Getting Results Faster with Lean People Analytics

A Major Success Story

Over the last 10 years, the field of people analytics has grown dramatically. In time I have worked in the field people analytics has gone from something invested in only at just a few large companies with data-insight-sensitive management teams to something that has spread out across the world at a rate comparable to some of the fastest-growing products and companies. Today most San Francisco Bay Area technology companies above 1500 employees have at least one person working on it. Companies in all industries, sizes, and geographies are working on it or talking about it.

Some of the energy for people analytics was the result of the well-promoted success of the work at Google as represented by beloved former HR head Laszlo Bock. Numerous articles about Google’s data-driven HR practices were placed in business journals that CEOs and business partnership minded HR professionals read. No List is necessary, you can “Google it”: search for (Google + People Analytics + New York Times), (Google + People Analytics + Washington Post) to see what I'm talking about.

Where Laszlo Bock and team left off, conference promoters around the world picked up. Some examples are:

·     The well-run Barry Swales promoted, David Green chaired Tucana Global People Analytics Conference.

·     The beloved Al Adamsen’s People Analytics and the Future of Work has been around a long time and continues to grow every year.

·     The Julian Brookes promoted HR Metrics & Analytics Summit is now its 20th conference. I don’t know if they are doing 10 per quarter or what?

·     The academic and professional milieu of the annual Wharton People Analytics Conference has probably the most rigorous speaker panel.

·     On occasion, I have even paid my own way to get on an airplane and fly to exotic places like Sydney Australia or New York City to attend or speak for the Mihaly Nagy promoted Stamford Global Workforce Analytics Series.

·     There is also the Human Capital Analytics & Workforce Planning Event and the innovation enterprise HR & Workforce Analytics Summit

·     Presentations on people analytics are increasingly included in broader HR conferences, HR Tech conferences and even has appeared sporadically at general analytics conferences like the Josh Berger promoted Minne Analytics…

There has been a virtuous cycle of energy from the increasing population of people analytics professionals, learning from each other at these conferences. I’d go to more of these conferences, but I am either mired in an interesting project I can’t let go or I am a starving writer. Rarely am I both at the same time. 

Stalling or Failing, You Pick

In the last several years, the field’s most vocal thought leaders have been ruminating about why so many of the people analytics efforts that started with great promise but seem to have “stalled*”.

*We tend to avoid using the word “fail” or “problem” in HR circles, unlike Startup circles where failing is a badge that indicates you are ready for the next investor and difficult problems are what the investors want you to start out talking about.

I wouldn’t say many people think the sky is falling on people analytics but there is a concern. There is a concern when you see smart and well-liked people (and sometimes entire teams) dispersing to other places quietly in the night. Some people make great presentations about a great project or their organization's journey and then we don’t hear anything from them again. What happened?

This concept of failing or “stalling” -- or a general malaise of what next -- has been written into reports by Josh Bersin, an HR Tech research and better thought leader than myself. Others have speculated. At various conferences, I have contributed my share of thoughts with starting, stalling, stepping away and beginning again. When I see Josh Bersin on stage I want to rip off a jacket to reveal a tight shirt with a large S on it, push Josh Bersin to the side and say, “Stand back, let me do it, nobody has failed more than me”. Looking into the distance. (I like to Troll Josh but Josh, you are doing fine. Keep on keeping on! Do you, it seems to be working. :-))

It is o.k. to be inspired and learn by examples, but examples don’t always apply well from one place to another. It is also another thing to understand how to create success in another environment. What I believe people analytics needs is a rigorous repeatable process that can be learned, remembered and applied by many different types of people, too many different types of companies, in many different contexts. In particular, the process needs to help determine what focus will produce the most value in a specific time and place. The process also needs to be able to tell you how to get from where you are to where you need to be to be, which is not universal and will change on you as you go. Is this possible?

