Skip to main content

Workplace Analytics and AI to Optimize Sales Energy

Author by Nathan Lasnoski

Ever hear that multi-tasking is a fraud? There is always an opportunity cost. When we choose to spend time on one thing, we lose the time from something else. The switching time has a cost, the time spent has a cost, and the choice between alternatives is always present. What we don’t always do is make the right choice because we listen to the loudest voice rather than the most important one. This is often true in allying the energy of our company and the energy of our sales organizations toward the right target. In the following example I’ll show how we can use AI and Workplace Analytics to direct our energy toward the right customers and the right opportunities.


First, let’s look at an interesting report within Workplace Analytics which paints the picture of our energy spent on various customers. Workplace Analytics is gathering the amount of engagement we have with each customer. The contrast is between collaboration hours and total billings. What is often true, as you can see in circle #1 is we can spend too much time on the wrong customers, in this case being those that spend little money with our company.

Then, in the following diagram, you can see we’ve highlighted our good customers. Note that they are spending much more money with us, but we are not directing as much collaboration hours in their direction.

Now, we could generally select every customer in the box and try to collaborate more, but we don’t have an endless supply of people. For instance, one of the good customers may not be able to produce at that level every year, so more energy may not pay off. Another may have had turnover so they are taking a pause. The hard thing is to predict what customer will be the best partner moving forward. The goal of a ML exercise is to then apply ML learnings to select these customers. We can do this by combining Azure ML, Dynamics data, and Workplace Analytics to select a number of customers. The outcomes is the following:

You can see that we’ve reduced energy in the low spending accounts and targeted specific accounts to increase energy with a payoff. To pick these with arbitrary understanding might have led to the same result, but it is doubtful we would have picked the right accounts. Granted, we might have lucked out, but we more likely would have missed. Machine Learning lets us apply factors and then even test if our prediction would be correct based on historical data and then proactively against future results.

After we’ve picked our targets, we then need to monitor and see how we’re doing against target. In the diagram below you can see we have our top customer targets and we’re monitoring how we’re doing against increasing the collaboration with the customer. You can see that some are increasing and others are decreasing. How much you want to be it correlates to growth? Few companies have this much intentionality and fewer still have the ability to monitor it.

With the ability to see the relationship between collaboration and results, the capability to apply ML, set direction, then track outcome, a business can really see the relationship between strategic decisions and results. But… wait… isn’t this like spying on our team? Where do we draw the line between performance management and micro-management? In this case Workplace Analytics really helps us. The very specific information is de-personalized and aggregated, so the goal of understanding relationships is maintained, but the individual-specific information is removed. If we want individual specific information, we need to go to a person-specific scorecard that is a shared expectation with the employee. For instance, a co-worker, Jeff Lipkowitz created this scorecard for a brokerage organization, based on shared expectations with an employee and commonly understood goals.

In this case you can see we’re tracking entries per day, standard deviation, and consistency. These are all things we care about in a brokerage scenario and are commonly accepted. We wouldn’t however track things that would be considered obtrusive in his scenario. The interesting thing is that what is considered obtrusive in one scenario would be certainly obtrusive in another. Where in brokerage we’re totally ok with tracking entries, in another industry it might be considered micro-management.

Net-net, the responsibility to be an ethical and appropriate organization is on the leaders of the business. We need to take seriously where the line is for each organization and working with the team members to find the right balance, in a transparent and honest way.

Nathan Lasnoski


Nathan Lasnoski

Chief Technology Officer