Dryers and Assumptions
I'm not going to make a joke in this post about asses, you, or me. Honest.
The other day I noticed our dryer wasn’t, well, drying anymore. Aha, I thought immediately, I’ve heard about this. I bet the vent is clogged. And since a clogged vent can lead to pretty bad (and fiery) consequences down the road, not to mention the annoyance of only half-dried clothes, I set aside some time yesterday to investigate.
After fighting through the cobwebs and muscling the dryer out of place, I very quickly learned two things I had not known about our dryer:
It was actually completely disconnected from the exhaust duct; and
It was, in point of fact, a gas dryer.
Now, we’ve lived here for four years, and I had no idea it was a gas dryer. That immediately opened up another major possibility as to why my clothes weren’t drying very well anymore.
So while I had the whole thing disconnected I checked for clogs (there weren’t any), hooked the exhaust back up, and put it all back in place. Today I’ll be reading up on gas dryers. But before I do that, let’s talk about making assumptions.
Where did I go wrong?
My situation with the dryer neatly illustrates the points I want to discuss about what leads us to make assumptions. Why did I assume it was a clog in the first place?
Presumed expertise: I thought I knew what was going on without actually investigating first. Why did I think this? Well, I’d been a homeowner for four years! I’ve lived in lots of places! I know what clogged vents are like.
Incomplete information: I had no idea it was a gas dryer so the possibility that something was wrong with the gas line or pilot didn’t even enter my thinking.
Sense of urgency: exhaust blockages are bad! Like really bad. I wanted to fix it pronto. Better get started now.
Now, the good news is that even though I was probably wrong about the root cause, the steps I took weren’t wasted. I discovered the vent wasn’t even connected! Over time that would have led to a big pile of lint behind the dryer, which itself is a fire hazard. And it led me to discover the more salient fact that I was dealing with a gas dryer. So all in all my assumption was fairly benign.
And let’s not forget that we make assumptions for good reasons: humans are pattern-matching machines, and an assumption is simply matching prior experience to novel situations. Quite often, your assumption will be correct! I might just as easily have opened up my dryer to find just the clog I expected to find. Most of the time, throughout your life, the assumptions you make are either correct or easily corrected. But it’s the cases where assumptions can go terribly wrong that we need to focus on, because technical customer support is an environment where false assumptions can lead to wasted time, loss of customer confidence, and in extreme cases the loss of a customer entirely. There are two key places a support engineer can go wrong by making assumptions: troubleshooting and interacting with customers.
Assumptions in troubleshooting
Troubleshooting, especially under time pressure, is a situation where you will be highly prone to making assumptions. Look at the three points above:
Presumed expertise: even when you know all there is to know about a given topic, or at least think you do, real-world situations are prone to throwing you a curve ball and not conforming exactly to what you already know. The more you know about DNS, the more liable you are to have blind spots for root causes that are actually something else entirely.
Incomplete information: when troubleshooting, you are very rarely going to have the complete picture. If the customer asking for help doesn’t know everything about the situation—and if they did, they could probably solve their own problem—then there’s no chance that you will. You will always be working in a state of partial or complete uncertainty, and that makes it tempting to jump to easy conclusions.
Sense of urgency: if a customer is down, and virtually watching over your shoulder as you try to wrap your head around the problem and figure out what’s going on, you can bet there’s a sense of urgency. Under pressure to figure out what’s going on, you’ll naturally gravitate towards the shortcut of assuming you know the problem and jumping directly to the known solution.
So let’s suppose you succumb to these factors and make an assumption. There are a few possible outcomes. Maybe you’ll be right, and you saved yourself and the customer a bunch of time. Good! But maybe you’re wrong. Best case scenario, you wasted a little time, but maybe learned something else useful about the problem, even if only to rule out one possible cause. Not ideal, but at least you’re making progress. But if you’re wrong and go on a wild goose chase, you’re wasting your time, wasting your customer’s time and burning goodwill. And in the worst case scenario, you may have (for example) directed the customer to unnecessarily rebuild their AWS security groups on the fly, leading to an outage in another system. Instead of saving time, you’ve given the customer good reason to doubt your competence. The rest of this troubleshooting process is going to be more difficult due to the loss of confidence, and the customer won’t forget about this when it comes time to renew later this year.