The Epiphany of Lean Startup

Over the last several years I have been influenced a lot by the Lean Startup movement – spawned by guys like Steve Blank, Eric Ries, and Ash Maurya. My favorite of the bunch is Ash whom I met in Austin Texas when I tried to start a company, or two, or three in people analytics. (Sorry Mike, Corey, Tim, Rick, Alvan, Adam, & Rich. I keep the candle burning for you. Last names not provided to protect the identity of the innocent. ) Ash invented a “Lean Canvas”, which is a one-page document designed to replace a business plan. It makes the old ways we used to do business plans look like the original typewriter. He didn’t stop there. He has great workshops, two great books and something called the “Lean Stack” (that signed up 70,000 new users last month or something ungodly number like that.) All this and you can still meet him face to face, get some advice and have coffee with him. When Ash swigs down the last of his coffee, outstretches his hand, looks in your eyes and says “You can do this” you get a pretty cool feeling inside. “I’m going to do this.”

What is profound about Lean Startup is how these clever people figured out how to apply a Japanese manufacturing practice to starting a company. It is hard to imagine two scenarios more different. I can imagine a factory is process-driven and standardizable. Starting a company seems like one of the least process-driven and standardizable things possible. What in God’s earth does an ice cream shop have in common with an outer space mining company? We don’t even have to be that dramatic, what do any two businesses have in common? In any case, they saw a repeatable pattern and they figured it out how to get that into the world. They developed an entire language full of new phrases like “Minimum Viable Product, “Problem Solution Fit” and “Product Market Fit”. Their success is only a result of how these ideas resonated with people who actually been building companies to something from nothing and those who failed along the way. You know you have nailed it when you give a speech and ? of the audience raises their hand to indicate, yes, that’s me, that’s how I lost 3 years of my life, my wife and my life fortune. In a few years, Lean Startup has become one of the most influential and important business ideas of our time. I guess it resonated.

The lean startup method proposes a counterintuitive process for starting a company that turns out to work much better than luck, brute force, piles of money and myths. The method is particularly compelling if you have personally experienced the pain of failure. This does not suggest having this observation of what went wrong immediately makes you a master at applying lean. Of course, there are a lot of different paths to starting a successful company and having a lot of money can help! Practice usually helps, and practice trumps theory every single day. I just added the “every single day part”, the rest was something Ash Maurya said first.

I know you didn’t start this reading this article to learn about starting a company or my difficulties to do so, so this is where I will tie back in. Founding a people analytics team is much like starting a company. Especially in relation to anything else that people do in Human Resources, which we have to admit was invented a long time ago and is getting kind of tired. I am not picking on HR people – I have a master’s degree in it. I have never handed a resume to anyone without HR on it. I try to be an example of good HR and I work with people like that.

The new methods that materialize from applying lean to people analytics is more of a book than a blog post, but I’ll just rattle a few thoughts to get us started and moving along. Books pending. Pending. Pending... Sigh.

How Starting People Analytics is Like Starting a Company

More often than not when you startup people analytics at your company you are a founder. By this I mean that often nobody you are working for or with has ever done what you are going to do before. In fact, you may be working on something nobody else in the world has ever worked on before. Let that sink in. You have to start something from nothing with no manual. It requires personal investment and energy. It requires innovation. It requires a big lift. It requires getting multiple things right at the same time. It requires learning from repeated failures before you have success and accepting failure as part of the process of learning.

Sometimes you produce things you think people will value and they are not valued by other people the way you thought they would be. Sometimes you start with back slaps for success and then people just stop using your stuff. Sometimes you invest a lot of money and work really hard to get this thing going and then a few years of new technology makes everything you have done obsolete. Sometimes you have a clear vision and never have the resources you need to implement it. Sometimes you have initial success but the people piling in the door outstrip your ability to meet the needs.

I want to key in on that last sentence: “sometimes you have initial success but the people piling in the door outstrip your ability to meet the needs”. Generally, we think of this as a “good problem". However, if you are looking at a line of unamused customers with crying children in tow it is still a problem you have solve pretty quickly.

How Starting a People Analytics Function is Not Like Starting a Company

One important difference between starting a company and starting a people analytics function is that if a startup begins to succeed it generates revenue that funds its expansion.

People analytics functions may begin with some small funding (or maybe just your pay and back sweat) but the work does not generate any revenue to fund expansion. You are trying to meet the needs of 10 times your number in a diverse “customer” base: HR Business Partners, Talent Acquisition leads, Total Rewards leads, Talent Management leads, Employee Relations leads, HR Operations leads, Diversity leads, CEOs, all the Division leads, etc.. These people all have different priorities, different data and different problems to work on, not to mention different personalities and proclivities.

