Welcome to episode #88 of The PPC Show where we interview the best and brightest in paid marketing. This week we're joined by Chadd Powell, Sr. Social Analyst at Hanapin Marketing, to dive into Facebook Analytics and show you how to find insights, along with other pro strategies to use when measuring the impact of social media on your business.
- The ins and outs of Facebook analytics and how to measure performance
- The metrics you should be tracking and measuring
This episode is taken from a webinar I conducted with Hanapin Marketing a few weeks so be sure to check out the show notes to see the deck and screenshots.As always you can find us on Apple Podcasts, Spotify, Stitcher, Spreaker, PlayerFM, Overcast, Soundcloud, & Tunein Radio.
Listen to Episode
Show Notes and Transcript
Three Announcements from F8
Chadd Powell: We're going to start with some update announcements. If you're not familiar, F8s Facebook's their big annual developer conference where they make a bunch of different announcements about the Facebook platform from a developer's perspective, et cetera. We do make announcements about analytics. Analytics has been around for a little while now, but it's been, in my opinion, a bit limited in terms of what it could tell you and things of that nature.
Chadd Powell: So these announcements were interesting. They sort of announced three. There's one that I think is really interesting. The other two were sort of add-ons. They're kind of cool but not sure that they're the most useful.
Chadd Powell: Journeys is the one I think is really important. It's a really interesting tool that helps you sort of understand where people are coming in to your conversion channels in your past and sessions, et cetera. So, where does somebody come in, where do they end up, what are the steps along the way? You do have to have faced a lot of different data in Facebook. You have to set up what are called the one source groups and all these types of things, but it's a really, really interesting tool to help us understand that. And yes, do other tools exist? Yes, they do. So this isn't some groundbreaking crazy thing, but it is Facebook's version of it and there is something to that. Especially with their people first tracking and data gathering versus cookie based.
Chadd Powell: The other two, the Facebook's analytics mobile app, cool. You can put it on your phone. Sweet if you want to check. I prefer to be at home with my big monitor and multiple windows open. That's how I like to do my analytics. So I probably won't be using this, but it's neat to have.
Chadd Powell: And then the last one is the automated insights. It's automated insights. So, cool to look at. Some stuff useful, some stuff not. Just sort of one of those things that it puts those insights sort of in the analytics. When you login and you go into the dashboard, I've seen these insights with a pop up in the top right. So they kind of pop down and it can be a lot of really random stuff but it can be interesting. I don't know what you've seen with those, JD.
JD Prater: Yeah, I would agree with that. I really like the idea. I really like this idea of really using this machine-learning to give us these automated insights. But again, some of them, like you said, there was like too granular where it's like, "Hey, males 35 to 44 in New York City where I spent more time on your website this week than last week." And you're like, "Cool." Like, what am I going to do with that? So, yeah, some of them are really good, but I like the idea of where it's headed.
Chadd Powell: I think it's interesting. I think that as the algorithms get better and stuff, these insights will hopefully improves and it service things that are more useful. And maybe they'll give us some options to sort of guide what kind of insights that we are interested in so that we don't get some of the off the wall ones. I haven't seen anything like that in there, but hopefully that might be the case. Sort of the man and machine combination rather than just solely the machine. But these are the three things that we think are really interesting, and we will talk about Journeys a little bit later on as we dive into analytics.
Sort of the man and machine combination rather than just solely the machine.
Chadd Powell: Did you want to go ahead and describe this latest announcement, JD?
JD Prater: Yeah, so just last week, Facebook analytics team up on their blog, make sure you guys are following that blog, just announced that they are going to be doing some really cool stuff around web sessions with bounce rates and URLs. And the really cool one here is really the UTM campaign breakdowns. What that really means is, you're now able to use Facebook Analytics and really understand how other campaigns are performing. Meaning, if you're using and tagging like your Bing Ads, or your AdWords ads, or Quora, or whatever it may be and you're using UTM sourced medium campaign, you're now can break that down into Google analytics. And then you can also build some really cool like segments off of that.
