How to Pick Key Metrics for Small Business

So everyone tells you you're supposed to have key metrics in business. You're supposed to have numbers that determine whether or not someone is succeeding or failing in the eyes of your team. But no one really talks about how you pick them, do they? And that seems kind of weird, doesn't it? Because key metrics, if we go by the definition, at least I have in my noggin, they are important numbers that are used to incentivize human behaviors. They're supposed to guide decision making, help people know this is more important than this thing. So if we get them wrong, how? I mean, that's really bad, right? We're going to have people incentivized to do the wrong thing, to care about the wrong areas. Now, I'm not an expert when it comes to key metric selection, but I've certainly done a lot of it. And when it comes to the work we do at ProcessDriven, when we are trying to build a process, human incentives are one of the key tools we have at our disposal. So I've picked out key metrics a time or two. So what I want to share in this video is a conversation I actually had with a client, Mallory, who's going through this wrestling match with her key metrics inside her business. What you're going to hear is an impromptu discussion around how to pick the right key metrics for the processes, specifically around customer service that Mallory's working on. But whether you're trying to figure out your key metrics for yourself, your sales team, your account management team, whatever the team is you're trying to organize, I think this conversation will be relevant as we talk through some of the challenges and questions we can use to narrow down our pick of which key metric we want to focus on. With that, I'll let Mallory take it away. What's on your mind?

KPIs are on my mind.

KPIs. Let's go. Okay, follow up from last time, right? We talked about leading and lagging.

Yes. And, like, push, pull. I'm a little stuck, though, on customer service. KPIs.

Oh, fun.

Okay. I use chat GPT and everything to try to brainstorm this a little bit, but I think the ones that we need specifically don't line up with, like, a lot of the ones that chat GPT is, like, spitting out at us, like, Net promoter for and stuff. Like, I just don't know. Like, we do a lot of our email and customer service messages through Facebook. Oh, okay. Yeah, primarily, like, we do have some that come through Amazon, but that's been like actually decreasing a lot because we now have a form that we direct people to on our website that goes into Airtable. Oh. And people have been using that, like way more reaching out way less on Amazon. So that's awesome.

We're going to go through the sticky notes verbally so you're not missing out on anything. Just want to make sure, you know, let's go back to the clip.

So there isn't really like a way to like have that, at least not that we've like discovered yet for them to maybe do some sort of like, oh, how did we do? Like, are you satisfied with your service or whatever that we provided? So yeah, this is kind of what I came up with though. So far it's a lot and I, I'm not sure if I'm thinking of it correctly because so like this would be like the lead measure right here is the customer average response time. But then there's all these lag measures where it could be things like, you know, categorizing the issue by SKU or just categorizing the different customer messages by issue. Cost, total cost of refunds, total cost of replacements, customer product rankings. Like, I don't know if you have any that you might suggest because I know that you guys do customer service too, but I'm just a little stuck.

By the way, Mallory's using the term lead and lag metric here. If you're not familiar with those terms, I'll give you a brief definition. Lead means it's something that you're able to influence as the person owning the metric, so you're able to control it in some way. It's kind of like how many times you went to the gym. That's a leading metric of other things. Lagging metrics are things you can only indirectly control. And so they're happening usually downstream of that effort. So going to the gym might be your lead effort. How much you weigh might be a lag effort. And so those two things are kind of examples to help you get this in your head. Mallory has color coded her lead at the the top and her lag at the bottom because she's kind of super organized and wanted to make sure these are visually distinct. If you want to learn more about key metrics, you can find more about those in the video here. If you're a client of ours, we actually have a whole lesson, a training area on key metrics that Mallory's actually been referencing throughout this interview, which you can find inside the process driven foundations program that we Will link here. It's in module three. Yes, module three. All right, let's go back. One thing I like to ask at this point is are any of these so important that you would like, stake your claim and say, I would die for this metric because it's so important to you? Maybe not that extreme, but I should say if you are somebody who's running a customer service department, you should not let your ability to track things limit your selection of key metrics. But for smaller businesses and smaller teams like Mallory, where this is done by a part time person, we don't want to distract ourselves by procrastinating, thinking, oh my gosh, I need to track NPS score, it's the only metric I can have, and then have to roll out a whole new software just to keep a metric. It's often worth that amount of effort for smaller teams or smaller, maybe like less. Less impactful processes. If you are someone who is like all in on customer support is the most important thing in your business, don't let this limit you. I just happen to have a little bit of knowledge of Mallory's background, so that's why I was being a little cheeky here.

