Binary Thinking in Data

"A free for all that's difficult to track, or a process that slower but more managed?"

The above is a quote from a thread I participated in recently responding to my relative indifference to having a formal request process for data teams.

This is representative of an entrenched thought process I often see in data leadership.  One I'm committed to defeating.  Catastrophizing possible outcomes from decentralization of data management away from IT control mechanisms.  Either/or dogmatism, as if there are no gradients or in-between's that can generate successful outcomes.

We talk about data assets and data products, but we fail to build a reliable data supply chain.  Instead, we focus exorbitant amounts of time and expense building administrative overhead like formal request processes before we've built the means to satisfy them well.

Let's say you want to launch a food truck business.  What are your priorities? 

Whatever's necessary to get you parked in front customers, making delicious products, and ringing your cash register as quickly as possible.

Would a food truck owner be overly concerned about having a free for all of demand and consumption of their tasty menu?  I think not.

What happens when a food truck is so popular that it runs out of product before all customers can be served?  Nothing but good things, if the truck owner has a solid business model. 

Customers start lining up early. 

Cash receipts and insane demand drive reinvestment and process refinement. 

Soon you've got multiple trucks expanding your market coverage.

Data teams that fear the "free for all" rather than embrace the exhilaration of building a loyal and returning customer base are missing out.

So launch that rattletrap food truck, but make sure that your product is fresh, tasty, and leaves them wanting more and willing to line up to avoid missing out. 

A rudimentary supply chain that wins with the customer is better than any well-managed ordering process that makes them wait.