Sometimes, I wonder if these are even "customers" at all. These “customers” do not pay you. Maybe they are something else: Fake Customers!

A Best-Case Scenario

There is a special place in hell for the people analytics pro that makes only one HRBP or functional lead look good. Another HRBPs stomps her foot down thinking, “What has this hot shot done for ME lately?!” Recruiting lead comes out of the meeting and says, “Wait a minute, when it comes to data I have nothing!...” A tactful Compensation lead stops you in the hallway and starts up, “Wouldn’t it be cool if …” This is just by 9 am.

As you sit down at your desk you find out the CHRO has sent out a red alert to you and the entire HR leadership team because emails were flying in the night. It turns out an important person left the company and there is a fight among the management team about what is wrong with the culture. She needs an intelligent reply ASAP. Two days ago nobody cared about employee attrition and culture. Today they do. These problems are not simple and you’re not ready. There has been no advance work to define culture and without getting into the details you are not comfortable there is a reply that can stand up to even a modest amount of critical thought.

Keep in mind this analysis is going back in front of the people who are trained in critical thought and carrying daggers. Two months ago they disemboweled an executive you thought was their friend and sent him to die elsewhere. With or without health insurance. They don’t care, they have a business to run. Get ready to put many iterations into your analysis of data they didn’t let you collect before they had their question – you just got invited to the world championships of critical thinking and it is go-time.

Meanwhile, all other projects just stopped and you aren’t going to complete all your previously defined objectives for the quarter. Other people you depend on to sing your praises are standing around with arms crossed, rightfully upset. You promised them something a month ago – it is on their objectives - and you didn’t deliver.

By the way, nobody in this story is the bad guy. I mean that! It is just an unfortunate story of a “successful” people analytics function.

As bad as the hellish scenario I just described sounds that that maybe your best-case scenario!

A Worse Scenario

A worse scenario is where you are hired to produce reports that nobody uses. You produce them month by month, week by week, day by day and send them off to people and nobody uses them. No tumult, it is just check-in, check out. God forbid you try to stop sending the report because … Wait, I don’t know what happens. Nobody knows. Let’s just change the subject. 

Maybe you have managed to stand up automated self-service reporting. Have you ever checked the logs to see how much traffic the reports are getting? How are you doing? I guarantee you the lunch menu is getting more traffic. Believe me - Taco Tuesday does not feel threatened by you.

An analyst tries to run an idea for a profound change up the flag pole and everybody looks at them like they are from another planet. Hey buddy, slow down. We can talk about the newfound importance data to HR, and we will hire a few people to take care of that reporting thing for us, but we are not going to change the way we do what we do. We don’t have time for that we have to get this thing done!

“Nobody makes me bleed my own blood” – Ben Stiller (Dodge Ball) Hmmm.

 All the decisions small and big are made on an entirely separate time, place and process than the reporting and analytics. The two never will connect. You are not going to hit this one out of the park.

“They call it Moneyball and I think they just bought a ticket onto the Titanic.” - Some Radio Critic (Moneyball)

Again, no bad guys. It actually really hard to blame someone; it is just the nature of people in groups. The lines that divide us are hard to see and most people want to get into work and out without conflict. They drag their feet to work, they moan, they get headaches, and may take longer lunches, but they are not going to fight at work.

I usually fight, and I usually lose. I can do that because I was raised by wolves, and I am confident I can occasionally take down a sick antelope when I need to. Gnawing on that dead animal's jawbone, I point it at you and garble out of the side of my mouth while still chewing, “You should make better choices than me”. 

However, I love my work and my life. Can everyone say that?

Value is What Binds Lean to People Analytics (and Everything)

The profound epiphany is that while both the situations I have described above are different, both are troubled by value and connected by value.

In the best-case scenario, you have a limited amount of initial success, followed by a high risk of failure because you can’t charge for your services to fund the expansion of your team to meet demand. In other words, people apparently find value in the work you produce, but you can’t extract that value and reinvest it in your business as any good entrepreneur would. So you get pinned down to choose where you want to deliver limited value. Your problem is that the definition of value is in the eyes of the beholder.

