Live Event

Essentials of Customer Reporting

Mat Patterson

Hosted by Mat Patterson

Customer Evangelist at Help Scout

You spend a lot of time and effort talking to your customers and prospects, but are you learning everything you could be from those conversations? Do you know why people are contacting you, or how the volume of requests is changing over time?

Moving from an email inbox to specialized customer messaging software gives you access to reporting tools that can help you answer those questions and more. If you’ve ever wanted a better understanding of your customers and your business, or wished you had real data to base your decisions on, reporting is for you.

Join Help Scout’s Mat Patterson for this 30-minute webinar, and learn how to effectively use your reports. You will be able to:

  • Understand the key information available in your help desk reports
  • Decide what to measure and how to collect the right data
  • Create a plan for managing your inbound customer contact volumes
  • Capture more useful insights from your customers

Attend live for the chance to take part in the Q&A session! If you can’t make it, please register anyway and we’ll send you the full recording.

Read full transcript

Well, welcome everybody, thanks so much for joining me. So, let’s get started, so my name is Mat Patterson, I work here at Help Scout. I’m talking to you today from way down on the other side of the world in just outside of Sydney, Australia. And Help Scout we’re a customer messaging platform for small and medium businesses.

So, we help you deliver better customer experiences, so we’ve got a help desk and knowledge base. And a really cool beacon tool for live chat, and messaging, that sort of thing. Lots of good stuff there, but this morning I’ve got some slides to share with you as we talk about customer reporting. I’m gonna switch those on now so you don’t have to look at my it’s five o’clock in the morning in Sydney face anymore.

Let me just do that and then we will get going.

Okay. so this is me, so before I joined Help Scout which was a few years ago now, I spent more then nine years starting and running a customer service team for a different software company. So this topic of customer reporting, pretty close to my heart.

I’ve done this a number of times, not always that well, so hopefully I can share some of the mistakes that I’ve made. You can avoid them, if you haven’t already made them like me. talking through customer reporting. What it is, the key information that it can get you, and then getting into how we can use that reporting to understand your customers, understand your business more effectively. And then hopefully at the end we’ll have a few minutes where you’ll be able to ask any questions that you might have using that same chat system there you’ve just seen.

And my colleague Moe is going to join us in the chat there to help out, she’s a product expert. But I wanted to start with, with these bears, and these salmon. Because in my experience, when you’re in the queues and you’re dealing with incoming customer emails, client contacts, live chats, it feels a little bit like this.

You’re in your inbox, maybe it’s Gmail, maybe it’s Outlook, maybe you’re using something like Help Scout, another tool. But, it just feels like a never ending flow of salmon customers, you know, leaping up out of the waterfall at you and you’ve gotta catch them and deal with them. And at first, probably not that great at catching salmon in a waterfall, but you learn. But, those questions keep on coming don’t they?

You can add more bears in, and maybe you start to kind of interact with each other and accidentally try and both take the same question at the same time, gets a little bit messy. But, the point is, no matter how good you get at helping those individual customers, it’s not a battle you can win in that way, right? They never stop coming and so you sort of need to figure out, how we gonna keep up? We need to understand what’s going on. And, so you need to get out of that waterfall a little bit.

Have a bit of a broader perspective, and understand like, who is trying to contact us here? What it is that they need from us? Which of these contacts is really important and valuable and the ones that we want to spend time on? Which of them are just necessary things we have to do to keep going? And which of them really are things, which the customer should never have to ask us?

Problems with our software, problems with the service we’re providing or things that they should be able to handle on their own without having to wait for our help. So this is where customer reporting can really help us. Help us figure out the answers to those questions and maybe other questions that you have, like how can we respond more quickly to our customers? Or, what should we be selling or providing that we don’t?

Or, where is all our time going in the lead up to these things that we’re trying to do? So, this is where our reporting practises can really start to help us. So what I wanna do, I’ll start by just running through some of the very basics, the typical reporting options that you’re likely to have. But, then I wanna spend most of our time really looking at how we can use those tools and make these reports more valuable for us.

