The Downside of Google's Data Obsession

| He didn’t announce it via cake, but Doug Bowman quit his job as head of Google’s visual design team last week, citing the company’s “reliance on data” for design decisions as the main reason for his departure. Bowman writes on his blog that he’ll miss Google’s “incredibly smart and talented people” and the “occasional massage,” but not “a design philosophy that lives or dies strictly by the sword of data.” We’ve asked before whether too much data can endanger patients and cause bad writing; might it also diminish a company’s workforce? (HT: Noah Harlan) [%comments]

Howard Tayler

Companies typically operate by making their decisions based on digested data.

Companies typically excel by making intuitive leaps based on the wild-haired ideas of a small number of charismatic individuals.

They also often fail by following wild-haired ideas.

Google didn't get where it is today by looking at data. These days they're just churning recycled ideas into the next quarter's business plan.


I'm not really seeing the downside here, other than that they're losing one employee because of it.


Google recently spelled out this very issue in its "official" blog. The idea being that data from web traffic is more important than assumed aesthetic qualities.


Seems silly to me. With millions of users and a system that is end-to-end customizable there is no reason not to test assumptions of the UI designer. Not to say that said wizard should not give direction and form, but that it's not out of line to validate those decisions with hard data from subtle user feedback/uptake of new design decisions.

Sounds more like the best being the enemy of the good in this case.

Dennis Rice

Eric, I think the key here is that the employee they lost is one they brought in to fill a very specific gap, and their reliance on data is preventing him from doing so. Essentially, it is not just an employee, but an entire approach that is being discarded.



"I had a recent debate over whether a border should be 3, 4 or 5 pixels wide, and was asked to prove my case. I can't operate in an environment like that. I've grown tired of debating such minuscule design decisions. There are more exciting design problems in this world to tackle."

He got bored of defending his decisions. Decisions that affect millions of users and can be easily tested with Google infrastructure. While he is probably right that the debating of such low level choices seems counter productive, every tenth of a percent of users count for the bottom line and with the tools they have at their disposal there is no reason not to validate.


Fair play to him. Google was created with enthusiasm and thinking outside the box. Now that it's reached "the" threshold in terms of business size, it can't operate that way anymore without scaring shareholders etc so it needs the boring reliability of data.

As a statistician, I do not recommend relying on data at all because there are too many people like me producing figures the way we want :)

Chris Miner

Data isn't collected in a vacuum. Data oriented people are driven by the same whims they claim strike and cloud the good sense of designers and artistic types. What data to collect? How to analyze it? What correlations to check? The only difference is that designers and artistic types aren't self righteous, hypocritical, and narrow minded enough to claim otherwise.

Bernd B

Another impressive description of Google fine-tuning design according to traffic data:


The problem is that data can be very myopic. Yes, maybe a green button attracts more clicks than the blue button chosen by a designer — this data is very specific and testable.

However, many other things are not: how does the green button affect a user's perception of the entire page? the brand? of green buttons? Which is more important to Google? Having more people click on a button or maintaining a consistent, positive consumer experience?

Data gives designers the power to make better decisions — there is no decision inherent in data. Google's culture apparently ignores this important principle, and I completely understand why a designer would leave in frustration.


Maybe this guy shouldn't be a staff employee anywhere, but work on a project basis as an independent contractor solely on new projects (where you determine how users will come to new major functions and how the elements in a function will appear), not established products in their maintenance phase (where you obsess over a 2 pixel difference in the next release because that's all there is to obsess over). Friend, if you were going to go corporate in Palo Alto, you should have joined IDEO and not Google.


Sounds less like an over-reliance on data as the primary problem here and more perhaps on an instance of Parkinson's Law of Triviality.

"Prove it" when it comes to selecting which is better of 3-5 pixel widths seems like a response from someone who just wants to be obstinate.

I think the real solution is not to change the engineering culture, but to set a clear chain of command, so at the end of the day, someone has the authority to say "it's 5 pixels cause that's the way I likes it, we've got bigger fish to fry, and I'm making an executive decision."

One of the best ways I've found to avoid this sort of nonsense is to avoid meeting creep- where everyone and their uncle's manager shows up for design meetings. You can't get any decisions in such a meeting. Keep it pared down to 2-3 principle people and you're much less likely to get people debating the color of the bike shed because they like the sound of their own voice.