In the worse scenario, you didn’t get value right from the beginning. You produced things that nobody wanted. Maybe they asked for them! That doesn’t change the fact that they don’t use them. In an entrepreneur’s world, you would run out of money and shut the doors. You may see that as bad - I see that as an opportunity to learn and produce something better. In the alternative world, HR lives in they just keep paying you and everybody is miserable. Don’t worry - eventually, the business will close or they will fire you.

In both the best and worse scenarios, we have issues with focus. In the best-case scenario, you need to focus because you cannot do everything and succeed. You need to establish priorities for what will deliver the most value. If you don’t have a rigorous defensible focus argument you are going to get chewed up and spit out in office politics.

In the worse scenario, you need focus to establish and communicate what will deliver value because you don’t want the company’s money and your life effort on work nobody really wants in the end.

By the way, in both business and larger life it turns out that when you take a broad perspective you don’t actually have that much time to figure out this value problem. What financial resources you have to invest in whatever makes you happy will be limited by the wisdom of the earlier actions you take. You want to get on the right path as quickly as possible or you will suffer a calamity. You might get away check-in, check out for a time but eventually, you will have to justify the value of your time.

There are a lot of debilitating cynical views in the world, but I will never let go of the view that the piper always gets paid one way or another in the end. Those who don’t pay the piper maybe sometimes start out looking great, but this disguises the failure that awaits if they don’t see the wisdom of what paying the piper would produce until after it is too late.

Speaking of this, seen this timeline lately? The rise and fall of Elizabeth Holmes

A Basic Definition of Lean

 Taiichi Ohno, the founder of the Toyota production system, described Toyota’s lean methods “looking at the timeline from the moment the customer gives us an order to the point when we collect the cash. And we are reducing that timeline by removing the non-value-added wastes.”

Another way to view lean is by analyzing capacity, the amount of product that can be produced in a given span of time. To Taiichi Ohno, the capacity equation is simple:

Present capacity to produce value = work + waste.

The lean way to increase capacity is to eliminate waste. Work is anything that that adds value for the customer; waste is anything that doesn’t. The goal is zero waste and 100% value-producing work. Rooting out waste increases capacity for work that adds value. We may never get there totally but the goal and the journey give us a fresh way of looking at the work that leads us to better results.

The absolute elimination of waste became the objective of the Toyota Production System, and it drove Toyota to become the largest and most profitable automobile manufacturer in the world in a time when American manufacturers kept bankrupting. You can give some of the credit for that to the Japanese and their silly ideas. To the extent that we have any sort of manufacturing success in the U.S., it is usually achieved by copying the manufacturing techniques worked out by the Japanese.

Fool me once, fool me twice… I’ll warn you now. If you don’t accept my ideas for lean people analytics in the U.S. I’m going straight to the Japanese. Does anyone speak Japanese? Any Interpreters in the building?

In their book Lean Thinking, James P. Womack and Daniel T. Jones define the lean way as a process, boiling down lean to five steps or phases:

1.    Precisely specify what customers value.

2.    Identify the value stream for each product.

3.    Make value flow without interruptions.

4.    Let the customer pull value from the producer.

5.    Pursue perfection.

Will the Real Customer Please Stand Up?

Know that the customer is not the requestor or recipient of reports or analysis. The customer is the customer: the person who buys products or services from your company with money. When we talk about customer value, what we are talking about? We are talking about the need to connect your analysis to that customer. Fundamentally any other definition of customer would not be lean.

Now, if you get into it there is some acceptable divergence in complex situations. A great example of this can be provided by advertising-driven social media companies like Facebook, whose largest user base doesn’t pay. We increasingly give companies like Facebook our time (and our data) because we derive value – warm interactions with our connections, if not intellectually stimulating. Facebook makes its money on advertising – the companies who buy advertising are their real customers. They deliver value to their real customers by producing us over long periods of time and in larger numbers. They produce us by providing us something we want for free. As we are all learning, it is not actually free though. VCs get around this problem by estimating the probable monetary value of a user base based on its size and engagement. 

