“It’s not me it’s the data”
“This report is different to that report, I’m going to quote the bigger sales figure”
“But I don’t know how we are performing, the data is wrong”
“But I can’t do anything about our poor performance, the data is wrong”
These are just some of the excuses I hear on a regular basis. Management can not ‘manage’ without data, but what they also do is use data as a smoke screen. It is very easy to create fog around performance issues by blaming the data. It’s often the first line of defence.
Thanks to the great corporate blame culture data has a tendency to come back and bite you on the ass!
So what can we do about it?
Get on the ‘front-foot’ with your data consumers
The most obvious initial task is to make sure that your data is squeaky clean, and without reproach. Be 100% confident in what you provide, and if required, be able to explain away any discrepancies between data entities.
How do we get to the point where we trust our data?
There are a couple of key points that I make to my teams on data provision.
- We must ensure that we provide data within a clearly defined context and definition, nicely wrapped up for the relevant data consumers.
- We must have the data governance processes in place that ensure that our data is as clean as possible and that quality is monitored constantly.
- Any discrepancies in reports must be captured and explained away through context or corrected.
- Get the data consumers to agree that the data is correct, get their buy-in.
- We must be 100% confident in our data, and ready to defend it at all times, don’t show any weakness when confronted about it.
Trust the data you provide!
If you trust your data, you will be able to get on the front foot with your data consumers. Being on the front foot is the best form of defence in data provision. It allows you to clear the fog that the consumers spout about data, and the more often you do that, they more the consumers start to trust your data, or at the very least, stop using it as a smoke screen to bad performance.
Have trust in the data you provide, don’t let it bite you on the ass!
Excellent post (as always) Charles,
The image of bad data biting you on the ass brings to my mind the image of the full body padded suit worn by guard dog trainers.
The trainers wear these suits to protect themselves while training the guard dogs to bite them.
Providing poor quality data to your data consumers in the past has trained them to bite you. They may even imagine you wearing the padded suit when they see you walking towards them with the latest report.
Providing high quality data not only encourages better behavior, but (no pun intended) also enables better business decisions.
Cheers,
Jim
Great post Charles!
One huge challenge I’ve faced is even when the data a team provides is smelling like roses, it is then passed between enough spreadsheets and mixed with other data of shall we say, “less fresh a smell” that the end result is poor quality.
But because the core numbers came from the data warehouse, it gets tainted through association. Unfair, but it happens.
I’ve found part of the key is to understand why all those spreadsheets are needed- usually its because you’re missing subject areas or reports in the main system, and users are having to “do it themselves”.
But I haven’t seen an organisation yet that managed to supply everything needed all the time- there just always seems to be more demand than supply…
Thanks Jim and James for your comments.
It is a constant task to ensure / protect your data from being tainted by constant rework and misinterpretation.
It’s time to drop the training suit and get out there and defend your data, but it’s best you only drop the training suit, you don’t want to be defending your data in your birthday suit, you need some protection!
Charles, way to go. I thought ”ass” was a bad word, but now I see it’s actually very suitable in a business vocabulary.
Hey Charles
Good post. Can I play a little devil’s advocate?
We must be 100% confident in our data, and ready to defend it at all times, don’t show any weakness when confronted about it.
I’ve never worked with or for any organization at this point. I’ve seen “general comfort” and “complete distrust.” Maybe I’m just unlucky?
Henrik, with some of the words I often hear in a ‘business context’ ass is a good word
Phil, you’re not unlucky, I’m just pushing the bar higher with the 100% confidence statement, gotta aim high!
Data is trusted through use after use, record by record, building the client trust that the data is structured, verified and maintained providing a single truth. Trust is achived through MDM processes including Data Governance and the broad category of Data Cleansing, each record processed in a standard method, import, referencing, classifying, matching, de-duplicating, verification, translation and delivery of reporting accuracy / use.
Good post Charles.
On a slightly different note, I’ve lost count of the number of times I’ve heard the disclaimer “you can have the data, but it’s not great quality”.
As if this disclaimer protects the provider from the the proverbial arse-biting.
Surely the antithesis of trust.
Love the graphic Charles, very nice post.
This posting really illustrates something so fundamental that it’s sometimes scary that we even have to talk about it: Trusting the data we provide is of utmost importance! With that said however there are so many people in organizations who just don’t get it. These folks want to “knock off a report” or “close out the project” or “get out of the office on time”. To those of you who are just knocking off reports and not 100% confident in the results, I’d recommend a different line of work because sooner or later a senior leader at your organization is going to get very grumpy because of the data you’ve been providing. These senior leaders won’t remember that you got the report “done on time”, they will remember how good or bad it was.
Trust, such a simple little word and like many four letter words comes with so much baggage…
Best…Rich
P.S. This post ties very nicely with Jim Harris’ posting “The Circle of Quality” that sometimes I think you and Jim are “in kahoots” together when writing them: http://www.ocdqblog.com/home/the-circle-of-quality.html
Thanks Jackie, Phil and Rich for your great comments and contribution to the post.
And Rich, you are right, this does tie in well with Jim’s post, but it is pure coincidence, I can’t even use the term ‘great minds think alike’ here, cause sometimes I’m not sure how Jim’s mind works like it does
Hi Charles,
Enjoyable, informative and thought provoking – Well done.
Ken
I thought it was “bum” across the pond?
I think Phil has a good point when he states that data perception tends to be somewhere in the middle rather than at the extremes. Data quality is an abstract metric, especially when it is not extreme. This presents a challenge when defending the state of data. If we are practical, data will never be 100% defensible. The nature of data creation in today’s reality is too voluminous to be “on top” or “out in front” of all of it. That is not to say that an attempted to address the smoke screens should not be undertaken, but rather to say that the “smoke screen” will always be there.
I think this post is another, in a long list, of reasons why MDM and data governance is a permenant part of Information Management. Thanks for the colorful illustration, Charles!
William and Ken, thanks for your comments and kind words.
William you make a good point re: data not being 100% defensible, and that the smoke screen will always be there. What I do however ask for from my teams is that they strive for 100%, not getting to 100% quality is not a failure, the failure comes when you stop striving.
Cheers
Charles
Like the nice roundup at the end there Charles, clear, concise actions every company should be focused on.