Which company are you: the Tinkerer

In an ongo­ing series of posts on the dif­fer­ences between large tech com­pa­nies, I look at the dif­fer­ent mod­els they take (refine,tin­ker, push, attach) and who their spir­i­tual chil­dren may be. In this entry, it’s all about the tinkerer.

The tin­kerer: Google

At Google, the algo­rithm knows  all. Their view is that data and com­put­ers are the best way to approach prod­uct design.

If a com­puter can do it, it is prob­a­bly the best approach seems embed­ded in the think­ing behind their prod­uct deci­sions, often result­ing in prod­ucts that appear bland but are used by mil­lions of people.

A reliance on A/B testing

The net-net of this approach is a strong belief in A/B test­ing, where two ver­sions of a page are shown to users, each with a very small set of changes. The one that results in the desired action ends up being the cham­pion, mov­ing on to the next round against another ver­sion, repeat­ing the cycle time and time again. So, at Google, deci­sions are not made based on what the prod­uct man­ager or peo­ple at Google think but rather as a result of exten­sive test­ing. For exam­ple, an arti­cle in the New York times pointed out:

a prod­uct man­ager tested a dif­fer­ent color with users and found they were more likely to click on the tool­bar if it was painted a greener shade.

As triv­ial as color choices might seem, clicks are a key part of Google’s rev­enue stream, and any­thing that enhances clicks means more money. Mr. Divine’s team resisted the greener hue, so Ms. Mayer split the dif­fer­ence by choos­ing a shade halfway between those of the two camps.

Her deci­sion was diplo­matic, but it also amounted to rely­ing on her gut rather than research. Since then, she said, she has asked her team to test the 41 gra­da­tions between the com­pet­ing blues to see which ones con­sumers might prefer.

That type of exten­sive test­ing may be good prac­tice (and it seems that it’s work well for Google to date) but it’s not the type of cli­mate every­one nec­es­sar­ily likes.

Trust the data

In a way, Google’s trust in the data given to them by their con­sumers can be seen as a large influ­ence in the way com­pa­nies now design games. The whole social games move­ment, for exam­ple, is largely based on data min­ing and an under­stand­ing of inter­ac­tion metrics.

Today, Zynga is arguably the most suc­cess­ful pro­ducer of social games in the world, thanks in large part to the ini­tial suc­cess of Far­mville. Zynga’s chief designer, Bryan Reynolds, recently explained how the com­pany was using a data-driven approach to its design:

Zynga doesn’t rely on gut instinct to zero in on what users really want. Reynolds said Zynga fol­lows an array of real-time met­rics in order to find out what play­ers like, and what they don’t.

One exam­ple was of a screen from Far­mVille that pro­moted another one of Zynga’s games, PetVille. The font used in the pro­mo­tion was orig­i­nally red. By exper­i­ment­ing with other col­ors, the stu­dio found that pink fonts, strangely, gen­er­ated an expo­nen­tially higher click-through count than col­ors includ­ing pur­ple, green, and red. With­out met­rics, Zynga would have never known that.

“Using the data min­ing, the met­rics, you are able to learn the things that are counter-intuitive,” said Reynolds.

The impact can also be felt across most star­tups today as investors now require you to track and hit spe­cific met­rics. In a lot of ways, this is a saner approach to invest­ment than what we wit­nessed in the 90s (when com­pa­nies were get­ting funded based on poten­tial exit instead of busi­ness basics) so one can’t com­plain about the con­cept but I would ven­ture that bal­anc­ing out met­rics with other forms of input is also impor­tant in order to estab­lish the best long-term strategy.