JD Prater: So everybody take a look at using the segment builder, when you can start viewing retention, live time value. So now you can say, "Hey, what is my retention on people that have come in through Google AdWords," or, "What is my retention for Facebook ads?" And you can really see, like are people coming back to your website and kind of build some new cohorts. And again, when you think about Journeys, which Chadd will be talking about pretty later, this is really cool stuff here. I think overall it's pretty exciting, and you can really ... As the analyst in you, have some fun here. So it's like, how many purchase sessions originated from my Facebook page? Or, what is the most popular landing pages for sessions that included a purchase? Or, which pages drive the most newsletter subscriptions, right? I think these are the kind of questions you'll be able to dive into within this new feature that they just launched last week.
Chadd Powell: Yeah, and the thing I thought was really interesting was the focus on session, sort of all the sessions, right? It's about that whole complete journey. So, they want to see is if a session is from somebody that comes in is at the point that they purchase or do whatever your conversion action is, is really trying to understand that. And you can do that with other tools, you can do it with this tool. It's interesting to see what Facebook's trying to do with understanding that, and how do they attract that?
Chadd Powell: So as far as from the analyst in me, being able to see that whole journey and understand how somebody came in on a sight, where they went, how they ended up, and then whether they bounced out or they end up converting is really important. So really interesting to see them develop this.
Chadd Powell: The one thing I do want to say as we go in now into speaking a little bit more about analytics, is remember the analytics doesn't mean anything if your data and your events and your pixels and stuff are not correct. So, I spend an exorbitant amount of time troubleshooting pixels and data collection so that these tools can be useful. So, just remember that. You really need to go through and make sure all of your pieces of data collection and things of that nature are accurate and set up the way you want in measuring what your want, or else none of this really is helpful.
The one thing I do want to say as we go in now into speaking a little bit more about analytics, is remember the analytics doesn't mean anything if your data and your events and your pixels and stuff are not correct.
Facebook Analytics - Funnels
JD Prater: I 100% agree with that one. That one's always a really tough one. So, let's get into some of the active features of Facebook Analytics. So, let's just say you're an ads manager and you're trying to figure out how can I even get into Facebook analytics and to understand these features that we're now going to spend the majority of the webinar running you through. So, let's say you're an ads manager. In the top left, you got the like little hamburger looking thing. Click on that and then make your way over towards the middle under measure and report, and you'll see analytics and kind of like that middle drop-down right there. Click on that, and within that now you're going to be able to be in Facebook Analytics.
JD Prater: One of the first things that you're going to be seeing is Funnels. So, funnels are a really fun way to kind of understand and play with understanding how people are moving through the news feed, your ads, your Facebook page, right? So this is like your timeline, and then ultimately to your website. So, here's the new …
JD Prater: A quick example that I've pulled together, and this is really probably for those of us maybe that are doing some organic type of social media. You're trying to understand is my organic really working, or the reactions that are coming onto ads, are they really working. So this one is looking at the post-reactions, is the very first step that I set up here. So, these are those reactions. Those are those likes, those are the hearts, that's the wow, that's the sad face, it's the angry face. And then, how many of those people that have done that action, what percentage of those went to my website.
JD Prater: So in this example right here that we're showing you, you can see that about 18% of people that reacted actually ended up going to the website. So again, when you kind of think through the news feed and you can understand that a lot of people really are just passively liking stuff. I know I'm guilty of being the anti-social social media guy, right? Where you're just kind of scrolling and you're scrolling, and you're not really doing anything. You're probably just wasting time, and sometimes you end up liking something. But does that mean that you necessarily mean that you clicked on that ad or you clicked on that link to go somewhere. So what this is really saying is about one in five people actually taking that next step.