Given the possible downsides, weighed against the chance of saving yourself some time, it seems clear that making assumptions is not something you want to do unless you have absolutely no alternative. (For more on that, see below.)
Assumptions in customer interactions
Of course, troubleshooting customer issues is not the entirety of the support engineer’s job. Managing customer expectations, monitoring the customer’s sentiment, and escalating customer frustrations to customer success is also a key part of the support day-to-day, and assumptions here can be just as destructive as, if not more than, assumptions made in the heat of troubleshooting. The same root causes apply here, too.
Presumed expertise: we deal with this customer regularly! If there were a problem we’d know about it, right? We can often miss flashing red lights because, as far as we can tell, the customer isn’t having any issues right now.
Incomplete information: the only interactions we tend to have with customers are in troubleshooting situations. So we are only seeing one specific aspect of the customer relationship, and it’s usually a stressed out one. Which leads to…
Sense of urgency: again, we’re usually only dealing with customers live in actually urgent troubleshooting situations, so that’s the baseline we have for interacting with them. This can, paradoxically, lead to the opposite problem: a sense that whatever lingering issues the customer may be having aren’t urgent. Why not? Because they’re not shouting about them. A quiet customer is a happy customer… aren’t they?
Because of all of these factors, support teams have a strong tendency to assume that issues a customer isn’t yelling about aren’t that big a deal, a customer is fine as long as we’re not having constant tickets from them, and a lack of followup on an open support ticket means there’s nothing we need to do. Perhaps it goes without saying, but all of these assumptions can be really dangerous to the customer relationship and we need to take steps to avoid them.
What if you have to make assumptions?
Look, I get it. It’s not always possible to avoid assumptions entirely. Especially in the troubleshooting environment, you’re always going to be working from a position of relative ignorance. Often you can gather the information you need from your own systems or from asking the customer, but sometimes in order to proceed you have to make some assumptions.
In a situation like this, the best way forward is to be exceedingly clear on the assumptions you’re making. That way everyone is aware of the assumptions being made, and there are ample opportunities to correct any misconceptions before they become a major problem. For example:
Thanks for writing in about your issues installing the kernel extension! Based on your message it looks like you’re referring to a Mac OS 13.4 system—is that correct? If so, the following steps should take care of the issue you’re reporting. […]
This both calls out the assumption you’re making and allows the troubleshooting process to continue without waiting for an explicit ‘yes, that’s the right version’ or ‘no, I’m using 12.6’ response.
Minimize assumptions
It’s impossible to make no assumptions at all. Sometimes they’re necessary, and sometimes they’re going to happen without us noticing, because it’s human nature to assume. But we can limit the incidence, and effect, of assumptions by paying close attention to what we do, and do not, know about the situation we’re in.
Build awareness of your own assumptions into standard support processes. When I formalized our ticket response guidelines in my last job, one guideline was specifically focused on avoiding assumptions or, if necessary, calling them out explicitly to the customer. By reminding the engineer to examine their own assumptions every time they address a customer issue, it is more likely to remain at the front of their mind in other customer dealings.
Get more pairs of eyes on the problem. When dealing with customer issues that aren’t strictly technical, lean on your customer success counterparts to help you discover and address customer relationship problems. Ask a coworker to assist if you’re facing a novel troubleshooting issue. Loop in engineering to suggest new directions of information gathering if your established processes aren’t serving you in a given situation. You’re a small part of a larger organization—take advantage of it.
Share the knowledge you’ve gained. When you deal with a new situation, whether by making the wrong assumption and paying for it, or carefully working through the issue, you’re going to learn something new by the end of it. Make sure that others gain the benefit of your pain! Share your findings in a ticket meeting, improve internal documentation, or propose a product feature to avoid this issue in the future.
While taking shortcuts by making assumptions is human, the downsides of getting it wrong far outweigh the occasional wins you may achieve by being right. By being aware of your own tendencies, and bringing in outside help as needed, you can limit the negative effects of guessing incorrectly and improve your overall record in addressing novel customer situations.
Thanks for reading Andy's Support Notes 💻💥📝!