See, and that's the thing with customer sir. I mean, obviously, like, we pride ourselves in like customer service, but customer service is maybe like, like out of all the orders that we get, we have maybe like 5% that actually reach out. So it's like a very small, small percentage.

Right?

So yeah, and like, even the person who does it, she spends maybe like, I want to say like 10 hours a week on this.

Of these. My second question would be, it sounds like none of these are like, I need this number, but of all of these, are any of these ones that you do not have the ability to track?

No, these are all ones we, we have the ability to track.

Perfect.

Like, they're all, they're all currently sitting in a report right now. So, so that's the thing. Like, and I think I'm maybe having a hard time distinguishing, distinguishing between like the report versus like KPIs, because, yes, we're starting to do like CEO reports and these are a lot of what we're tracking. So in my mind I'm kind of like, oh yeah, this, all these make sense. Like, these are what we would want to look at and these kind of all tie in together. But I know we're trying to like narrow it down to like three.

Yes. I would say when it comes to reports versus key metrics, key metrics are generally there to incentivize a human. And reports are there to inform a human. So when it comes to these, what I would probably think, things like customer issues by sku, customer messages by issue, those are things that are all outside of the human's control. And if I know that SKU 1, 2, 3 is getting a lot of support, tickets me as the customer service person. I don't know that that would incentivize anything other than me dreading us selling product. 1, 2, 3. So to me, those ones, those. Those core ones seem like something that would be in a CEO report and maybe important to that person or someone who's in charge of product, but maybe not being the key metric for the person who's in charge of customer support. If we're trying to do lead and lag or push and pull, a great way to narrow this down might be to think about what is the opposite. Like, what would happen if customer average response time was exceptional? If we were, like, responding in one second, what harm could that cause? What thing could we see a negative reaction to? That's usually the one I would want to pick. So of these, do any of them seem like they could get maybe worse if we respond too fast? Now I have to give credit where credit is due, where I'm talking right here about picking metrics that complement each other. I'm actually quoting a concept I learned from Leila Hormozi's YouTube channel. And you can learn more about her whole approach to key metrics and her hiring process in this video where I did a deep dive on researching her process and recapping it on the channel for you to steal. But she was the first person I had heard who suggested this balancing act of making sure that the metric you want people to have a lot of, you also put them in charge of the metric you want them to have a little of. So credit where credit is due. Watch the full video there. But thanks, Layla.

If we responded too fast, I guess I don't. Okay, so like, if you're saying if we responded in like one second. Yeah, I'm sure people could maybe think like, oh, there's not a human behind this or something.

Which of these might be impacted by that, though, Right? So if I. I get a message, I'm like, oh, my God, I gotta respond in one second. Spin it off. That'll make my average response time excellent. But it might lower one or multiple of these. Do we see any of these that might be inversely related? No, I'm looking at customer messages. If it. If I respond really fast. But then it takes 50 messages for that to be responded to because I'm not taking the time to really digest the stuff people are giving to me. We might see the number of messages to resolve an issue go way up even if response time goes way down. Similarly ratings perhaps. Maybe people have a negative experience. I know you're not doing NPS scores, but maybe the types of feedback you're getting could go way down. If it seems less thoughtful, I would pick one that is inversely correlated to what you've already picked out.

That makes sense. Can you see there being any other leg measures that I might be missing or not lead measures, I'm sorry.

I think when it comes to leading customer support, most of these things will be happening outside of the department. So the person is going like they have very little control because it is by definition a reactive department. A lot of the stuff that could be done to prevent it would be on marketing side. But you could look at things like the number of self help articles so people can self manage their account or self, you know, if you have an FAQ on how to install this product so you don't need to contact support help desk articles is things I've seen before. Just the quantity. The number of users who visit your self help site, the number of AI conversations as opposed to human ones. Anything here that is focused on resolution pre contacting a human is typically what we'd look at for customer support.

Okay. Yeah. And most of our customer support is just direct contact to a human.

Right. So implementing an AI chatbot.

Yeah. And I think because we spend so little time on it, we haven't been focused on like upping any of, you know, the, that technology that would help us speed things up or even like, you know, give direct them to like a FAQ FAQ section.

