Ernie Arboles

A challenge in assembling a diverse group is overcoming our own (or your leader's) bias. Be honest, we all have biases. Problems emerge when we are not honest with ourselves about this. What to do? Seek out perspectives, opinions, arguments that make you uncomfortable and that you disagree with 100%. This is not about being right or wrong, it's about hearing all perspectives. Ensure they get a seat at your table. That way you can fashion a solution, outcome, path forward that's fair, balanced, inclusive, and most importantly, makes your business, stakeholders, and customers better off.

Critically important to start with the right problem statement, hypothesis or whatever you want to call it. It's a clear articulation of what's NOT happening that needs fixing. Next, brainstorm the data you need to provide insight to that problem statement. Diagram the inputs & key activity steps at the core of process & decide which metrics (think good, cheap, fast) give you the most insight. Don't be fooled by easily available data. Get the data you need, even if it's going to take longer, cost more, or ruffle some feathers. This is better than relying on "lazy data" that may lead you astray. Also, consider stratification factors (regions, business units, customers, etc) to help create compare & contrast factors for later analysis.

A great best practice here is to start with a Simon Sinek-ism: "Start with why?" Why do you need the data? What decisions will it inform? What's the value (think customer/revenue/share) of that decision? Then you make the trade-off judgment on time/energy/effort to get that data versus the decision-value it offers. When earning my Master Black Belt in Six Sigma at GE, we used to joke about "quality projects." You didn't want to waste time analyzing where the water cooler should go in the break room. But you did want to spend time understanding why a commercial went to black during a prime time on a Thursday night in May (ratings). As they say, "follow the money." Same applies to the value of data today.

Listen to those who know the process best. They'll you what works, what doesn't, where the pain points flourish unabated. The problem is that in many companies someone else knows better, wants the fix done yesterday, which usually leads to treating symptoms and not root causes. You can expect to be back addressing the same problem in a few months. But when you listen, and gather the input for the process experts you not only get a better solution set you got something else for virtually free: Their buy-in to get the idea implemented.