JD Prater: So, Chadd, go ahead and click one through here. Another example of this one is also how about post comments. So, now you can actually and go and see well, what percentage of people that have commented on one of our actual posts went to the website. You can see right here, it goes up 27%, so people that comment are more likely to go to the website and actually click through. So again, you're kind of thinking through engagement metrics that ultimately one in three are going to be going to your website. So this one's a really cool one. Play around with it.
JD Prater: There's quite a few that you can actually go through and kind of understand through your events. If you use it for Messenger, so this is a really good one as we think through over 2018 and really wanting to understand how Messenger is contributing to maybe website value or to thinking through events or through revenue. You can go in and play around with that as well.
JD Prater: And then this last one here is post shares. So we just talked about reactions, we talked about comments. How about sharing? And again, sharing, about one in four people that shared this post are going to be going to the website. We know that it's not as many people, but something that's also kind of cool, something I wanted to bring up, you can see right next to about that 25% mark right there is the medium completion time. So people on average, it takes them 1.2 weeks. Let's call that, I don't know, one week and a day-and-a-half. Somewhere in there. So it takes around eight days for someone that has actually come in and shared something, to go and visit the website. So it's kind of interesting when you think about it, they're not necessarily doing it immediately. And again, they're using median and median is that 50% mark. It's not the quickest, but it's also not the slowest. It's that 50% mark is the median. Another fun one to go in and play around with when you're thinking through comments, shares, and reactions.
Chadd Powell: That's really interesting, because when we get down to lifetime value, that delay between when people do things, I know that the focus is usually just on immediate behavior. Sort of, oh, they liked something, they're immediately going to go buy it and go to my site and do something right now. And sometimes that's how we try to build the advertising, by understanding that there is this lag and how best to reach out to people during that lag can be really important with your advertising.
Facebook Analytics - Retention
JD Prater: Really cool. Yeah. I agree with that one. That's a good one. Let's talk about retention. So retention and cohorts. I'm going to talk about cohorts in a second, and they're going to look very similar but really the point here that I wanted to kind of drive home is really what retention is. So you get like this kind of fancy looking heat map. Within this fancy looking heat map, it's really important to understand, this is really user that you've retained in your product over time, right? And so one of the ones that I've got here is looking at age breakdown. So this one is 25 to 34. You can see it, it's in the very top right-hand corner there of that slide. It's very small, but this is just an age breakdown. So when I look at 25 to 34, how likely are they to come back after week one?
JD Prater: So you'll see there on zero, that is when they actually entered the website for the first time. Week one is seven days later. Did they come back? What percentage did? So you can see in this example, it's a 4 to 5% that had visited the website came back the week after, and you can see about half of that come back two weeks later and then three, then four. All the way down to like week 12. Week 12 is the longest look back that you can piece together. But it's kind of interesting to break down and to play around with these different types of age ranges, devices. Again, segments you can create and really kind of understand those user behaviors as they are interacting with your website or events, for example.
JD Prater: You can go ahead there, Chadd. Then, here's another example of looking at this one. So this is just a different age range. So again, I was just talking about 25 to 34, and now this one is 55 to 64. So you can see this one isn't nearly as uniform. You can see where people are coming back, where they're dropping off and they're not really coming back as often. So again, interesting data points kind of understand who your most active users are even by age.
Facebook Analytics - Cohorts
JD Prater: Cool. Let's talk about some cohorts. So again, very similar to retention as far as heat maps, but this one is, again, we're looking at that exact same example where we talked about in the funnels. So I just said, how about post comments and now page view. So again, these are people that have commented first and then went and visited the website.
JD Prater: So you can see right here within this example, we can see what percentage of people are doing this. And you'll get like a weighted average, which is the pink line there. Then you'll see this kind of some week selections that I did there. So we can see like March, it was early March for those two weeks. And you can go through and you can select all those different weeks to also add more lines, but for the sake of visibility on this webinar and a kind of a small looking graph, I just went ahead and selected those three, which is an average and then two different weeks.