So, I realise that listening in right now, we’ve got a probably a pretty broad range of people. We might find some of you have been working in, in Gmail or Outlook in a shared inbox situation, and so you wouldn’t of had a lot of access to reports at all. So, some of this might be new to you, but others of you will be using Help Scout already or using some other sort of help desk tool. And you will have seen these things before so just stick with me, just for a few minutes, if that’s the case. We’re just gonna cover the kind of basics here and then we’re gonna dig in a little bit more.

And for me, I think the two big areas for built in reporting are volume and performance. So, let’s look at those, so. Volume metrics, now I’m showing Help Scout screens here, but any other help desk tool’s gonna have something pretty similar. Volume metrics, really just what sort of work load are you dealing with here? And these, these reports they will help you answer questions like, are we getting busier over time? Are we seeing certain days, certain times of day, or certain times of year be much busier?

What did this month look like in terms of volume compared to the same month last year? What should we expect for next month based on our historical data? And you can click in and understand kinda of the shape of days, the shape of weeks and then into individual channels. So, channel reporting is super helpful for understanding where are our customers asking for help most? Like, are we seeing a shift from people wanting email to people wanting live chat? How much is, how much time are we spending on phone calls compared to live chats or compared to emails? Which channels are we performing best in?

So you start to combine your channels with your performance metrics to dig out all sorts of interesting information. Lots to look at there, but the second big category is performance. So performance reporting it’s all about, you got this big work load, how well are you doing it, dealing with it, and handling it. So, I break performance reporting down into sort of responsiveness and satisfaction ratings.

So, responsiveness pretty obvious, helps you understand things like; when are people waiting the longest? how many replies is it taking us to resolve conversations? How long are we spending between the time we pick up that conversation and the time we discover the answer and are able to send it back?

Now, I’m looking at some pretend sample data here with a very poor resolution time, hopefully yours is much better then that. Satisfaction ratings, also fairly self explanatory, good way to keep an idea, an idea of your general trends. Are people happy with the level of service you’re giving them? And, you can also use this to dig into individual feedback and start to look in the comments there for themes, things you might wanna continue to do, things you might need to address.

So for me, I remember in my last job looking at these sorts of reports, and just one day seeing that on Mondays our satisfaction ratings were significantly worse and try to figure out, what did that mean, why? Does everyone have a bad case of the Mondays here, is this like a Garfield situation? But, it turns out what was actually happening was our responsiveness on Mondays was worse, too. And, that was actually not to do with Monday, so much as it was to do with the weekend. It turns out we had higher volumes coming in over the weekend and our existing weekend support team were a bit overwhelmed.

So, stuff was piling up into Monday morning and then the poor people who emailed in on Monday morning weren’t getting help until much later in the day. So your, your volume reports and your performance reports, that kind of built in stuff. Most of the help desk they’re gonna give it to you and they can be super helpful. But, reports they’re just a tool, they can’t tell you what you should care about.

Or what to do with the information that’s in those reports. And, sometimes those overview type reports, they can almost be a little bit misleading and I’ll give you an example. You can have an overview and I’ve seen this, an overview which looks fantastic. It’s showing you’re, you’re really quick to respond, you’re resolving things in one or two emails, customers are giving you great ratings, which sounds awesome! But, you can be getting all of those great results and actually what’s happening is, you’re hiding a problem.

So, a simple example and something I have seen before would be if you’re getting a lot of very basic questions, like maybe you get a tonne of forgotten password emails. So you get that team, people asking for these very simple things. And, from the customers point of view they can get an answer back really quickly, because the support person, the customer facing person there, they identify that obvious repetitive problem. They’ve probably got a pre-written answer they can pump out for it.

They feel good, they’re getting through their numbers, their metrics are gonna look good, customers happy, everything kinda shows up, but the result is you’ve got that for something that they probably should be able to do themselves without ever having to ask at all. And so, what the report can’t do is tell you that, So, the reports are just a tool that will show you what you ask them for. But, it’s up to you to ask the right questions. And to understand the answers that you get back in the context of your business and what you’re trying to do. It can help you answer questions like, you know, when do I need to hire someone? What product should I be selling?

