GA 2.2 Please sir, can I have some more? How to create reports that keep that budget flowing.
- Problems With Bounce Rate and Time on Page
- Segmenting Data for Reports
- Segmenting by Sessions vs Users
- Providing Context in Your Analytics Reports
- Building More User-Friendly Reports
Aiden: Hello ladies and gentlemen. It's good to see you all again, for this, the fifth in the Google Analytics series for SEMrush. This one's around how to create a report that keeps the budget flowing. If you don't report effectively, and it is our jobs as marketers to report, we get fired.
Again, I welcome you, before I pass you to Jill, to How to Create Reports That Keep the Budget Flowing. Jill quick, take it away.
Jill: Thank you Aiden. I'm going to now attempt to share my screen. I am going to do a talk for the next 50-ish minutes... on how to create reports in Google Analytics. Now as Aiden had mentioned, the issue that a lot of people have I think with Google Analytics is that the language can sometimes be quite impenetrable. Like it's lots of big jargon-ey words.
When you're trying to present that information over to people that are going to sign your checks, or approve your agencies, or keep your job, then you have to display that information in a way that's meaningful, that somebody can understand.
How does Google Analytics actually calculate certain metrics? There have been a few times in past of understanding how to use analytics, where I've had my ass kicked, essentially because I didn't fully understand how this gorgeous computer program actually calculates certain metrics. I assumed, and it was the wrong assumption.
We're going to look at the difference between session and user data, and that's really important because of the way that Google will assign a value to particular assets, and you can end up backing the wrong horse if you're not careful.
We're going to be looking at the best ways to use the reporting interfaces. A few hidden gems in there. Then, because I want to stop the death of data vomit and just creating crappy dashboards, we're going to go through some UX techniques on how to actually create some reports.
I've always wanted to be a marketer, I've known I wanted to be in marketing for a very long time, but at no point did I sit there and whimsically think, "Oh, I just can't wait to grow up and create a really big report that takes me bloody ages to create, and nobody reads it, nobody understands it, nobody cares what I have done.
Problems With Bounce Rate and Time on Page
One of the metrics that I really wish people would kind of relax on a little bit is bounce rate. I'm not going to diss this beautiful little metric, but you've got to remember what bounce actually means.
This is somebody that goes to a webpage, and they don't go to any other webpage on that website, or if they click on a third party link, like let's say you go to a webpage and you sign up to Eventbrite. You send them to a third party website that isn't yours, that's going to get counted as a bounce.
I've been in many situations where we have been looking at the work that our content marketers have been doing, that our search engine optimization teams have been doing. Somebody, maybe somebody a bit more senior that's from maybe an older school way of marketing, or they don't understand the metrics, they'll look at this and go, "Well, that's got an 82% bounce rate. That page is terrible. Sack the team. Off you go."
You've got to remember that it's all about intent. There are so many people that are talking about this at the moment. Google won with best search engine because they were really good at usability. They were really good at understanding intent and matching that intent with the user.
You've got to look at your web pages and think, what was the task that the user was trying to achieve when they were on this particular web page? This is one of our blog pages. The bounce rate is really high, but the time on page is really, really high. In terms of meeting the intent of that user, we've done it. It's already happened.
You've got to look at your website pages and think, where would you naturally see a high bounce rate, and where would you want to try and drive people further into the funnel? It's not necessarily a bad thing.
The way that this program is going to work out what your time on page is, is by going, "Right, what was the time between page one and page two?" There's your difference. There's your average time on page.
Obviously if you've got people that are going to, say a blog page on a mobile phone in the morning, and they've been sent there from Twitter, or from your social media, or from an ad, or whatever, and they read the bottom of that web page and then they get on with their merry day, that's going to then register as time on site is zero.
Actually, that's not the best metric to look at, is it? Sometimes the intent is that they're only going to be on that page, or maybe you've done a really good job that there's a sign up on that page and people have completed a goal for you.