In complex situations, you are permitted to think about a customer as trading time in lieu of money as a measure of value. Time = engagement.

A lot of things boil down to time. You can play with this idea of customers and value and maybe think about how to measure the value people find in your people analytics services by how much time they give you. If they are saying, “Oh, I’m sorry, no time!”; they are probably really saying, “this has low relative value to me!”

Lean People Analytics

What is Lean People Analytics?

Lean People Analytics it is not about being cheap. Lean People Analytics is about focus and speed. It is about having an original vision and then testing it through tightly controlled experiments, which turn out to be efficient, effective and FAST.

Lean People Analytics is not about asking employees what they want through surveys or focus groups. Lean People Analytics is about testing an original vision about how the people management practices you apply will influence people and measuring what people do as a result of that.

Lean People Analytics is not cherry-picking a few analysis techniques that seem awe-inspiring or hiking up the analytical maturity curve. When you talk to people about starting People Analytics you hear a lot about technology infrastructure, the right composition of teams, data quality & ownership, the analytics maturity curve journey, getting out and finding out what executives want from you, and so on. That is not what Lean People Analytics is about. Lean People Analytics is about focus. It is about focusing on the right actions and changes.

At any given moment there are only a few people related actions and changes a company can take that matter. However, when you start to look at the way people contribute to organizations you find it is inherently chaotic and complex. Lean People Analytics is about finding those actions that will matter to produce change and ignoring the rest.

Human Resource teams that succeed are not the ones that perform the most complex analysis, or set up the best technology stack first, or implements a perfect checklist of best practices. Human Resources teams that succeed are those that manage to take the information they receive and make or influence the changes that are necessary to demonstrate measurable business value before the company has spent off all its allocable time and resources for people on wasted efforts, fails outright, and/or tires of YOU.

Here is are some beginning thoughts about how some of the principles of Lean can be applied to People Analytics:

Step 1 - Precisely Specify Value

Dig deeper. Think about your company’s real customers. Why do they give their time or money to your company versus some competitors? What are you good at that they are not? How do you increase this value?

Those are great questions and they sound like something out of an MBA textbook, but what you generally don’t find in an MBA textbook is that that value is ultimately created by people, nor do you find the mechanisms through which this transfer of value occurs.

Here is an example. U.S. Merck, a pharmaceutical company that develops lifesaving drugs. People pay for those drugs because they are better than the alternative. Patents protect U.S. Merck’s ability to charge for drugs. Patents expire 20 years from the day filed. On average, it takes at least ten years for a new medicine to complete the journey from initial discovery to the marketplace, with clinical trials alone taking six to seven years on average. The average cost to research and develop each successful drug is in the billions. Most drugs fail to meet effectiveness and safety standards. If they succeed, by the time you get it to market you only have a few years left before you have competitors who have found a way to copy the drug without breaking your patents. In any case, you now have less than 10 years before the patent will expire and anyone can use the patents you filed. This is when you see the lower-priced generic alternatives. They can do this because their costs do not include any of the expensive R&D and initial awareness campaigning. Everyone buys into the system because it is a fair trade in the end – that does not say it is easy.

Why has U.S. Merck survived as a company over 100 years under these operating conditions? Merck is in the science business - more specifically the scientist business. Merck does not produce value by manufacturing drugs. Merck’s value is created in making novel scientific discoveries that generate methods that are patentable, some of which will provide a path for the profitable manufacture of drugs. Patentable drug discovery and manufacturing methods are produced by scientists. Scientists are people. Merck can’t stop producing new science or they die. Merck has survived over a hundred years by being the most prolific research institution in the history of the world*, and a lot of that is as a result of how they think about people. How they hire people and the environments they create for people to work in. If you know them, Merck is not perfect, but they are a lovable and fierce tribe.

This is not an exaggerated claim – you can check this assertion with the U.S. Patent Office.

Google has a similar people model but applied to a different problem space: how human beings interact with information. Google’s value is produced in novel innovation, ideally innovation that is defensible in some way. If Google does not get that right then in a short amount of time they will end. They know this so they make bets on intellectual property that is produced and developed by people and the environments they create for these people to work in. Alphabet? Cheeky. 