Should we be changing shipping providers? So, we can go beyond those basic reports and getting to making that reporting a little more useful for your individual business. So we’re back to salmon now! And I realise these, I don’t think that these are actually salmon. But, I wanted the picture of the bucket and the fish. So, what we need to do, all those incoming salmon questions that are leaping up at us. Looking at them all at once is useful for understanding volume, but it’s hard to understand really what’s going on at that level. It can hide a tonne of information just by trying to look at everything all at once, so. We wanna take all of that incoming customer contacts and just break them into a handful of kinda big fat categories. What those categories look like, well, it’s gonna depend on the type `of business that you have.

So, here are some examples. Maybe you’re a SaaS company and you’re looking at you know requests, and billing, and bugs, and knowledge gaps, and presales marketing type stuff. But maybe you’re an E-Commerce company, you’ll have a different set of big buckets that kind of cover most of the things that you talk about. And events again, completely different set of buckets that make sense. So, if your in the queues, if you’re dealing with customers in your role in your company you’ve probably now have a pretty good idea of what those buckets would be. And, what you’re looking for is probably no more then five or six of them, because if you have too many it becomes impractical. You should try and cover the majority of questions, you don’t have to try and capture every single thing in a specific bucket. In fact, here’s an example.

We just put an other bucket here in this demo, that’s what we use to catch things which are not obviously fitting into the other buckets. And setting this up pretty quick, a custom field like this in Help Scout or a similar, similar tool in another product helps you just to start to lump things into big, big fat groups. What do we do with those? Well, the buckets are really just a way of identifying, breaking down that workload. Helping you to understand where is most of the work coming from and where should we focus our time if we want to do something about that workload. If we want to either fix things or identify the obvious places that we can make some improvements. And, as soon as you start breaking down all of your incoming support into like big logical buckets like this you’ll see some obvious issues.

Like, “Oh man, we spend a tonne of time on billing issues,” or, “We’re getting so many shipping questions.” Or, “These one on one conversations with sponsors that we’re having for our event.” Actually, they’re all asking the same question and we could answer them all at once if we could figure out how to do it. So, if you do nothing else after this webinar, or if you disappear right now because you think you’ve got the best thing, you have got it really. Build some buckets, you will learn something, absolutely! But, we don’t have to stop there, you know, everyone will have those top level buckets. Probably you will get a bunch of improvements, and fixes and automations out of that. But, we can go beyond that. And, I wanna talk through three, kind of key, key things that we can do here. First one, measuring the right things then taking control of kind of incoming volume, and capturing customer insights. So, lets get into those.

What should we be measuring? So, those default reports that we looked at, those are great, super helpful. The buckets are great too. But then, we can go a little bit further and start to record just a little bit more information about every incoming conversation as they arrive. Because, that can help you understand later and be able to refer back to it later, and group things in ways that the default reports can’t do for you. I remember one mistake that I made when I first started doing this sort of stuff, was to realise, “Oh, I can tag and categorise things, and filter things.”

And, I went a little bit wild, and I ended up with tonnes and tonnes of tags and categories, and it very quickly became burdensome to actually, you know, categorise anything. There was too many choices, so. Don’t do that. So, what we’re looking for is just enough, right? Fast enough so that it’s practical, so that I don’t have to spend twice as long trying to filter and tag the conversation as I do actually helping the customer. We’re gonna make it pretty clear which tag and with category to use, so they need to be distinct from each other, so that it’s fairly obvious.

And then, the categories and the tags that you choose need to be specific enough to make those reports actually useful later on. If they’re too broad, they won’t tell you any new information. And so, it ends up looking something like this. Where you can start to break down all that big volume into much smaller categories. Now, what should you choose? What tags? What categories should you choose? I can’t tell you that. But, let me tell you how the Help Scout team, the customer’s team at Help Scout, figured out what they wanted to measure, and why they chose that, so.

