5 Important Differences About Google People Analytics

Last evening I was asked to speak at a meetup in Austin, “People Analytics and the Future of Work”, set up by the Talent Strategy Institute and hosted by Whole Foods. Also there to speak was Robert Lanning (People Analytics at Rackspace) and Chris Butler (OneModel - a recent Techstars graduate and innovative HR tech solution of the year). I enjoyed the others talks, the audience, and the chance to get out and speak about what I love and do: People Analytics. Preparing to stand in front of other people, with all eyes fixed on me, forced me to spend time thinking about what of value I could share in a brief interchange and to organize my thoughts more than I usually would. I was fortunate it was well received. Having made this investment, I share what I came up with...

I have been lucky to work with data in HR with a number of companies under various banners. The industry has varied, where in the organization I reported, title and situation, but the central unifying question I have worked on for over 15 years is, “How can we use data to make better decisions in HR?.”

It has been a diversity of experiences. I have seen good and bad. (Sometimes I have been responsible for the bad!) 

You fail, you fail, you fail, and if you don't quit, eventually you figure it out.

The central question I want to discuss today is what is different about the places where People Analytics is going very well, versus places where it is not going as well. Given that I don’t have exposure all companies working in People Analytics I have to base this on my experience. My experience can be described most simply as Google on one side (high degree of success) with all of the others on the other side (modest degrees of success).

To say Google versus all others may be a bit dramatic - clearly we have had some success everywhere I have worked - however Google is a stand out experience that begs to be described.

If you are not familiar with the Google Storyline here some catch up on Google People Analytics.

Here are 5 important differences about Google People Analytics:

#1 People Analytics is not another thing they do in HR, it is how they do what they do in HR.

What does this really mean? This is a book so I can’t really do it justice briefly, but to help unpack this thought I will offer you the following principles:

If you can’t measure it, it probably doesn’t exist. We can debate this, but if you can’t measure it, you can’t manage it and if you can’t manage it it probably isn’t worth working on. Your opinion doesn’t matter. If I have learned anything in HR it is that opinions are like lips and assholes - we all have them, they vary. You actually don’t know what will work. We think we do, but that is just pride. There are no perfect templates for HR. For every great company with a “Best Practice” you can find a company with the same practice that is failing.

Google has a really interestingly model of Human Resources (they call HR People Operations). They deliberately hire three different types of people proportionately: A.) people with deep domain expertise in HR, B.) people with STEM expertise (science, technology, math and engineering), and C.) people who are great consultants. They put these different types of people together to work on problems and they rotate through different roles. The result is that you get people working together and you get this cross pollination of ideas. Eventually you have everyone on the same page - at least that is the goal.

An interesting thing I heard Laszlo Bock (Google’s SVP of People Ops) say one day is that one way he measures his success is by the number of HR people who are hired into other business functions at Google. Given the nature of people who Google’s hires outside HR this is an impressive goal. How do you get impressive STEM people to come work in HR AND/OR how do you get the rest of the business truly impressed about the people that came into your business through HR. Well, that’s the whole problem in a nutshell.

This is very different than what you find in HR at most other places.

#2. They have a problem focus versus a tactic focus.

 There are two ways to improve Human Resources

A. Implement what you or someone else thinks might be a good idea and hope this has the desired effect. (Call this tactics)

Or

B. First figure out WHERE the biggest current business constraint is, then proceed to understand WHAT the people constraint in that problem is. Proceed to understand HOW you can eliminate or reduce the constraint. (Call this strategy)

The tactics are only as good as the strategy they are based on.

#3. They look at each source of data as a single point of reference in a longitudinal human experiment.

Data are/is connected to what we are actually working on in HR, to our HR strategy - and really is just a measurement of either your strategy or your execution of that strategy. You are using data to address the following questions repeatedly: is our current strategy working and how can we make it better?

I don't know if others agree with me but I think "Action Planning" is ludicrous. Didn't you start the year with a strategy, what happend to that? Is this a competing strategy, an addendum or a revisal? I vote for revisal - anything else is confusing.

Data are/is connected to each other - we can answer questions better if we look across our silos and boundaries.

Data are/is connected over time. The best analysis allow you to observe change and then ultimately what the effect of those changes are. Great analysis considers time, occurs over time and takes time.

#4. They have a heavy emphasis on math and science.

This should not be a mystery - math and science are tools worked about by humanity over hundreds of years to give us increased certainty in a changing and uncertain world. We want to be right more often than we are wrong. Isn’t this what People Analytics about for HR?

The case for scientists:

Scientists are naturally curious.Scientists are trained to ask great questions.Scientists have exposure, and often appreciation for, prior research that may be applied to a problem - they know why, where and how to get it.

#5 They invest a lot in it.

 Google invests a lot in People Analytics.

Here is the bad news…

Here is the good news…

Still, you may never be able to imagine a time where you can make the same type of investment as Google. If this is you, I offer you the some practical advice.

Practical Advice

Start with a clear vision of how HR can help YOUR business win through people.

