Decide the action before you know the problem

There once was a traveler named Wright,
Who traveled much faster than light.
He set off one day,
In a “relative” way,
And returned the previous night.

– James Thurber (I think).

What decisions would we make if we could go back just day? This blog post is about decision making. It was inspired by a question posed over at BPM.com. Recently Satya Nadella, the new CEO of Microsoft, said:

“Any organizational structure you have today is irrelevant because no competition or innovation is going to respect those boundaries. Everything now is going to have to be much more compressed in terms of both cycle times and response times.”

When my daughter’s were growing up I encouraged them to make their own choices and to live by them. I used this children’s riddle often in those conversations. “There are five frogs on a log. One decides to jump off: how many are left on the log?” The answer, my kids will tell you, is “five.” Deciding to jump and actually jumping are two very different things.

In modern business we are too frequently stuck on the “deciding to jump” stage. We get caught in the analysis paralysis of needing more data. So we develop ever more elaborate ways of reporting and visualizing the information our BPM systems deliver and all the while the our more nimble competitors, who have nothing to lose and everything to gain, are jumping off their own logs onto better ones and on to our log. And the more they are successful the more they reduce their decision time. And the longer we wait to act the harder it becomes to act.

Despite all of this we still feel superior because we have more data.

Colonel John Boyd of the USAF added two more steps to the Decide/Act process. He added Observe and Orient before Decide and Act. We now call this the “Boyd Cycle” or “OODA Loop”.

Observe: In business we should be constantly reviewing how we collect data and interpret it. We must observe the trends in the internally gathered telemetry as well as publicly available data. Anecdotes from sales calls, information from job interviewees and blog sites like this one. Data needs to be shared among all the decision makers in a consistent manner.

Orient: Is where we filter that information with what we know about the business, the market, the technology capabilities. But this is this everyone’s weakness. We tend towards “happy ears”, we treat with higher value the information that confirms our point of view. Scientists call it confirmation bias: we have have to teach ourselves to embrace ideas that conflict with our own and improve our filtering system.

Decide: The decision making process also needs to step up its cadence. If we delay we will miss the market, miss the first-mover advantage: we will not get the worm. But in Boyd’s process we should not be making decisions serially: we should make the next decision and the one after that and the one after that so that by the time our competitor responds we have already pivoted and seized the next market advantage while they are uselessly trying to address the old ones.

Act: Empowerment is key. Once the decision is made teams should be empowered to act and execute the plan. Again it is vital to train the team that the tempo of business is going to accelerate and the rate of change will be rapid.

Loop: Throughout this cycle we are gathering more information, filtering it and analyzing it, modifying the next (and next-next and next-next-next) decisions and adjusting the actions in place.

At the USAF Weapons School in Nevada, Colonel Boyd was called “40 second Boyd” because he bet that he could defeat any military pilot in air-to-air combat in 40 seconds using the OODA principles. He never lost that bet.

BPM and ACM are powerful tools that observe the information our business is collecting and orienting that information for executive consumption. We do let those tools take the next step and decide how that information should flow and we permit them to act on our behalf without human intervention.

But we don’t do that very deeply into the decision making process. The horror stories are real: Knight Capital losing $400m in 45 minutes when the automated trading system sold low and bought high. If it had made $400m in 45 minutes we’d celebrate it as a victory of automation. (My own confirmation bias is showing up very clearly here).

So there is room between superficial automation that manages hand-offs amongst team members and complete abrogation of decision making to technology akin to SkyNet.

Satya Nadella is right about one thing, cycle times are short. We can respond with our gut, or we can respond with technology, or we can follow Colonel Boyd and respond with both.

Last week I was at a conference where a VP of Customer Support spoke and said “If we don’t hear from the customer for a while, we call them to see how things are.” This is being inside the OODA Loop, this zigging before the customer zags, this is solving problems before they are reported, this is compressing cycle times and response times. This is using BPM and ACM to inform Customer Support Advocates where they should direct their outreach.

So I would say Microsoft’s new CEO is also wrong, or at least out of date, about one thing too. We should not be focused on cycle time and response time at all but perhaps in pro-active, prescient time. It’s brave, new, science-fiction-come-true, world out there.