And I didn't really do a great job explaining this in the moment, but since we're recording this video now, I might as well expand on it a little bit here for you. We wouldn't typically think of number of self help articles being a way to inversely correlate to how long it takes us to respond to messages, but maybe it could, if we're getting fewer messages, we could spend longer on them or we could just spend less time per week in general on tickets because we have people being able to self serve and self solve their problems. That's a mouthful. That's probably like super terrible sounding in the microphone. But these are the kinds of examples of creative key metrics we can have if we really Think about what do we want to incentivize for this support person in their 10 hours a week. So if you're thinking about this for your own business, think creatively. You don't have to just do. As Mallory was explained at the beginning, she didn't want to do just default NPS scores. She couldn't do it for technological reasons. It felt like a pain in the butt. It didn't seem like the best representation. Hell, there's whole websites dedicated. The fact that net promoter scores, which is NPS scores, are not a very good indicator of how your people actually feel anyway. So get creative like Mallory's doing here and see if there's ways that you can incentivize behaviors that might not be in the box or the standard way that chat GPT will spit out. But let's get back to Mallory. I mean, based on that, a number of customer resolutions might be an interesting metric if you're still trying to narrow this down. Because then you could see if maybe in the future you look at preventing tickets because you want to do a little push and get AI or have a help desk or something. If you started tracking response time and number of tickets, those would be two interesting metrics to see change over time. If technology being incorporated might be something in the future because those two would really change. Maybe people are spending longer on a ticket, but they're having fewer tickets, so the quality is up because the other things are going to AI that could be very interesting because we do want.

To figure out like when we should focus on implementing some of those things. Like what? Like when is it going to get to that point. So. Yeah, that definitely sounds like a good one. I also really liked what you said about how, you know, the, the lead measure or sorry, getting them mixed up, the leg measures are going to incentivize versus just give information. So.

Right.

I think that that really changes how I'm thinking of a lot of these now, so. Awesome.

Yeah, and they might change over time too. Just so you know, we might pick two today and then as you implement AI, you're like, oh my gosh, scratch all of this. The help articles is the best thing I can do. And maybe it shifts and that's okay.

Yeah, absolutely. Well, thank you.

What we doing here at ProcessDriven, just for anybody else who's curious about these key metrics is every quarter each person picks a key metric. And that key metric, they might have two or three normally in their job, but for this quarter, that one key metric is their like their top priority and it gives them a space to focus. So I wish I could remember what Sarah's is right now, but I think Sarah's. Not to speak for you, Sarah, but I believe it's the number of people who start foundations, like the percentage of people who begin the course. And Sarah might have quite a few metrics that she's managing and making sure that everything's humming well. But for this quarter, that's her focus, is to make sure the people who get started on the course don't get overwhelmed and paralyzed. And you know, sometimes it can be when you've got education to do, you're like, eh, procrastinate. She's making that her focus. And so for Mallory and others who are playing with key metrics, we've had a lot of luck with this and I just want to share it here with anybody else who might find that a helpful way to help triage key metrics. So to recap, metrics are things that can be used to inform anybody in or outside your business of how things are going. But key metrics are your tool for incentivizing things you want more or less of when you're making key metrics, making them complimentary and balancing lead and lag metrics can either empower people with leading metrics or demoralize people when they only look at those lagging metrics. And by connecting those dots for folks, you can help people make better decisions by helping them know what is actually the most important. And yes, this can be challenging, but as you kind of saw with Mallory, and maybe it was a bit cut out of this clip, but Mallory's going through and doing that for each system and process in their business, which is awesome. It's the kind of legwork that we all, as leaders and managers owe our teams to do. But of course, like this channel, this consulting call was intended to be a conversation amongst peers trying to help us all collectively do better. So if, if you'd like to join that conversation and share your tips about key metrics, I'd love to hear them in the comments below. And if you enjoyed this kind of unconventional video, I know we don't do much like this on the channel. Go ahead and write the word Mallory in the comments below because I would love Mallory to come see this video at some point and be like, what is happening? Why is everyone yelling at me? In case until she gets to this point, I just think that'd be funny. And by writing the word Mallory, preferably in all caps in the comments below, then I can make sure that we produce more videos like this on the channel showing you behind the scenes of some of the challenges we are solving inside Process Driven foundation to hopefully help you solve those tiny little issues that really can be a thorn in your foot when you are operating a small team. Of course you can also just skip the line and join us inside Process Driven Foundations where I do calls like this every single month with multiple people solving challenges back to back to back. You can join in the description below, but even if you're not ready to learn from other people, make sure you take some time to learn from yourself in this next week. In fact, there are little habits you're probably doing right now that are interesting introducing chaos and disorganization into your business because of incentives you don't realize are there. For more on that, watch this video on the end screen where we talk about how process chaos might be infiltrating your business. Until the next video, enjoy the process.