What we actually do in our analytics, and we would recommend that you look into, is create something called a custom metric. You can just Google custom metric, “time on page.” It's just another way of calculating in incremental values what the actual time on that page is.
When you're building reports to people, just bear this in mind. I try and avoid reporting on bounce rates and time on page because I think there are other meaningful metrics that are much, much better for you to actually work on.
Segmenting Data for Reports
The next thing is to think about your reports in terms of segmenting your data. There's a bit of a tendency to show everything. Let's say you have 1000 users to your website, and you'll go, "Fantastic. Here's everybody on my web page."
You want to start segmenting your data so that you can actually start to understand what your users are doing on your website. Then you can start adapting your strategies to make them do what you want them to do.
For this you want to be thinking about, well what does good look like for you? Every business is different, but I would want to look at, who are my best customers? If we think of a chart where you've got an axis for time, and you've got another axis at the other side that's for your conversions, you can start to work out, if I've got everybody in the middle that's an average, what does really good look like?
Who spends a lot of time and a lot of money on my website? How many sessions did they have? How many times did they come back and complete those tasks? Who are the worst customers? Who are the people that don't really do what they want you to do and they cost you a lot of time and money? You want to be doing less of them.
Hopefully, you've got some questions that you specifically want to answer as well. You might be saying, "Well I'm focused on a campaign. How did the last campaign do? Show me all the users from New York. Show me only the organic traffic."
You want to be thinking about the work that you're doing and how you want to slice and dice this information in a way to make it meaningful, and a way to show some context to what you're doing.
Segmenting by Sessions vs Users
Now one of the things that we need to be mindful of when we are segmenting our data is how Google defines a user segment and a session segment because there's a big impact in how they actually attribute values.
I've seen this in a lot of reports, where people only focus on session data, or there's only focus on user data, and they still sat there scratching their heads thinking, "Why can't I get my budget signed off? Why can't I get people to understand the value of the work that I'm doing?"
Let's start by thinking about, well what is a user in terms of Google Analytics? Users are your user segments essentially, they are going to show you all the sessions that belong to that particular user, and all the interactions that they had with that particular website.
Whereas when I look at my session segments, it's only going to show the final session, where the users interacted. Especially when you're building something that's going to say conversions, so if you're saying, "Show me everybody that came from organic traffic." They converted. Google is literally sieving through all this information to just pull out a segment, and if you specified it to be a session based segment, that came from organic and converted.
Now the problem with this is you can absolutely knock yourself out with this. Because maybe in the particular business that you're in and the user journey, the first session of a user was organic, and then the second session was PPC, and then the third session was email, and then it was a retargeting message, and then they came directly and converted.
If you're going to be attaching conversion values when you're looking at your sessions, bear in mind that Google is looking for the last hit. I've seen people build reports where they haven't done like a kind of structured approach to say, "Here are all of our users that came from organic search. Here are the sessions from the search." And looking at how they converted and what they did. You've got to be very careful about how you balance these two.
Also be mindful that, when it comes to conversion data, you can only look back 90 days. If you've got a longer conversion path, then you just want to look out for that. You'll find that information in your multi-channel funnels report and timeline.
For the majority of businesses... I would say you would be looking within a 90-day window anyway.
Providing Context in Your Analytics Reports
Now, on to the reporting interface. You need to give context with everything you are doing, and you can get context from looking at date comparisons.
When you go into your reporting interface, in the top right-hand side where you've got the date ranges, you want to be looking at this month versus the previous month. Here's January, and then you're going to compare that with December.
I'd also want, if you've got the data, to look at how does this January compare to the previous January? 2018 January compared to January 2019. The reason why you want to give both those comparisons and then an average is because you're trying to tell a story.
You could compare your January 2019 to January 2018, you might only have 50 people. Then you're showing the growth. Then if you start adding segments to this data to see where you're gaining or not necessarily gaining, then you can start to make better suggestions with your data.