They went through this process again not too long ago. And, when that came down to two key areas, one was, what is good service mean? They wanted to provide high-quality service and they needed to figure out, well, what do we mean by good service? Because, sometimes your customer thinks they’re getting the best service, but you know you could have done better. You might have higher standards.

So, they sat down and tried to figure out well, one of the things we think it makes for good service is, is first response time. We want to get that down as low as we can. And so, that first response time, it’s built in, conveniently. But, you don’t have to just look at the overview. When you can look at first response time on different days. Maybe, on the weekend. Look at first response time for different time zones. For Europeans, for North Americans, for people way down here in Australia, like me.

And, they even added some kind of real-time feedback and some real-time reporting on to response time, using a work flow, to identify when conversations look like they were getting a little bit too old. Like, no one’s responded, it’s been, you know, a period of time that we can see that too much time, and so, push that into a folder, and have that pop-up and we can see that number right there saying, We need to do something about it. So, that was the quality goals. The second goal that the Help Scout Customer Team had was, to present a really strong data-backed argument to our product teams, they wanted to be able to go to the product team and say,

And now, we’re finding it hard initially, to be able to pull enough useful information to back up that business case, and say the product specialists on our customer’s team reworked all of the custom fields on the tags to break things down into, you know, what type of request are we getting? And then, what particular part of the application is it in reference to? And, then they can do reports and see where there were spikes in activity, where certain parts of the application were creating more customer service work, or where there was opportunity to make improvements for particular types of customers. So, what does that mean for you? Well, you’ve gotta explore your own priorities. You know, what matters to your particular company.

Is it speed like, like Aussies at retention? Or, relationships that you’re trying to build? And then, figure out from there what are the metrics that we can measure that are good proxies for the sorts of behaviours that we want to see, or the sort of information that we want to capture. Think about the business problems that you’re particularly tackling. Maybe, you’re like our team here and you want it to influence product, and so you need to record that sort of information.

But, maybe you’re trying to scale and you would like to identify places where it could add some documentation, or we could add some self-service tools that would really impact on the volumes that’s coming in. And so, then you would need a way to identify those things as they come in.

Or, maybe you’re looking for new ideas and new opportunities, and you need to make sure you’re tagging those and collecting those in some way. And don’t forget, not everything in the built-in reports is going to be relevant to you. There’s gonna be some numbers you don’t really mind if they’re high or low, because that’s not your priority. That’s totally fine, don’t have to use them. And, don’t forget to talk to other teams in your business. The product teams, you marketing teams, your company management, because they will have questions, they would love to answer, and if you can help them answer those questions, using insights from customers, they will love you for it.

And don’t forget, that reporting is not just about the big graphs and numbers. Sometimes it can just be. Can you show me specific people talking about this particular thing? That’s also valuable reporting. All right, second one. Managing those incoming volumes. Like, maybe in that last minute or so, when I was talking about, you know, getting into complex questions that you can answer. You might be thinking,

Which is fair point. Reporting, luckily is a pretty good tool for helping you understand your queue and take charge of it. So, let’s say you’re an e-commerce retailer, and you’re feeling a bit overwhelmed with the volume coming in, and you cannot just hire five more people. Because, who can? So, how would we approach that? Well, start with those five buckets. I mentioned for an e-commerce seller, maybe they’re shipping, product questions, refunds and returns, supply relations, and reviews and feedback, so.

So, we’d start there. We use that bucket, we figure out, all right, which one’s the biggest? But also, which bucket is not necessarily the highest volume but has conversations that take the most time. Or, conversations that are the least valuable for everybody involved. And so, in this example well, maybe it turns out to be shipping. Shipping is the problem.

Okay, well, we just need to understand what could we do about this. Well luckily, for the last couple months, we’ve been tagging these conversations and we can tell that it’s not just shipping in general people are struggling with, but that it’s international shipping and we’re looking at our, our set of data here. And we can say, yes, we’ve done a fantastic job reducing the amount of Unicycle-based shipping questions, but the international shipping, still a problem for us.

So, now we’ve got a possible target. We can use this view to dig immediately in and see actual conversations here. Because they’re reporting, the way we set up our reporting, it’s made it easy to get down to this level, thanks to those buckets and those tags. And reading in those conversations, maybe you learned, you know what? They’re all asking questions that the answer is already on the website.

