As I mentioned last week, I am a Halloween enthusiast. When I’m not forcing my wife to hide her face at scary movies, or shoveling pumpkin seeds and pulp into a bowl, I’m thinking about ways that I can traumatize my young neighbors with home haunts and giddy little frights. So if there’s a holiday season worth spilling ink onto across two consecutive weeks, this is the one for me.  

So, having discussed the most explicitly service-oriented spooky movie imaginable last week, where do we go from here? Rather than talk about another pop culture property, let’s talk about what is truly (at least here in the USA) the reason for the season: trick-or-treating.

I, for one, am very excited to take my daughter trick-or-treating. Of course, my daughter is six months old, so we will not be doing that for a few years at least. And though I abandoned trick-or-treating around age 12 or so for “cooler” Halloween activities, near the end, I took great pride in canvassing as broad a swath as possible of an extended network of neighborhoods with my friends to maximize candy acquisition. 

And that’s a distinction that we need to point out, as it’s quite different than service management in how it handled optimization. Yes—both service deployment and trick-or-treating reward maximizing quantity, but there are obviously more dimensions to consider with respect to service—SLA agreement, job complexity, parts management, the list is lengthy. Obviously, all of those elements are important when it comes to true optimization, but for our purposes, let’s assume that that each ounce of candy equals one positive metric of service delivery, whether that be dollar of service revenue, client renewal, SLA compliance, or whatever makes the most sense for your business. Use your imagination. 

With that in mind, what service software capabilities could trick-or-treaters use to be successful? Let’s discuss:

Optimization of Schedules
Young folks might be inclined to move as a single group in order to maximize candy weight output, but in reality, often schedule engines show that what seemed like a conventional wisdom was in fact a hidden inefficiency. Smart scheduling engines will dispatch each trick-or-treater in a way to meet the specific requirements set out by the system, which we have defined as maximization of candy. At a basic level, this can be accomplished by ensuring that each trick-or-treater visit the most houses, and the system could dispatch several kids to specific dense neighborhoods to ensure that the most houses are hit by each individual with no bottlenecks, then the candy can be united into one big pile and divided up. But there are other things that we can use to improve that schedule optimization.

Using Historical Data
My wife insists that we be “the house in the neighborhood that hands out full-size candy”. I find this personally very annoying, because we can purchase, hold onto, and distribute far less candy than we could otherwise (also meaning your humble author has less candy to sneak at 3AM, which is actually for the best), but also because now, years in, some kids know we’re the house in the neighborhood that hands out full-size candy.

Of course, that sort of information, when maximizing candy weight, means more candy with fewer stops. For that reason, logging historical precedent and building those assumptions into your optimization engine will naturally increase the output. To do so, the scheduling engine may prioritize those houses first, even if they’re not on a linear path, in order to make the best use of time. A good system will of course benchmark doing that against following a liner path. 

Simulated Assumptions
That same historical data can be used to build some assumptions if you’re looking to trick-or-treat in new “territories”, so to speak. Let’s say we know that every fifteenth house hands out toothpaste rather than candy, and we have some criteria for what defines those houses—perhaps the public record indicates that 75% of dental employees hand out toothpaste rather than candy. We can build those assumptions into our models, and attempt (through simulation) to avoid any houses that generate a less-than-optimum output of candy weight. 

All these little things may be too much for the average eight-year-old dressed as Captain America to be concerned about, but when it comes to effective service delivery, there’s certainly a great deal of small things that can make a big difference. Taking optimization seriously might help prevent a truly terrifying outcome for your business.

Happy Halloween, everyone. 

Tom Paquin
Author

Contributor, Future of Field Service