The next thing that I like to do in reports is to add a secondary dimension. All this is, is I'm adding another line to this chart essentially. That's all I'm doing, but it can really help to give better clarity to your reports. I did it a lot when we did our webinar on tracking, but this is another example of looking at hour of day.
This is the “hour of day” report in Google Ads, and it's really helpful to be like, zero, one, two, three, four, five, six. Because you can't assume that everybody understands what does zero represent? Does zero represent Monday? Is that a Sunday? Is it a Saturday? You're not necessarily sure.
You can just click on the arrow to say secondary dimension, and then you can type in the dimension and metrics that you would like to layer next to that particular report. In this instance, I can just start to type in day of week name, and then in my report, I then get the day of week, but then I've got the day of week name next to it. It makes it a lot easier for me to digest that information.
All the reports allow you to do this. You can go into your acquisition reports and you've got the default channel groupings, you can add a secondary dimension to put in source or medium. It's just going to make this reporting interface a whole lot easier to understand. Play around with your data ranges and start looking at secondary dimensions in your reports.
The next favorite button that I've got, in terms of making it a lot easier to digest all of this information in analytics, is the “compare to the site average”. I really like this button because as human beings we're not really designed to process pages and pages and pages of numbers. It's just really difficult.
Some people are like mathematics wizards and they can see all of the insights. I need a little bit more help. Instead of just seeing all of the data, I might want to say, "Compare to the site average."
This is thinking about your segments again. Average being slap bang in the middle. Here are all of my users and the average time on page is X. The average order value is X. The average time on page is X. What is better or worse than the site average? This is where you want to see a lot more in terms of ideas for your segments.
If I click revenue on this report, how much clearer is that? To look at my hour of day report and say, right, day of week, and that's really hard for me to understand with the zero, one, two, three, four, five, so then I add a secondary dimension to say day of week name. Then on the compare to the site average, I can say, "When do I make more money? When do I make more revenue?"
This, visually, is really nice to do as screen grabs, to give it to people to say, "Look, I'm going to start playing around with my budget and when I'm going to be spending my cash, and this gives you a little bit more of an easier way of looking at the data."
Now, my lovely ladies and gentlemen, it's time for you to actually build some reports.
Building More User-Friendly Reports
You might want to build reports directly in Google Analytics using the custom dashboards, or you might be using the lovely Data Studio. I love Data Studio. There are still a few little kinks and quirks with it, but my god is it making things a lot easier.
Because in theory you can now have a blank canvas to build a dashboard that you can repeat across different reporting properties, and views, and marketing channels, and it just makes things so much simpler instead of recreating the work over and over and over.
I think that marketers are quite bad sometimes in building reports. We just throw everything in, give our client that 55-page report, or the boss the 55-page report, and we're done with it. I'd like to see nicer reports. I think some of the techniques that can do that for you are taking some sort of a bit of a lean, a bit of a wink, from the user experience angle.
When I'm talking about user experience, UX is a particular discipline, it's all about designing digital or physical products that people are just going to find a total joy. If you've ever like had an experience where you downloaded an app, or you've been on a website, and you've gone, "Man, that was smooth. That was delicious. That was lovely."
You want to have the same across not just all your customer-facing marketing assets, but you want to do this internally, with your internal communications as well. Because you want to make sure that somebody goes from, "This is really complicated and squiggly, and I don't really know what's happening." To, "Wow, I got from A to B real easy there. I did not have to think at all."
When it comes to our reports, I think that we can start to think about our reports and apply the same structures. We're having clear headings. We are having clear subheadings. We are making sure that we are taking people from A to B, and they know exactly what they should be doing, and they should get the point.
We're going to focus now on how to actually apply these user experience techniques to actually create these fantastic reports that don't waste your time, so we can spend more time doing something with the insights.
Where do we start from this? There's a process that we like to go through, and it's something that hopefully you guys will get on board and start using. We've got a template for this as well, a reporting template on the website. Aiden will put a link in the chat box for this.