Some companies are Murky (Merck), other companies are Chic’y (Apple), Google is cheeky, all of them know a lot about how to acquire and equip people to produce the product features that people who buy things want to pay money for: which represents value. How it works is going to be a little different for every company and gets into the nuances of who the real customers are and how they define value.

Formulaic “Best Practice” based HR is a pretty blunt and dumb tool, and contrary to popular belief “Best Practices” do not deliver much real business value. Real business value is produced when you fully understand the value chain between employees and customers and bake that into how you run HR. I can’t tell you how to do that from the outside – you have to get into the details of it.

Step 2 – Find the Value Stream

 Once you understand or have a hypothesis of customer value, map out how employees contribute to that value chain as it is being created. You might recall the Sears Employee Customer Value chain from ten or 20 years ago.

Keep in mind that once products are created you only have a short chain from there to the customer -- through a store or sales representative – because this is just the end of the chain. The end of the chain can be analyzed. These simplified analyses are a great place to start, however, you might start squeezing there and after a short time find no more juice to squeeze. The real customer value is produced upstream from the sales team. How did you get the products and how much time do you still have to figure out the next one? Usually, you are going to find some people in that early value stream and those people are doing something. Their qualities, techniques, and performance vary, as do the organization of their work into jobs and teams. That’s your ticket to the value train. Board the value train.

You can guarantee Sears is thinking long and hard about why their employee-customer value chain stopped working. I am probably not the first to notice that for some reason smiling happy customer service representatives stopped being enough to bring people into stores. The paying customer of the products you sell in stores is really the customer of the products you sell and to the extent, these are not differentiated from anywhere else they could to get them, and you are also not the most convenient solution, then you have a real problem. Consider that sears product is really not any of the products people buy with money. Sears product is the store and customers demonstrate the value of the store to them with the time they are willing to spend there. Amazon figured that out - Amazon's product is the store and they keep improving it. Of the brick and mortar retailers that I know well, PetSmart is the best example of a company that has mastered this equation. Despite competing against the same forces that have taken the wind out of the sales of companies like Sears and Toys r Us, PetSmart has not only survived, they have thrived. How? The answer to that question is the ticket. Get on the train. I'd be happy to provide clues if you haven't picked them up already.

In recent years I have taken to drawing models – or whiteboard sketches – of connected measures that relate people to business value. I suggest you do this to clarify your thinking about what matters and share your thinking with others for expansion and feedback. Ultimately the model defines the opportunity and forms the outline of your questions. You use your analysis to test the theories and you will use the results of your analysis to refine your model. Then you use your model to tell the story to explain your findings.. I can’t say enough good things about models. It is not a side thought, it is the whole thought.

The value you produce in people analytics is not your analysis, and it is not even in the interesting insights or stories told from your reports. Ultimately the value you produce from people analytics is in the usefulness of your analysis to illuminate areas of focus in the employee value chain model that cause constructive change.

You could probably just shorten this to say, the value people analytics produces is “the change”. Anything not necessary to produce this is not value and by definition waste. If the reports do not produce change, no matter how perfectly constructed they are, they are still waste.

There is some room for complexity – maybe you need certain things ready to go but you don’t need them all the time. There are various ways to deal with that, but If nothing else, the time and resources you use to produce them need to be minimized. Also, if you say you need them for hygiene purposes or “just in case” then don’t belittle the people who produce for not producing insights or impact because it was agreed that was not the point of putting effort into those reports – it was readiness and efficiency. We call this "Scale" for short.

People analytics has it own unique value chain to produce change, which connects somewhere to the real customer as defined by your business and situation. Map out each step required to get to a rigorously tested employee value chain and common agreement about what corrective actions that implies. Don’t skip or leave out any steps. As an analyst often we define our success in terms of the complexity of the analysis. The massive blind spot of that definition is that it ignores what inputs are needed before and after that complex work, and what is required to push or pull that work. Blind spots lead to injuries and waste.

Step 3 – Create Flow

Now scour your map of people analytics value chain for waste, defined as anything that doesn’t produce value as defined carefully above. Once you find waste, plan to root it out. The goal is to create a process to make value flow without interruption. 

