Why Being Data Driven Is Nonsense.
Generating more unnecessary buzz than a Queen Bee.
Data-driven is the latest buzz word.
We must all be data driven.
Well of course. How else would you make decisions?!
The truth is companies have been data driven for decades. Just ask the accountants.
For every new revenue stream or cost reduction initiative, the CFO is likely to have the last word. If he / she believes the numbers ‘do not add up’ or ‘make sense’, then it is unlikely to happen.
This in its purest form is being data driven — using data to drive the decision.
The problem is, like Big Data or Machine Learning, Data Driven has been catapulted into the business lexicon.
Whether it be a McKinsey study, a Gartner research paper or a blog post, the chances are significant emphasis will be made towards data driven cultures.
I’m not adverse to such publications, far from it.
I regularly read the McKinsey publications, although fully appreciate these are often more theoretical than practical.
Yes, there are some companies and instances which are exhibiting the behaviours captured, but in reality, they are far fewer than they would have you believe.
Being data driven is here to stay. That much is true.
I think the more appropriate approach would be to improve data literacy.
Data literacy as described by Dataversity:
“a business state where data is used to power decision-making and other related activities efficiently, in real time.”
In much the same way we learn about verbal reasoning at school, why should data literacy not be taught too?
It is presumed we should know what to do with data when it is presented to us. The shocking truth is most don’t.
Being able to use data in an appropriate manner to create a compelling and coherent argument to a hypothesis is the absolute baseline. This, however, is often lacking.
Companies need to acknowledge where the weaknesses lie.
And it’s not with the data.
That said, what happens if you do become data driven but your data is ‘bad’?
It’s not rocket science — you will make ‘bad’ decisions.
That’s why being data driven, is not the answer. You should aspire to being data-enabled.
That means having the necessary data needed for a decision, but with the correct level of skill and understanding to know how to use it appropriately.
This relies on the person looking at the data.
And that’s why I’m a contrarian.
I believe that people are more important than data.
I know that sounds oxymoronic as a data leader, but it’s true.
No matter how good your data is, without the right people in the right place doing the right things, your company will never become data driven.
And that’s why I think the Private Equity industry is in a data echo chamber.
Everyone is focusing on the wrong thing.
Analytics, Self-Service, Machine Learning — take your pick. I bet each company operating in the PE industry is focusing on one of those initiatives.
In much the same way, 95% of machine learning projects never make it out of the ‘Proof of Concept’ (PoC) stage, the spotlight is being shined in the wrong place.
Companies need to focus on the lowest level of the data pyramid: basic data management.
Without being able to call on the data you want, when you want it, at a satisfactory quality, the rest is superfluous.
It’s boring but the bedrock of your business.
Even simple analysis such as the interrogation of a single dataset, is hamstrung if your data quality is bad, for example.
How do you expect to do analytics or leverage advanced analytics if you can’t get the basics right?!
The brutal truth is it’s a people problem.
People across businesses don’t understand their responsibilities in being a good corporate data citizen.
It’s not about off-loading responsibility to the Chief Data Office or a federated data team, it’s about each and every person in the company, doing the right thing.
I can’t tell you how much time my team have wasted over the years correcting data errors.
Just because someone can’t follow simple instructions and put the figures in the CRM (or source system) correctly, it means we spend an inordinate number of hours resolving this issue.
If you put a date in a name field or choose the first drop down because you’re too lazy to do it properly, don’t be surprised if the Chief Data Officer, calls this out publicly.
No technology will ever solve all our problems — much to the chagrin of every software salesperson.
We need to educate.
And this is where I use my Rubik’s cube analogy:
“Everything we do has an effect on others whether you realise it or not.”
Let’s say, I’m focused on turning all of my side blue. That creates clarity for the teams: we know what we are trying to achieve.
The only problem is that every time I move one of the squares to be blue on my side, it will affect the other sides.
Some of the effects I may know, but others I may not.
And therein lies the problem.
I am undoing or making worse a problem I am not aware of.
That’s why every part of the business needs to acknowledge that what they do has an effect, whether they appreciate it or not.
The incorrect name put in a source system by itself is not a big deal.
But if the data warehouse can’t distinguish it and subsequently causes problems in the transformation and loading processes, this IS a problem.
If the roles were reversed and I incorrectly put in a figure the sales team used, which was for an old client or threatened the closing of a big deal, I reckon they would lose their sh*t.
This is the same problem.
Perhaps if data leaders lose their sh*t more often, the sales teams or other parts of the business may sit up and take more notice.
The equilibrium is very much out of balance (but that is a conversation for another day).
To make serious gains with data in an organisation, understanding the people dynamics is the first step.
- Do your people know what their data responsibilities are?
- Do we know what the data is trying to help with? i.e. business value
- Do we have a culture of learning or just following what has been done before?
If you answer no to any of these questions, there is some serious work to do, because no amount of technology is going to resolve them for you.
Data and culture are so tightly interwoven but often never acknowledged.
That’s why every data initiative is really all about the people.
As Ben Horrowitz says, culture is what people do when no one is looking.
And when no-one is looking in your business, how your people act will determine if you become data-enabled or not.
Culture is one of those terms which gets bandied around; easy to say but hard to do.
When the new CEO took over at Unilever, he surmised that it would a year to change the culture for every level of management.
At the time, there were up to 10 levels, which meant it would take 10 years (!) to change the culture. That was simply not acceptable, so he set about reducing this to 5 over his 1st 12 months.
It took time but Paul turned Unilever from a staid slow moving FMCG business into an innovative, dynamic and forward-thinking business. This was driven from embedding the right culture.
So if you want to unleash the power of your data, the question is, do you have the right people & culture…?