You've got to start by answering a couple of questions. What's the actual point of the report? Like, don't report for the sake of reporting. Don't report because you feel it validates your work. Don't report because you have to somehow show the tangibility of what you're doing.
We need to get away from this. We need to start getting better, so what is the point? If there's a reason for you to do this, then do it. Otherwise, you are wasting time.
The next thing to think of is, who is getting the report? Because this can really change what is inside of that document. A little bit like how you've got customer personas.
We're all familiar with this, and if you're doing user experience, you've got the empathy map. Actually, I'd argue that your personas and your user experience should be mushed together anyway. I want you to start thinking about your reporting personas.
I've been in meetings where I have put a really beautiful dashboard on the board to try and get something signed off, and the board have gone, “What's that thing in the right-hand side?" Let's hypothetically say, "Oh, that's the pie chart for bounce rate." "What's a bounce rate?" T
Then you end up spending five minutes explaining a metric on the dashboard that you didn't want to draw attention to. Then the entire meeting has been hijacked and you end up doing something that you didn't want to do when you came out of the meeting, which is not the purpose.
You've got to think about, who am I reporting to, and what do they care about? We call this process, at The Coloring In Department, “Post-it note” reporting.
Okay, so there's a point in building this report. What's the persona? What does the CFO care about? I'm like, oh, he just cares about cost. He really couldn't give two hoots about how many people played the video. He was like, "How much does it cost to get a customer Jill, by the visit? What was the total return on investment?"
You've also got to be thinking about the language that we use. Sometimes as marketers, and I'm sure this is the same in any sector actually, you can end up throwing acronyms and jargon at people that won't necessarily tell you that they don't understand what you're talking about.
The next thing is to think, what are the key messages? Out of the work that you're doing, out of the reports that you're building, what is the one thing that you want them to come away with? What's the one positive thing that you want them to come away with?
You've got to think about how you can work out who you're reporting to, what they care about, what kind of gets them excited and happy, and then how can I not make them think and get those messages to them real quick? Then how can I make it really easy for them to get the message?
I think a gold stamp of reporting is, if you give me a report and I don't have any questions about that report, then you've done a fantastic job in terms of usability. You want to be in a situation where you give them a report, it's just the right amount of information that they need, and you've done it in a way that is visually appealing.
One of the things that I like to do, and this is again stealing a UX technique, you might be familiar with building wireframes for your website pages. I've built wireframes for my email campaigns. Then after spending far too many hours building an Excel dashboard, and then it moved into a custom dashboard in Analytics, and now with Data Studio, sometimes people don't really understand how long these things actually take to build. Some of them can be really fiddly, and real time-sinking holes of doom.
A way around this is I start to wireframe the reports. What's the best way to represent this data? Does it have to be a bar chart, or would a metric work? Would a pie chart help? What if I added a heat map to it? Wireframing helps you clarify, how am I going to structure this report.
A real basic wire framed example, I could sit here and go, right, I'm going to give it a title and I'm going to think about my user experience fundamentals. That's going to be the biggest heading, header one, SEO review.
I'm then going to have headers for each of my different metrics. Don't be afraid to write it in plain English. Don't assume that the person that gets the report understands. Remember, we're thinking about that, don't make me think. Instead of writing something like, "Conversion rates." Like, did they convert? Where did our traffic come from? What pages did they land on? You're almost making it idiot proof when you're looking at that data.
Within Google Data Studio you have an option to put a date range on it. If they go, "Oh Jill, this is amazing. Can I have a look at the data for February?" Do it yourself, Bob. You can just click on this little link, and you've got your own little date range for it.
Then thinking about those messages, I'll be saying things like, "Okay, what's the best way to visually represent this data?" I'll do a pie chart so that very quickly they can go, "Oh right, we get all of our traffic from Google, so your recommendation on why we should opt for Google is valid."
To do this, I'm talking a lot about Data Studio. I think personally that Google is probably going to slowly phase out custom reports, because they've spent an awful lot of money on this tool, and it's really, really useful. Previously I would have to create this report for one particular view, and now I can apply it to all views. I can share it. It's visually very, very appealing. Because that data vomit is really hard for some people to digest, whereas this has been a lot easier to get across.
