A handful of times a year, clusters of Twitter engineers will line the walls of the company’s cavernous main hall, gathering around smartphones, monitors or poster board setups. The large, open-air room buzzes with excitement as employees show off the slap-dash project they hacked together over the past four days. It is Twitter’s officially sanctioned “Hack Week” – a time where all bets are off and employees can work on whatever they want.
Project Mixtape, for example, was a way to suggest and send a list of good people to follow for any friends you’ve invited to try Twitter for the first time. Another group’s project involved a small Vine button stuck inside the Twitter app, a quick way to take and tweet videos inside of Twitter proper. And one set of engineers developed a way to play with location-based discovery of nearby users and tweets from people you weren’t following.
Many employees see the event – the Friday that ends Hack Week – as Twitter at its best, a testament to the company’s capacity for innovation.
But some view these days as among the most depressing of the year: A parade of Twitter features that will never see the light of day.
Simply put, Twitter has a product problem. Engineers inside the company have long grumbled that there are few direct paths for moving product changes up the ladder at an efficient pace. Twitter’s occasional “experiments,” or ideas for change, sometimes sit in their requisite testing phase with 1 percent of users for an inordinately long amount of time, stagnating without any decision being made to move the product forward or kill it.
In a nutshell: It’s a product culture that’s basically the opposite of Facebook’s. Where the social giant is willing to “move fast and break things,” sometimes pushing out new features in the span of weeks, changes to Twitter’s core product often languish for months – or even years – in limbo.
While the issue is hardly a new one for the company – product pipeline problems have plagued Twitter for years – the stakes now are especially high; Twitter has to answer to the public markets, as well as to its public user base. Being slow to innovate, or even to iterate, could affect the company’s ability to push its microblogging service deep into the mainstream. And to some degree, it has inhibited the company’s ability to retain talent. Twitter has already lost a number of engineers and product managers who were frustrated by its failure to ship things in a timely manner.
The problem’s origins are two-fold, sources said. The product divisions are messy, with an absence of efficiency and sometimes a lack of clarity around which executive has the last word on whether a feature change is good enough to ship. Many in the organization have also expressed frustrations about working with Michael Sippey, Twitter’s VP of product, whose hesitation on pushing changes through has created problems with some middle managers. (My colleague Peter Kafka interviewed Sippey at our Dive into Mobile conference last spring.)
The other problem is perhaps the larger one: Twitter’s aversion to risk when it comes to its marquee product, the core Twitter timeline. According to sources, some of Twitter’s top people – including a few from the company’s executive ranks – fear any major changes will meddle with the company’s revenue products. These products, in just a handful of years, took Twitter’s revenue from zero to hundreds of millions of dollars.
As one insider put it: “Twitter hasn’t been playing to win. It’s playing not to lose.”
When asked, Twitter declined comment on any current or future product plans.
The situation isn’t hopeless.
Twitter has shown some willingness to change in just the past few months. A few weeks ago, IT pushed out in-stream photos and Vine videos in a major enhancement to the home stream – an area the company rarely touches.
It should be noted, however, that these changes had been ready to go for years, according to sources, born out of an idea called “Project Skyline” – the result of yet another Twitter Hack Week. Due to its existential product struggles, however, Twitter didn’t push the new design out the door until this October.
The company also introduced a thin blue line to its app recently, an attempt to make it easier for infrequent users to get up to speed on conversations happening between people they follow.
But that, too, was controversial. According to one senior Twitter executive, the company considered killing off the blue line altogether just weeks after introducing it to widespread dissatisfaction. (Whether Twitter will actually go through with killing the conversations line isn’t clear. For now, Twitter swapped the blue for a gray line, one hopefully less obtrusive to users.)
Twitter is also working on its personnel problem. Sources tell AllThingsD that VP Michael Sippey now reports to COO Ali Rowghani on growth and product issues instead of directly to CEO Dick Costolo and the board – an apparent demotion. That may help with some of the complaints around process and expediency. Multiple insiders, however, said this move was a head-scratcher, seeing as Rowghani has little background in growth and product issues. Not to mention that it is atypical for a tech company’s VP of product to report to the COO.
The biggest indicator of a product division shift: Twitter’s upcoming redesign, a makeover which truly attempts to court the mainstream. We’ll hopefully see that, or steps toward it, before the end of the year.
Keep in mind, this “Twitter reinventing itself” story is nothing new. Twitter has already undergone multiple internal reorgs in the span of just two years, removing some product and engineering execs and replacing them with others – all to no avail.
Perhaps, if the product division can truly change its mentality, the company will actually start to live up to its somewhat ironic “#shipit” motto that some employees chant.
As always, though, there’s no guarantee it’ll actually work.