Curating Signals to Drive Focus
On the importance of focusing scattered management attention on the things that matter
Management attention is the top constraint in many organizations today. Human attention is scarce. There’s only so many things you can keep track of in your head. That’s why it acts as a constraint. Yet even in the best organizations, management attention is at best scattered.
That’s not me saying it, that’s Nobel Prize winner Herbert Simon writing about in 1973!
The information-processing systems of our contemporary world swim in an exceedingly rich soup of information. In a world of this kind, the scarce resource is not information; it is processing capacity to attend to information. Attention is the chief bottleneck in organizational activity, and the bottleneck becomes narrower and narrower as we move to the top of organizations.
– Dr. Herbert Simon
Half a century later, organizations have gotten even more complex while attention is scarcer than ever. Faced with such complexity, managers do the best they can, but too often they focus on noise. Unimportant things that rarely move the needle.
Scattered management attention probably accounts for more than 80% of poor decisions made today.
What’s missing is FOCUS.
Data and analytics teams were supposed to provide guidance on making better decisions; instead they created even more noise. Thousands of dashboards and reports, inconsistent metric definitions and confusing insights. Yet I do not blame them. Data teams are rarely empowered to be opinionated.
The problem is the methodology they use. Instead of telling managers exactly what to focus on, it scatters their attention even more. You want focus? Here’s another dashboard; here are more tables and more workflows!
I mean that’s the point of self-service right? It makes for great marketing material. Vendors love it when you create more dashboards, more tables, more workflows. It makes them more money.
Ok, Ergest how do I provide more focus?
Curating signals
I believe the most important job for the data team to do is to curate signals. What that means is data teams need to sift through the data, separate the noise from the signal and curate the most important aspects management should be paying attention to. They need to provide focus on things that matter.
Do this regularly and you will soon gain their trust, to the extent that you can slowly transition from being a data servant to a true partner in the business.
So what is FOCUS?
According to Theory of Constraints focus means “doing what should be done” while at the same time “NOT doing what should NOT be done.” Given a clear objective, what management attention should be focused on is far less than you’d expect. Pareto’s rule of 80/20 dictates that 80% of the output comes from only 20% of the inputs.
By the way this does NOT mean have fewer metrics. It means find the right metrics to pay attention to and provide recommendations on what to do about them.
How do you curate signal?
Imagine if the data team can provide a weekly report (maybe a 10-15 min video) where they walk through all the areas of the business management should focus on. What could such a report look like?
Here’s my suggestion based on my analytics goal tree:
Overview
The overview should answer the question “What’s happening in the business.”
Provide a high level overview of the key metrics in the business perhaps the first two levels of the dashboard tree. Make sure to mention what’s working well vs what’s not. The best way to do this is to have goals for each of the top level metrics and compare goal vs actual. The other way is to look for anomalies in metric variation. My friend Cedric Chin has a very long essay on the topic of variation.
Drill down on problem areas
Once you cover the top level metrics in the overview, drill down on areas not performing as expected. The moment you tell someone what’s not working in the business their natural next question is WHY?
To answer this question you need to do root cause analysis (RCA) which I covered in a previous issue. This analysis is best done ahead of time, where you know the drivers of all of the metrics and already have a metrics tree set up.
If there are multiple metrics that need attention, the data team should prioritize the most important and impactful ones that need to be addressed first.
Highlight areas of opportunity
The next section should cover areas of potential opportunity. Best way to do this is to build a causal model of the business. It can be a metrics tree or an Excel sheet; what matters is finding causal relationships between input metrics and key output metrics.
Finding growth opportunities could also involve highlighting potential areas for data science research.
Imagine providing this every week and being RIGHT all the time! How much trust would you build for your team as a data leader?
On the other hand imagine you’re a business executive and you get this trusted report every week. How much time would this save you? What would be its value? I believe its value to be immense, way beyond creating more dashboards, more reports and more tables.
This is exactly the type of thing I consult with organizations on. If you’re interested, simply reply to this email and let me know.
Until next time.
Speaking of signal, thanks for yet another extremely high-signal post Ergest.
In the absence of properly curated signals, how are decisions made? Think "performative decision making". Or just "going through the motions". And the results might not show up for a while.