a user base. In the case of a product like Facebook, Reddit, or even Seamless, outrage is a side effect of continuing to use the product anyway. “The users will come back because there’s nothing else that can replace them,” Neeman adds.
Back in 2006, when Gourville first talked about the 9X effect, he also spent some time showing how companies have avoided it. The lessons were simple: Pay attention to how users behave, and then try to replicate it with the new thing, even if it’s different. “Until businesses understand, anticipate, and respond to the psychological biases that both consumers and executives bring to decision making, new products will continue to fail,” he wrote.
Which brings us to the part of the new Seamless that can’t be rationalized: The reason why it looks so different, even if it acts pretty much the same. In the end, completely overhaul design just because you’re adding some functionality and cleaning up the backend is just asking for trouble. Maybe that’s Seamless’ biggest sin: In the process of fixing some stuff users had forgotten was even broken, they fixed something that worked just fine.