How can I be a Data-Driven Leader?
13 Steps To Data-Driven Success
Step 1. What is Data-Driven Leadership?
This step is captured in the article I published yesterday. It is a quick read to feel free to give a shot before you continue OR if you already know what Data-Driven Leadership is, feel free to read on.
Step 2. Do you know your business… quantitatively?
As I write this step, I recognize that it deserves an article of its own (now available here) but wanting this article to deliver as well, let me summarize. You need to know two things:
- What does my business do?
- What are we trying to do?
The first question should be answered by statements like $XX per day in sales/revenue, attract/acquire XX new customers, or some group of similar metrics. A small handful will suffice.
The second is about growth. If you are not growing, you are dying. You are either growing margins or customer or sales or all three. Maybe the company goal is to lower cost per acquisition — that will grow margin, as with many other cost cutting concepts.
Wait — what if I am an HR leader and my job is to keep happy employees? Good question. Why do you need ‘employees’ and why do you need to ‘keep’ them? Develop metrics like sales/employee, revenue/employee, and cost/employee. This is why rates and ratios exist — to tie secondary functions to primary measures.
Step 3. Use the metrics you identified to make decisions.
It is incredibly difficult but remarkably easy. Judge everything against the metrics you just identified. It is easy because it is just that simple. And yet, if it weren’t difficult, everyone would do it. You just need to recognize that this is not optional. Every decision you make that does not impact those numbers is a wasted one.
But wait… there is more to this than metrics and discipline…
Step 4. Instrument Everything
We live in a digital world. If it is not captured digitally, it might as well not exist. Your organization needs instrumentation — on the website, in the call center, at the register, and in the warehouse. Do not build it unless it documents every customer contact point.
If you show the customer an image — log it. If they give you information — log it. If you ask them to click a button, to create an account, to provide an email address, download software, register a product, download an app… whatever you do OR they do — log it!
Note — if that requires people with clipboards and pencils, my condolences — but log it!
Step 5. Store it
We live in a world where each year we invent new words to describe the size of storage. B became MB then GB then TB… and it just keeps going. You can get that storage for pennies — so store everything!
There should be no compromise here. BUT, not everything needs to be at your finger tips. At most 13–15 months of data is critical in detail. The rest can be stored in archive or on tape or in aggregate or whatever IT claims is most cost effective.
Step 6A. Organize it
This step is critical. You won’t know that 4 & 5 have even been accomplished without this. Organization is a series of articles. It is likely something you should consider out-sourcing or at least get expert oversight. Also know — how IT wants it organized, the business needs it organized, and analytics will want it organized is different. If whoever you task with this is not clear on that, give the job to someone else. You want a Production (or Operations) database, an analytic database, and a BI database (or mart).
Step 6B. Understand it
This step arguably needs to happen before (or in parallel) with the last. Typically the Prod or Ops DB is well understood by the developers, engineers, and dbas. The next two layers require more work.
Too much emphasis is put on ‘Actionable’ analysis. As a Data-Driven Leader you first want your analytic support to educate you. How is the process, the product, the business working? What information are we collecting (hopefully, all it… but that is more of an aspiration than a reality)? What are our data capabilities (sorry — this is an article long question, too)?
Step 7. Measure it
Daily and monthly reports do more to change the culture of a company than anything else — short of a new CEO. What is measured is acted upon. What is measured is understood. These reports are critical and sadly, often overlooked or delayed. They create the baseline to measure your decisions by — don’t skimp or shortcut here!
Step 8. Measure it some more!
Reporting is a multi-layer function. At one level, you want the measures and metrics you determined in Step 2. But these also create baselines to test and measure decisions and strategy. You don’t need fancy test & control infrastructures to recognize whether a major new program, product, or process is having an impact on your company. The reports should reflect the lift (or deterioration, if things went poorly).
Step 9. Reward it!
Create an incentive structure based on measured improvement. Even the most hardened, stubborn employee will adapt to a data-driven culture once they realize that is where the bonus money is won. Note — this is only one piece of true accountability — sorry, more articles soon but here is a starter:
Why Your Analytics Team Should Be Creating Accountability
And Many Of The Reasons They Probably Aren’t
Step 10. Talk about it
This is really part of all the steps — but be sure to communicate. Data-driven Leaders should foster feedback. It flows both ways AND is not just about data, numbers, or money. Communication is critical.
Step 11. Pat yourself on the back
If you have gotten this far, you are likely ahead of most of your competitors. Be proud. Now be more driven!
Step 12. Test it.
Once you get here, you should strongly consider investing in an infrastructure to support test and control. This is another intensive topic. Test & Control is about producing marginal and targeted wins — often under 10% and typically limited to smaller subsets of your business. These require statistics and tools.
Almost all businesses can and should do this, BUT the amount of money and complexity that they should create to do it is a ROI analysis of its own. Not every company should hire a 50 man analytics division full of PhD’s and neural networks — honestly, very few should.
Step 13. Adapt and Repeat
Congratulations! You are now head and shoulders above most of your competition. You are getting strong and measured results from most of your largest decisions. You have a culture of data-driven. Now just roll with the punches.
Things will succeed. Some will fail. Things will change. Things will break. Everything will change eventually. Just keep at it. Without a data-driven approach, all of this would be much worse — likely overwhelming. Recognize that the stress of maintaining data-driven discipline is minor in comparison to blind-management or gut-instinct. Learn, grow, change, and preserver.
Thanks for reading!