This is the 8th post in a series on support systems (though it’s evolved to be more about business systems), which began with Your First Support Model. The most recent post in this series was about High Performance Standards.
If you’ve been keeping up with the series on support systems—which evolved into a series on business systems in general—you’ve become familiar with terms such as stocks, flows, and feedback loops. We’ve modeled systems with boxes and lines and circular arrows, and talked about boundaries of the systems.
But what if I tasked you to focus on improving an area of the system where your only concern is to make the system more efficient? What would your model look like and what would you draw? Would you even have feedback loops?
If you’re wanting to model a business process so that you can ensure the process is efficient, you’re now entering the arena of “process analysis”, a function of operations management. Today we’re going to take a small peek at process analysis and how it can be used to complement an understanding of “thinking in systems.”
Modeling business processes
I have to admit that one of the things that was in the back of my mind when reading Thinking in Systems1 is that there seemed to be a dearth of online resources modeling business systems using the tools outlined in the book. It could of course be that I’m not looking in the right places, but I also suspected that perhaps the systems language used in the book isn’t the only way—and perhaps not the most common way—to model business processes.
A couple of weeks ago, I virtually attended an MBA class at Georgia Tech called “Manufacturing & Service Management”. In preparation for the class, one of the students was kind enough to send me a few papers they were planning to discuss during the class. One of the papers was “A Note on Process Analysis”. I tried to find an online resource that discussed the concepts in that paper as clearly and succinctly as the paper itself, but came up short. This site does a decent job introducing the terminology, but it’s not as thorough as the paper. If this topic interests you, maybe consider spending the $8.95 to pick up the paper.
Anyway, the class was really interesting, and it opened up my eyes to a way to model business processes that’s different from what you see in Thinking in Systems. Here’s a diagram from the paper that shows a dry cleaner’s process-flow diagram:
There are a lot of steps to the dry cleaning process. When you start looking at multi-step processes in this way, you can start identifying bottlenecks, that is, the resource that limits production. In the image above, the bottleneck is the packing station, which can only handle 40 orders per hour. Even if the pressing station is at full capacity and cranking out 48 items per hour, the packing station can can still only handle 40 items per hour, meaning that 8 items per hour are going to build up behind the packing station.
A super simple process analysis of a support inbox
When we start looking at a basic support system in a way that’s similar to what we explored when we modeled the support system, the language of process analysis can start to seem a bit out of place. For example, in a manufacturing process, you might have a machine that can produce a batch of 1000 widgets in an hour. A customer support agent, however, can only process a ticket in a batch size of 1, so it almost seems silly to talk about “batches”, “setup time”, etc. when referring to a support agent.
Let’s model it anyway, sure why not.
I mean, it’s pretty simple, right? You have an inbox that represents a store of tickets. Those tickets go out to agents who process tickets in batches of one, at an average run time of one ticket every five minutes, or a capacity of twelve tickets per hour. The entire “Responding to Ticket” team has a capacity of 36 tickets per hour. If more than 36 tickets per hour accumulate in the ticket inbox, the team isn’t going to be able keep up. You’re going to have to purchase a new “Responding to Ticket” machine, er, I mean you’re going to need to hire a new agent.
Alternative solutions
It’s weird to think of agents as machines, but let’s keep pretending for a second. Maybe you can upgrade your “Responding to Ticket” machines to get them to process 16 tickets per hour instead of 12. Now the whole team can handle 48 tickets per hour instead of 36. You saved having to hire an agent—not bad!
But what if there’s a machine you could buy that had an increased batch size, that could serve multiple customers at the same time? If that machine could also serve customers at a faster rate, it would potentially have a much higher capacity than a “Responding to Ticket” machine.
If you shift the model from thinking about “Ticket Inbox” to instead thinking about “Customer Questions”, you can find ways to answer customer questions more efficiently without necessarily needing to respond to support tickets. For example, you can add online documentation which can answer customer questions before they ask. Or you can improve the product, precluding the need even for online documentation.
Systems and processes
I’ve shared just a handful of concepts from a single paper on process analysis, which itself was just a part of a single class on Manufacturing & Service Management. You can get an entire master’s degree in Operations Management and make a career out of it if you want. By taking the time to model and visualize processes, you can find opportunities for minimizing costs and making those processes more efficient.
Where does “systems thinking” come in? I’m still thinking a good bit about that myself and haven’t fully synthesized how systems thinking and process analysis work together. The best I have so far is that process analysis is focused heavily on the balancing feedback loops of the system, that is, the areas of the system that are focused on minimizing. Minimizing is all about reducing costs and making operations more efficient.
A business can’t thrive by only minimizing costs; it’s also going to need to identify opportunities for strategic growth and alignment2, and that’s outside the scope of pure process analysis. That’s where identifying the reinforcing feedback loops come in (i.e. “how will we grow?”). How do the reinforcing feedback loops you identify interact with the business processes you’re trying to analyze? How do you model that?
I mean these as genuine questions—I really don’t know the answers. If you have thoughts on this topic, I’d love to hear them. Feel free to leave a comment or reply by email.
As mentioned throughout the series, I’m borrowing terminology and concepts from Thinking in Systems. Highly recommend.
This also makes me thinking about Michael Porter’s article, “What is Strategy”, which I covered last year in the series on Is Your CX Strategic?. As Porter succinctly states, “Operational effectiveness is not strategy.”
The first thing I think about is the costs associated. The reinforcing feedback loops increase absolute operating costs. In your example above, if 1 new customer means you need a 4th agent, is that 1 new customer paying enough to be worth it? In growth, the answer is almost always yes, since it's not just 1 new customer, but it's good to know what those incremental costs are, and where those break points are. Like if that new customer is in APAC, do we then need to stand up APAC support? Or does that new agent mean we need to reorganize the team and promote a new team lead?