Hopefully, that's been useful. Final tip, and if there's anything that you take away from this webinar, go and follow Helpfullee. This guy has done so much for the Data Studio community.
If you go on his Twitter now, he's got a whole resource that he has collected. I think at last count it was something like 280 different templates. Templates guys, as in you can click on them and copy them, and borrow them, and edit them, and tweak them. Don't just take them as like, oh I'll have that, and just run away with it.
Still, work through the questions. Still, think about, what's the point of the report? Who's getting it? What's in this template that I can then tweak to my requirements? How can I get that message across to people? Then wireframing them, having conversations with people, and then building, so that you're not wasting your time. There's nothing more soul destroying for me to spend so long building reports, and people take the numbers the wrong way, or they don't even look at the data.
We have gone through in this session, how does Google calculate certain metrics? Bounce rates, time on page, the difference between sessions and users, in terms of how Google's going to attach those hits to them.
You've got to be very careful when you're building your reports to make sure that you get the right output. Looking at the reporting interface. Playing around with the date ranges. Add secondary dimensions. Start using that lovely compare to the site average. If anything, it's fun. Like it just makes the report look better.
Then when we're thinking about reports, not just dumping a load of data and metrics for the sheer sake of it. #Those user experience techniques have really helped me create much better reports and have a better output when I'm trying to tell a story with my data.
Aiden: A few other things have come through in various different ways and means. What's your actual favorite report? What's your go-to report that you like to look at the most?
Jill: Oh, I think it's going to be the assisted conversion report. The reason for this is you can drill down not just by your campaigns, or your sources, or your mediums. You can actually look at this by landing pages as well. A bit like, what was the page that converted? They get 100% of the credit. You might have written a load of blog content, and that blog content actually converted, but later on in the journey.
Aiden: It's really funny actually. We differ somewhat on that one a little bit, and I guess a conversation that Jill and I have quite a bit. Jill loves assisted conversion reports.
Jill: I really enjoy those.
Aiden: I like MCF reports, so multi-channel funnel reports. Speaking to probably my background a little bit, in media, I love MCF reports so much. Now whilst neither the assisted conversion report nor the MCF report, are where you should start for sure, they're really, really interesting to start to dig into them. Both of which live under the conversions tab on the left-hand side in the Google Analytics interface, in case you wanted to go and have a little look into it.
Aiden: Is it worth reporting on bounces in your opinion? Is there any context that you can think off the top of your head, generally speaking, where a bounce is a valuable addition to a report?
Jill: As I said, it should be in context. I'm sure I'm not the only person who has done this, if you wanted to like some people just really get a bite off bounce rate and they will not let it drop. They're like, "Our bounce rate is 70%, and your entire worth as a marketer is going to be judged on said bounce rate."
You can just fire an event in Google Tag Manager to say, "Hey, if somebody plays a video or downloads something, pretend they went to another webpage." You can have it as an interactive hit, and your bounce rate could drop to 30%, and everybody's happy again. You can fiddle with the numbers if you wanted to.
I don't really put them on the report really, because it goes down to the intent. That's where I'd look at content groupings over bounce rate. Show me all the blog pages. Show me all the product pages. Show me all the contact pages. Some are going to have a higher bounce rate, like the contact page.
Yeah, I think it depends on who you're reporting to. As I said, there have been some people that I've reported to, and that's been the one metric that they've taken over the last 15 years and they've just stuck with it. You can tweak it. You can tweak that number, and then they think you're amazing, and then they'll just let you get on with your job.
Aiden: With that, getting on with your job, I'm going to park it here Jill.
Aiden: Huge thank you today to absolutely everybody for joining us for this, of course, the fifth in the Google Analytics series for the wild, the outrageous, the wonderful SEMrush. Thank you very much for all of your questions throughout the previous hour, because it's been an hour together. Take care.