“Until you’re ready to look foolish, you’ll never have the possibility of being great.” — Cher

Let’s name it, claim it, and tame it. Data Shame… we all have it. And with all the pressure now to integrate AI applications into our enterprises, our unaddressed data woes are closer than ever to being exposed. The news these days is that Corporate America’s adoption of AI has been sluggish. But why? AI builds can be overly complex, change management can be difficult, and executives can get lost trying to articulate what they want AI to do. It turns out that all of these important factors have a common insidious element that drives them: Data Shame.

The Mighty Ducks

There’s strength in numbers and to know that you are not an outlier is important. The truth is that most enterprises struggle mightily with their data and systems. Even the most polished companies have underbellies of band-aids, manual workarounds, and black holes. If the outside world really knew, it would be enough to gasp “how can a company run like that?” But in fact, these companies have found ways to run exactly like that. We have all learned to compensate and make concessions, forming a delicate balance (as the actor Michael Caine once famously put it — “Be like a duck, calm on the surface, but paddle like the dickens underneath”)

Enterprises have operated for years without addressing underlying data woes and adding new ones to the mix, for utterly understandable reasons:

• The daily pace of business was too fast to ever be able to pause and fix
• Dollars and talent were constantly diverted to more externally visible needs
• And ultimately, as the problems mounted, the situation seemed almost incurable and impossibly complex

And so here we are, having accumulated tremendous data and technology debt, built entire human departments and accompanying processes to manually untangle the data, and reluctant to reveal the duck-like nature of our current state to those above and around us.

If the Glove Doesn’t Fit…

And into this stew of a difficult, and difficult to explain, situation comes AI, that says give us your clean data and we’ll train our models and bring you speed and cost savings. And each layer of the organization, from the CEO down to the Data Analyst thinks to themselves “what clean data?”

And yet, pursue the AI dream we must, lest our shareholders think we are not forward-thinking or our competitors see weakness. So despite our unaddressed duck-like state, we venture into AI implementation, which usually, indeed, leads us to these outcomes:

1) Overly complex AI solutions: if the data isn’t straight-forward, then the solution will either follow the squirrely path that the data takes, or worse, try to deliver on the snake-oil promise of ingesting dirty data (no problemo!) and figure it out. Either way, the supposed solution starts to grow hair

2) Change management struggles: putting aside workers’ fear of replacement (which is a legitimate issue) for a moment, if corporate change initiatives aren’t efficient and carry quick-wins, then resistance sets in fast. But how can an AI initiative be any of these things if no one is admitting the elephant (or duck) in the room — that the data isn’t ready?

3) Unclear AI goals: when unvoiced doubt is already in the picture, it’s difficult to clearly articulate goals, especially ones that are both impactful and achievable. No one wants to set themselves up for failure. So hemming and hawing ultimately leads to neither-here-nor-there solutions whose benefits are hard to articulate

And the headlines ultimately turn to this type of click-bait — “What happened to the AI revolution?”

There is No Shame

Once we understand that there is no shame, because there is no reason to be ashamed, true advancements can begin.

It’s time to let the outside world see the state of our data. The winners in the coming wave of successful enterprise AI integrations will not be those that have the best data, but rather those whose leaders shed their data shame and bravely venture out to find the data solutions necessary to honestly tackle the problems.

Just remember:

• Even large companies have similar data issues. Theirs are just bigger and more complicated
• Openly share your data woes with those that are there to help. You are not the only one and the more honest you are the faster the fix (it’ll feel great too)
• As a leader, create a safe space for those around you to share their data woes. Saying “I thought I asked you to fixed this years ago” is probably not the right motivator for honest dialogue
• Develop a complete intolerance for manual repetitive processes. Adopt a “do it once” strategy: that means do the manual process just one more time — enough to document the steps and then either automate them or fix them
• Connect to the goal. Ambiguous or conflicting goals lead to ambiguous or conflicting data. When you clarify the goal (what decisions, what metrics and why) the knot will start to untangle

The truth is that we need to get over our data shame and wear our badge proudly. Cooper can be an invaluable tool in your data detox journey, beginning with common sense data assessments and AI-readiness grading through to AI-solution implementation. Cooper helps you shed the paralysis of data management and tackle what is necessary for your business to thrive in the fast-paced data-reliant AI-era.

‍