Make sure you don’t forget to scrutinize carefully how these things ultimately connect to real customer value. It will completely reshape what work you do and the way you do your work. If not, you probably aren’t doing Lean People Analytics - you are just doing people analytics stuff.

If you want to produce something from your work it can’t just be a series of tasks. It has to be a series of tasks connected to some objective, going from somewhere to somewhere.

Maybe you have a messy series of tasks that you do that you imagine aligning with value somehow, but it feels awkward. Maybe you think about the work and you think, “I don’t have anything that meets his definition.” Great! Stop what you are doing for a minute. Forget about all of it. This is a chance to define it as a process that you think will get you to real rigorously tested business value.

Every process needs a beginning, middle, and end. Start with the end in mind. Draw the employee-customer value diagram. Meet with people and get their take. Get their theories. Figure out where the perceived people leverage points are. Get a reaction to the range of possible people improvements in the leverage points are. Ask where the biggest risks are. You are going to learn a lot.

If you have done that you have all that you need to figure out where YOU can add value with data and experiments and define a process for YOU to get there. Map your journey. This is a map that may involve steps to confirm your research questions, define measurements, define data collection techniques, define data workflow, and define the statistical analysis you need. This is not a map of ongoing technology architecture. That comes later.

At the end of the day, however you get there, you need a defined process and then you need to take all the waste out of that process. You are not going to get anywhere if you are running around without a map. Remember the employee to customer value chain – don’t think about creating the perfect team or perfect tech infrastructure stack if you haven’t confirmed what data you need or what actions you are trying to influence yet. Missing this point leads to dead ends - it is the biggest mistake in the field.

Step 4 – Go with the Flow

In Lean manufacturing, they talk about letting customers pull your supply chain forward. You do this because don’t want excessive inventory. You also don’t want to have components stacked up because different teams are running at different speeds. You also don’t want to push investments in features or quality that nobody really wants to pay for. 

In people analytics, it is a little more difficult to apply this concept because we are further from that real customer. You may think the best thing to do is to have your internal stakeholders pull you, but this has substantial unseen risks. One problem is that the stakeholder may ask for things that it turns out they never really use. Another problem is that they may just tell you what they think you want to hear. Another problem is that they don’t pay for anything, so they may ask for more than you can deliver. Another problem is that there are a lot of stakeholders and every one of them needs or wants something different. Another problem is there are a lot of things you can do and you can rationalize any of them. Many of them seem like great ideas and have some rational value, but you can’t do them all. Doing a little of everything is a recipe for failure.

Your best option is to pull out your map of the employee to the customer value chain, however poorly it is currently conceived, and sit down with stakeholder’s person to person and try to improve it together.

Here is what your conversation should look like: “Here is the system perspective of the employee to customer value chain for our company. Is this right?”, “What is missing?”, “Where is the companies greatest risk of executing on its business strategy, and why?” “How can we do this thing better?” “Where does what you do fit in?” “What related theories about people do you have?” “What related problems do you see?” “What related questions do you want to add to the list this model already implies?” "What did I not ask you about that I should have asked you about?"

This conversation is letting the real customer pull your work by proxy through your model, not talking with fake customers about report colors and features, that they aren't going to use anyway. Maybe the conversation influences you to add a variable to your model. Maybe they suggest a simple model of what they do that that connects somewhere to your model. This is not telling you that you have something new to work on yet, but it gives you a starting point to determine this and to see how it all fits together. In any case, the reflection the conversation produces may cause your stakeholders to think about their own work differently. That could lead to change without you doing anything more. Eureka!

Your second-best option is to let your nonpaying stakeholders pull you anyway. Ash Maurya told me one day, "Don't measure pull by what people say; measure pulls by what they do."Often people tell you about features they honestly believe would be good, but when it comes time for them to buy what you built they are nowhere to be found. In our case they don’t pay with dollars – they pay with engagement. Measure engagement. You need to increase engagement.

Here is one way you can establish and measure engagement in different areas of work. Manually produce 10 different analyses and put them on a secure website where you can measure their use. At the end of 90 days stop doing anything that nobody has looked at. Prioritize your time in producing a more efficient (scalable) version of the items that are more frequently used in priority order of that magnitude of their measured use. Keep iterating until you have it.

