Google’s Predictable Failure

Does Google have a Magic 8-Ball?
Image by Carolyn Sullivan

The ability to predict future market trends is a talent many long for, but short of trusting in magic 8-balls, there is no way to do it. In lieu of accurate eight-ball readings, businesses turn to historical analysis to learn from past trends, and similarly, many designers rely on past experiences to approximate users’ future decisions. This isn’t a negative thing – analyzing past trends is a great starting point to understanding user behaviors. In his book Predictably Irrational, Dan Ariely describes this analysis as necessary to “understand irrationality” – a crucial step in understanding – and altering — future behaviors, both personally and in marketing. But learning about the past is not, in and of itself, enough to predict the future.

The Science of Predictability

Google offers an example of a company that seems to understand what users want, and when to offer new innovation. From their initial launch as a search engine in 1998, Google has been pushing the envelope on what should be expected from a product or service.  In 2005 the one-year old Gmail was well on its way to eclipsing Yahoo Mail as the top email client. Much of Gmail’s early success came from offering all that Yahoo Mail did, but with more storage for free and an invitation-only registration to create the aura of exclusivity many users can’t resist. 2005 was also the year Gmail added the Labs feature, allowing users to test out new Gmail additions while in beta form. The success of Labs is unsurprising– Labs is marketed as yet another exclusive “first sneak peak” club, and by introducing users to potential future Gmail features Google paves the way for future innovations. Time and again, Google has proven that user trends can be predicted with the power of historical analysis.

Yet even Google doesn’t have a perfect track record. In 2009, Google released Wave, a platform designed to be stunning, magical, and everything a user could want in an email client/messaging system/document production tool. Released at Google IO 2009, Wave was met by awe and excitement, and yet in 2010, only 3 months after fully launching, Google stopped further development on Wave. Wave was officially dead in the water. What happened? Did the users suddenly change, or did Google’s system fail?

Making Waves

There are many theories as to why Wave failed where Gmail and so many other products succeeded. One possibility, suggested by Mashable’s Christina Warren, is that Wave did not have a clear target user. To take this one step farther, maybe Google did have target users, but the users were less predictable than history had led Google to believe.

Google followed many historical trends from their previous product experiences when creating Wave.  By combining disparate and independently popular applications such as chat, email, text documents, and social networking, Google created Wave as a veritable Frankenstein’s monster of previously successful products. Wave’s success as a product was based on the prediction that users would prefer an all-in-one application to the myriad networking sites, time saving apps, and procrastination devises cluttering their browsers. To a culture growing overwhelmed with excessive input and clutter, Wave should have been a dream come true.

And when Wave was first unveiled in May of 2009, the reviews were almost across the board enthusiastic. Invited users jumped on board, grabbed the available invitations, and from all appearances loved Wave, as predicted. However, 2010 brought a decline in usage that was anything but predictable. Though the initial hype was just what Google expected, the lack of long-term adoption shows that maybe the users were looking for something else.*

Waving Goodbye

Another theory is that Wave was simply too far ahead of its time. While it’s possible that this is the case, it’s an incomplete theory. It leaves open the question of what made Wave ahead of its time, or why Google made the mistake of introducing it at the wrong time.  A more complete theory can be drawn from the reactions of the users.

At Google IO 2009, where Wave was unveiled to glowing reviews, the audience was comprised primarily of developers. These same developers jumped at the available invitations to test out Wave, and many continued to make use of this valuable product. Wave, according to many developers, was (and still is) a very impressive development platform. Whereas previous Google products began their beta lives in the hands of developers and quickly spread to the public through the use of those exclusive invitations, Wave did not make that transition. From May 2009, at Wave’s unveiling, until May 2010, at its Public launch, the product remained contained in the development world. Google may have assumed success in the development world to be a promise of public success in the future, based on their previous experiences. In reality, Wave might have had a stronger user base of developers, had they focused their marketing there.

The Science of User Research

There is a lesson to be learned from Google’s experience with Wave. Google had reason to believe it’s users to be predictable: with Wave they would follow the same pattern of adoption from beta to public release as other well-received Google products.  However, predictability among users is not a reliable metric. If Google had chosen to find out which users were engaged in Wave from the start and focus on their needs, Wave might have succeeded as a development platform. Instead, Google relied on the past trends of other products.

What Google has hopefully learned is that successful products are less reliant on following trends, and more concerned with creating the best overall experience possible. Gmail’s success was built from researching the previously successful Yahoo Mail, and making improvements based on research that showed what users were missing. By doing thorough user research, and filling the gaps for users, a successful product becomes predictable: you know definitively what the users want, so all that’s left is to provide it. Some of ATF’s tried and true methods of research include usability testing, field studies, or ethnographic interviews.

User research is a science, but really listening to users is an art, and perfecting it takes practice. At the end of the day, listening to users requires that a company weigh their own assumptions about their product against what they learn from their users. Here are a few tips for how to balance your assumptions with your users’ needs and feedback:

  1. Be aware of your assumptions. Collect assumptions from your team and use them as a starting point for understanding your users.
  2. Conduct Ethnographic Interviews to learn about who your users are, what time they have, the space they work in, their motivations and goals, and their needs when using your product. Use this research to create personas to help identify your target audience
  3. Set up a round of in-house usability tests to validate your assumptions. (Why usability tests and not focus groups? Dan Ariely reminds us that “we respond differently to a question if it’s worded differently”; usability tests focus on open ended questions to understand what a user sees, alleviating this potential concern.)
  4. Listen to what users said (and didn’t say) during testing.. Hone your listening skills to read between the lines of each usability test; what were the users really struggling with? Is the problem they saw really the root cause? What are the most important things that are preventing users from doing what they need to do?
  5. Revise your assumptions (and your product) and test again. See how your new (or revised) assumptions stand up to the test.

User research as a necessary step in product development is a step towards becoming “the next Google.” Wave aside, that team has a strong track record. Clearly the benefits to learning the “predictable” possibilities are a solid jumping-off point, and Google’s many successes are proof that they also possess an in-depth understanding of the users’ needs… even the unpredictable ones.

Why do you think Wave failed? Share your theory in the comments.

*As an interesting side note, Google has since created the Google Bar, and GTools for Google Chrome, both of which link the disparate Google products for users. This is a smaller step than Wave, a completely new application, and a less invasive one, while still offering some benefits for users looking to have fewer steps between Google tools. Sounds as though Google has learned from the user response to Wave. Read more about the Google Bar and GTools.