JD Prater: Just to kind of show you how different weeks are different as far as impact. So we can see with the green line, we had a really strong start and then it kind of tapered off. So for me, I'd want to really dive in and kind of understand what was happening within those weeks, like two through five and maybe even six and understand that within that month period, what was the comments? What did it look like? What was everybody posting? What was the content? What did it look like? And really kind of understand how people are coming back into it to really understand those cohorts.
JD Prater: So go ahead and click one more time and I'll show you whenever you have this graph, right below it you'll see, again, a heat map by this. So this is the exact same thing. These are people that have posted a comment and then clicked on the website, and then can see here as an example.
JD Prater: So in this example, a really great one is if you look down like March 15th through the 21st, during that week, a really strong week. People that came in that week had been more likely to come back. Looking at even down to week three and seeing 17% of people that came in that week are still coming back. So really strong when you think about what were we posting that week that people were commenting on that got them to go and then go to a page view?
JD Prater: So when I look at this, I will understand what content was produced that week and it would really help drive insights. But also really help me with my strategy and understanding what could I be doing better at, because when I look down at maybe April 5th through 11th, that wasn't nearly as strong of a week when we look at comments and then coming back to the website. So that would be something I would definitely want to take a look at when building out these cohorts. And again, play around with comments, shares, Messenger, page views. There is a ton of options there. You can glean a lot of good results.
JD Prater: The one thing I will tell you, make sure you save some of these. I have a really bad habit of creating these goal cohorts and then not saving them and have to go back in and recreate all of them. So that's true for everything, I'm going to say. And for Chadd when you're building funnels and cohorts and retention analysis. If you like what you're seeing and you think you're going to use it again, make sure that you're saving it. You can also share it and export it as well.
Chadd Powell: Yeah, I also want to say really a question kind of relating to cohorts. I just want to stick on it for one second. Running the cohorts can be really good for ... I think you sort of touched on it when you talked about March 15th to the 21st, seeing that sort of strong week where people seem to stick around. I know whether it's something you've put on your organic Facebook page, something you've talked about, or maybe a blog post you posted there, or in the page side maybe you're running a special, or you introduced a new product or something. If you keep track of when …
Chadd Powell: Say you ran something in the week of March 15th or 21st and only ran during that week and didn't run again, then that cohort is going to tell you how potentially that whatever it was, how that performed. And if you see you did something different the next week and the cohort doesn't perform as well, it gives us some idea of what impact over time that whatever you did that week had. That's one way to use cohort analysis, is because you can test different things and different groups of people at certain times.
Chadd Powell: Cohorts can be, they're doing time here, but you can look at cohorts and you can look at it based on age. So you can say that this age bracket performed better than this age bracket if showing them each the same content, because 35 to 44 over time performed way better after seeing that content than 18 to 24 or 55 to 64 or something like that. So cohorts can be really good. If you don't know what you posted at that time or you're not setting it up in that way, then cohorts can not be as useful because you don't necessarily know what you're trying to see in the data. It just looks like, "Yeah, there's a bunch of colors. And this week did better for the first four weeks, but this other week didn't do well. But I don't really know what the treatment was or what I did that would've created that versus other weeks."
Chadd Powell: So that's really a way to use cohort analysis. It requires a little bit or organization upfront in terms of knowing what you did, and then going back and looking at the tools that tell you what the potential impact of that was.
JD Prater: Very important there, Chadd, is really documenting those changes. So maybe we mess with targeting changes? I mean, that would be something that week. Maybe we opened it up, maybe we got more fine tuned. So it really is just kind of understanding and documenting what you're doing every week or with your status with your ads, or your organic as well. So yeah, good point.
Facebook Analytics - Breakdowns
JD Prater: Let's move on to breakdowns. We're going to get into breakdowns, and again, breakdowns are fun. Again, it's going to look at lot like cohorts, it's going to look a lot like retention, but it's really important to understand that you can do something slightly different with a breakdown that looks slightly different than a funnel.