Another way to establish and measure engagement is to schedule 24 “agile people analytics data strategy sessions” two weeks apart and invite any reasonable stakeholder. Describe these sessions as the place where you are going to collaboratively define and refine the analysis you are going to perform over the next several weeks. Those people who value your work and show up can collaboratively help determine what you work on. Those who don’t show up get less and less right to your time. We talk a lot these days about increasing agility in HR. Well, what I just described is Agile in a nutshell. If this sounds as much little like herding cats to you as it does me you may want to get an unbiased and lovable facilitator.

An additional benefit of the strategy session proposal is that if you have multiple parties showing up to these sessions you might cause people to see the world through others' eyes and find connections. The sessions also cause people to work together more and be more willing to offer their own resources to help you get things done. When push comes to shove all these people to become your advocates and defenders. This is some pretty awesome stuff.

These are just a few creative examples among many. Keep in mind you need to establish a common understanding of the concept of a customer value chain model, the problem area of focus, and prioritization rules at the outset. You can’t arbitrarily implement a prioritization methodology without getting buy-in to that first.

It is easy to get caught up in the work we do, and work begets more work. Always keep coming back to look for problems to work on that will help the company deliver more value to the real paying customers. If the connection to customer value is so abstract as to not be measurable then it is probably not worth doing at all. Sorry, that’s the truth. I can say that because I am confident most things can be measured. If you tell me it can’t be measured then you don’t yet know how to define what it is, so we either need to work together on that first or you have more thinking to do. Let's come back to this when the thought is complete.

Step 5 – Aim for Perfection

Develop a culture of continuous improvement with the goal of achieving perfect flow, which means zero waste production. This statement can apply to how you eliminate waste in people analytics, but also in how you use people analytics to identify and root out waste in HR. HR is also producing something, the question is what. A savvy people analytics leader can solve both problems at the same time! Lean is meta like an infinity loop or Russian doll.

The techniques you apply to define focus is beyond the scope of this article but imagine for the sake of argument you have already decided to work on a specific problem you know to have value to real customers. Here are some examples: 1.) how do we reduce high attrition among people working on a key future product line? or 2.) how do we create uniformly enthusiast sales representatives (without making them inhuman and creepy)? 

On either of the two examples you can use a key driver analysis, or a conjoint analysis, to identify that there are 10 things that matter, 30 that don’t, and of the 10 that matter the company is already doing great on 8. That gives you a whopping total of 2 things to conduct more analysis to determine what needs to change. Working on anything else is wasted effort. Aim for perfection overall, but just work on those 2 things that matter and need improvement. Think it through. Test one or two ideas. Run experiments. Measure change. Share findings. Measure change. Keep going until you get it right. Of the other 8 that still matter, but that you are doing fine on, just make sure they don’t drop. Of the 30 things that don’t matter at all, divest effort.

Keep measuring the outcome and the 10 things that you know matter to make sure nothing goes wrong. If the outcome slips and the 10 things are still going great, maybe something else has materialized that can influence the outcome that you haven’t seen before. This informs your need to test new theories and update your model. This defines the parameters of your next analysis and provides some clues about the path you take.

Lean People Analytics Series

“Introducing Lean People Analytics”, July 6, 2018, LinkedIn.

“The Ten Types of Waste in People Analytics”, June 19, 2018, LinkedIn.

“The Five Models of People Analytics”, July 7, 2018, LinkedIn.

“Making a Business Case for People Analytics (with the three A's of Lean People Analytics)”, July 2, 2018, LinkedIn.

“Getting Results Faster with Lean People Analytics”, June 15, 2018, LinkedIn.

More

• Find more of my writing here: Index of my writing on people analytics at PeopleAnalyst
• Connect with me on LinkedIn here: https://www.dhirubhai.net/in/michaelcwest
• Check out the People Analytics Community here: https://www.dhirubhai.net/groups/6663060
• Buy my book on Amazon here: People Analytics For Dummies, directly from the publisher (Wiley) here: People Analytics for Dummies, or from other places where books are sold.