They’re just not seeing it. And we can confirm, now we can jump over to our knowledge-based docs report, and have a look here and say, Well now, we’ve got an addressable problem, right? Getting people to read this particular page, which has the answer on it. We can try some different things to do that and we also have now a set of reports that is going to show us if that’s working.

We’ll be able to tell. Are we seeing more people reading it? Are we seeing fewer questions in the queue? That’s just one simple example, let me give you a few other possible targets here. Look in your queues for low-value questions. How can you identify and tag questions where self-service would have worked? You know, can you set up a system to capture those and report on them?

Look at your most read help documents, they can indicate, you know, repetitive problems, things that people should be not running into. You can use tags and look for recurring tasks, things that could be automated like account look-ups, like password resets. Look at your contact rate.

This is just, of all our customers, how many of them are having to ask for us for help in a given week, or a given month? And, is that percentage rising? Which suggests maybe we’ve broken something, or we’ve added something more complex, or are we improving it, are we reducing the amount of support per person. And again, volume trends, like I was saying. Maybe you need weekend support, maybe there are key periods where you could intervene and make a difference to the volume.

So, let’s move on to the last key benefit of reporting that I want to mention today, and that’s capturing customer insights. Questions like this, what do our customers love? How do they compare us to the competition out there? How do they describe, you know, what we do and the problem that they’re having?

And what do they struggle with? All these sorts of questions. Everybody in your company wants to know the answers to these questions. Who has the answers? Of course, it’s the customers themselves. And, who are they talking to? They’re talking to your customer facing teams. That information is coming in, usually it’s coming in, but it’s getting stuck inside that help desk, or inside your email inbox, and not going any further. So, what we want to do is figure out how we identify that information. How do we capture it? And then learn from it, so we can actually do something about it.

We can use reporting to excavate some of that information out of there, make it findable, able to be passed on. And again, we’re just gonna use the same set of tools here. Big buckets at the top level, custom fields, tags, and then some reporting views to break it down. You just need the right set of tags and a process for applying them consistently so you can do this sort of reporting, so. Say for example, you’re an event manager and you’ve got three conferences a year, your next one’s coming up, and this is your chance to make some changes and maybe it’s something like,

Do we need to upgrade our catering? Is the people happy with the food? That information you know it’s in there somewhere. People have said stuff about it before. But, there are thousands and thousands of email conversations and chat transcripts to look through. And, you really just want to be able to do something like this. I want to look at the feedback for particular summits, from attendees because it’s not about vendors or suppliers here. It should be in the last year so it’s up-to-date information, and it should be about catering, right, so. All we’ve done again is make it just make it really fast to go from that huge mass of conversations down to a more manageable level. Because, you’ve done the work ahead of time to make sure that those things are tagged. Now, you’ve got something you can spend an hour reading through, and you can make an informed decision about whether or not you should get those tiny hamburgers again.

And I say, yes. Because they’re delicious. So, the key to capturing those insights is to first know what it is that you’re looking for. It’s almost impossible to go back in time and figure out how to answer a question that you weren’t thinking about in the first place when you recorded that information. I know it’s a really common request from businesses to say, “What can you tell me about this particular thing?” But, if you didn’t know that they were interested in that, you probably didn’t capture it in a useful way. Reduce the friction that’s needed, so, just be careful not to put in place a system that relies on a support person. Noticing the information and then taking it and copying it, and pasting it into another form, or doing something complex. You want to make it as easy as you can otherwise, it just doesn’t happen. Encourage curiosity.

So, in the example of, you know, the food situation at an event, if your team knows that you’re interested in answering that question, and you give them permission to do this, as well as answering a direct question about the event, they can also ask, how did you find the food? And then they can capture that feedback. They need to be able to be curious like that. You can use workflows to identify and tag things automatically, instead of relying on someone having to do it manually. Saves a lot of time pulling other data from your help desk, your purchase history, payment plans, interactions with your product or your service. Put all that stuff in makes it easier for a support person to ask a more sensible question. And then, probably the key thing, use your reporting features to be able to make that information accessible to people who aren’t in the customer team. So, reports that you can export or reports you can screenshot.