My new favorite definition of strategy : “Strategy is a carefully planned plot to murder your competition.” John Lyons - Guts - Advertising From The Inside Out. What he is trying to say is that, in the context of advertising, a good ad strategy does not fit a preconceived template or produce an incremental improvement. A good ad strategy will find that unseen piece of truth that will help you obliterate your competition. After seeing a good ad, there really is no longer question in the customer's mind what the difference is and whom they want to give their business. I offer this is how HR should think about strategy as well. I am not saying it is easy - it is a creative journey. You have to search but when you find it should seem simple and seemingly obvious. Don’t settle for mediocrity.

I think we are way too nice in HR. The word strategy comes from warfare and if we are going to operate in the realm of strategy we will need to take our mr. nice hat off. The best examples of strategy are in the history of warfare. In modern warfare it boils down to something in the nature of: first we are going to shut off their ability to see us, then we are going to shut off their access to a particular domain of the battlefield (perhaps the sky), then we are going to rain down bombs from that area, then we are going to surround them with tanks, then they are going to surrender. If that doesn’t work we will shut off the water, then they are going to surrender or die. There is nothing equivocal or vague in strategy. If so, go back to the drawing board.

Who said HR strategy has to be vague? I think it should be a carefully devised plan to dominate some area of the business battlefield. For example, “We will win/own/dominate this particular talent space, XYZ, which is central to our future business success and we will tenaciously pursue that mission at whatever cost is necessary.” This is to put your head in the right place but is still not clear. In real life these plans should be constructed with many carefully planned steps (tactics) pieced together to arrive at a clear destination. 

From strategy will come the objectives, problems and questions to formulate how to work on what matters to your business so you spend your time on what will be relevant. If you don’t have a clear vision you will end up spending your time on a little of “this and that” and never get anywhere. In a few years, your solutions will be replaced, as will you. If you do not currently see a clear vision for HR at your company I suggest you stop everything you are doing and insist that you cannot, will not, operate in those conditions. I’m not asking for negativity here, I’m asking for you to be part of the solution. Help the people you work with to find a clear vision. People Analytics is a great tool for this.

Work on a single problem at a time.

 Ideally, try to work on the most important problem first.

Treat your unsupported baseless opinions as an educated hypothesis and test them with data systematically.

You may have the a great intuitive understanding of a problem. This is great. Pull out from this a theory and then go to work on testing the most suspect (riskiest) aspects of your theory. Keep going until it is clear that the solution is a pretty good bet.

Yes, this takes more time. However, you will lose less time here then you would lose by implementing the wrong solution.

Every million (or billion) dollar problem is just a $1000 analysis too late. 

 I acknowledge it is absurd to create a business case for the use of data, without data. This is a true “Catch 22” - your argument that HR decisions should be made with data collapses on itself because you have no data to prove that. Rather than dwell on this problem, feed off the thought that you are already making decisions based on some premises, they are just poorly constructed premises. We should be able to find some fraction of funding from the HR programs to ensure that our time and money is being deployed in a way that will have maximum effect.

Regardless, I suggest you start with something small and show everyone how data has helped turn an erroneous, potentially costly, assumption on its head. People are inherently interested in this; when they see it they will ask you for more.

Q/A

In my talk, someone asked, “What are examples of assumptions turned on their head at Google?”

Here are a few examples of assumptions that Google has dismantled with data:

Mind puzzles will help us select smarter people who will be more successful at Google. (Not true)The best way for us to increase our odds of hiring high performers at google is to select people with high GPAs from Ivy League colleges. (Not true)As many people as possible should be involved to make the best hiring decision. (Actually, the answer is 5 structured interviews. Problems with bias / perspective. )Only the manager's opinion is really important to make the best hiring decision. (Actually, the answer is 5 or 6 structured interviews. Problems with bias / perspective. ) Managers should evaluate employee performance. (Actually, a committee is better. Problems with bias / perspective. )We don’t need managers at all. (Wrong)The best way to know what can help us increase retention is to ask people why they are leaving is when they are leaving. (Not true, other scientific & mathematical methods are more useful) All the common reasons why we think people leave companies (More often than not, not true) Benefits are a retention tool. (Not true, may act as an attraction tool, but difficult to tease out, and don’t confound purpose.)Benefits programs can be selected entirely on the proclivities of the management team. (No. What are we really doing here? A lot of mistakes occur this way. Mistakes in certain benefits program investments relative to other options for spend given increasing headcount and changing employee demographic profile.)

 ---------------------------------------------------------------------------------

Join the Linkedin People Analytics Community group (~700 strong)

Connect with Mike West on Linkedin

---------------------------------------------------------------------------------

People Analytics Question and Answer Series

What is People Analytics?

What is not People Analytics?

Why People Analytics?

What is the history of People Analytics?

What are the key questions of People Analytics?

What is the actual work of People Analytics?

Lean People Analytics Series:

Introducing Lean People Analytics

Getting Results Faster with Lean People Analytics

The Ten Types of Waste in People Analytics

Making a Business Case for People Analytics (with the three A's of Lean People Analytics)

Making a Business Case for People Analytics at Your Company (Get Started Guide)

Fueling People Analytics with Feelings

The Five Models of People Analytics

Talent Acquisition Analytics Series

Creating Competitive Advantage with Talent Acquisition Analytics

Measuring Employment Brand and Attraction

Measuring Candidate Experience