Views that you can give them where they can dig in and explore, just the subset of conversations that’s relevant to them. Okay, so. Summing it all up. How do we build a solid base for customer reporting? Get to know those built-in reports, consult with the rest of the company about what they want to know about the customers, about the business, and then pick the right set of big buckets, thoughtfully add some custom fields, add some tags, and always be balancing how easy it is to record that information about against how much you’re gonna use that information later on. And then you’ll find, no matter what system you setup, you’re gonna have to adjust it as you go.

Even if you get the perfect one for this month, next month you’re gonna have a new challenge to solve, you’re gonna have to adapt. So, that’s it. Now, we’ve got a few minutes left where I’m happy to answer any questions or to post some questions to you. I’m gonna stop sharing right now, and we can come back to zoom. Okay, and I see there’s been lots of questions that I haven’t been able to see there. But, Moe’s been on the case. So, let me just… See what’s going on. Okay. Just first to answer Meliah’s question, Will there be a replay?

Yes. You will get a recording of this at the end. Okay. “How do you improve consistency in tagging?” That’s a great question. So, for me, part of the problem is, if the tagging’s unclear, like when you, when people don’t do it, it’s often because they can’t figure out which tag to apply. And, if you have too many tags which are too similar, and you’re kind of making that decision like, Uh, I don’t know which way to go. It starts to seem futile. So, make sure your tags are pretty clear, go through and clean them up, maybe do some training, and also, think through what you’re actually using the tags for and show people how you’re using those tags. If they don’t understand the relevance of it, it’s hard to get people to commit to doing it. So, “Should you use custom fields and tags?” I think, yes. I think they can serve different purposes. Having custom fields with some just drop-down options, makes it really quick to roughly categorise things and then you use tags for smaller differentiation within those buckets off them. Okay. Not sure I understand that question, can you rephrase that one? Resolution time, someone asked, How can you find the resolution time? That one is built-in, a built-in report in Help Scout.

If you’re not using Help Scout, I’m sure they have something similar. Time to resolution is probably what you’re looking for. (muffled laugh) Okay. “Do the buckets come preloaded with salmon?” No, they do not. No salmon there. (laughs) Sorry if I confused people with my use of the words buckets, tried to use something which is not product specific, but yes, I do mean essentially just a category or a custom field. Well, that depends on the state that they’re in , really. I would say, anytime you have that conversation about what do we need to know from our customers? It’s a good time to revisit what are our tags, and are we capturing the information that we want to know.

Particularly, I had this conversation with an AI company recently. If you’re thinking about furthering the future we would like to be able to, maybe look at how we would use machine learning, something like that. All of these systems rely on your tags being nice and clear, and consistent. And, if you’re not tagging things consistently now, when you get to that point, wherever that is, even if it’s years down the road, you won’t have the data set that you need. So, we are running short of time. Got lots of great questions in there, so what I’m gonna do, if you’ve got a question, we will capture all of these questions. If you’ve got a particular question that we didn’t answer it today, feel free, when I, what’s gonna happen from here is that we’re gonna get this recording, I’m gonna send it out, we’re gonna get it captioned, and we’ll sent it to you.

And I’ll include in that email some links to articles that I think will answer some of these questions. Including some of the help documents for Help Scout if you want to know exactly how it works. But, you can also hit reply on that email when it comes to you and just ask any question that you have. Happy to help you from there, or we can point you to useful answers, and we can always have a follow up, so. If you did want to get in touch with me, in particular, again reply to that email, or email me, mat with one t, I’m gonna type it in the chat right now for you. And otherwise, thank you so much for spending time with me here today, hope you have a great rest of the day wherever in the world you are, and I hope to see you again in our very next webinar.

Get Started with Help Scout

Looking for a better experience for your customers? Get a free trial of Help Scout for 15 days—no credit card required! Go ahead, kick the tires...we don’t mind.