JD Prater: So this is really thinking about dimensions, and it's going to help you answer questions about how people are using your app or your website. So one thing we didn't mention, but Facebook analytics was really built for apps. When it started out a couple of years ago, it was really built for app acquisition and app advertisers, but also just apps in general. And now they're really transitioning it into a website, and so they're kind of merging the two together.
JD Prater: So a lot of times, understanding that will really help you whenever you looking to do this. Let's take a look at how we can use these breakdowns and really understand these things. In this example, on the left here, we have an event. The event that we have here is a post reaction. If I just wanted to see, okay, hey, on the post reaction, what was that device type?
JD Prater: So this one is going to say iPhones are definitely carrying the weight here whenever we look at what types of devices people are using when they are reacting to our post. iPhone. Then you can see tablet, you can see computer. I'm not too surprised by this data. If you guys are following anything with Facebook, about 90% of their revenue comes from mobile. What do you know? Just over 90% of people that are reacting are on mobile. That's really important to understand.
JD Prater: Now, let's go to skip forward, or one click forward here, Chadd. You can see here what we really can do. Again, post reactions. Now it's device type, and now you can do a secondary breakdown to really understand the next part of understanding reactions and then also reaction type. So now we can say, of those reactions that we got, let's break those down. So think of it really as like a pivot table, or really slicing your data and really going, okay, I've got this one thing, well let's break it down into more granular information.
JD Prater: So reactions. How many of those were likes? How many of those were loves and ha-has and sads and wows? Now we can see percentages and everything. Again, no real surprise here, likes are the winner. It's the easiest one to click. If you know within the mobile experience, you have to hold it down. You have to push that like, hold it down and then you can pick a different reaction type. So I think even just that little bit of barrier to entry is enough for people just like. How about yourself, Chad? Have you got to play around with any of the Breakdowns?
Chadd Powell: A little bit. The reaction type is kind of interesting if you're trying to understand something like that. This was definitely more high level. If you're just trying to understand there's some high level insight of wow, for some reason people are still using computers on our site to visit our site and things of that nature. That's really crazy. Even though it's a mobile first world. That would be something that might be ... You might need to understand and make some optimizations to your site and your advertising or something like that.
Chadd Powell: I think this is one of those things I just I grew up in. I click into it. I do some breakdowns. I say, okay, it either confirms what I thought or maybe there's a couple of things that are different and why I might go somewhere else to poke through. So that's kind of how I've treated breakdowns.
JD Prater: Yeah. Then with the secondary part here, we could add in age, we could add in gender, we could add in country. So we can breakdown reactions by reaction type, by gender, by age, by country, by whatever we have in that next part. So you get three different breakdowns. So you can sit there and just go granular, granular, granular. For this example, I just kind of wanted to show an example, again, of something. At least something I thought was interesting was where do these things happen, and which ones do people actually click the most? So that's the reason I chose this one as a pretty easy example for people to go in and test out. Because with Facebook analytics you can get lost in there pretty quick.
Chadd Powell: Yeah and for advertising purposes, going a little bit deeper with the which age or gender particularly are interacting the most in terms of reactions and commenting and stuff like that, could help you understand who you maybe need to be targeting on the advertising site. Who are the people that are really going to share your message the most, or share you ad the most and react to it the most and most likely maybe come to the site and do what you want them to do. So it can definitely help in that regard. Definitely interesting to see all the different types of reactions. It's sad that wow isn't really pulling it's weight there.
Facebook Analytics - Journeys
Chadd Powell: We'll move on to Journeys. This was the new product. I do apologize. This is the screenshot from Facebook, because being a new product, we had some issues trying to get it to populate Journeys. So the one thing I did learn was you have to set up what's called an event source group, which is kind of what ... Think of it as like a container for all your different pieces of tracking. The different pixel, your page, your create soul, your ad account, and all the different things that you can bundle into there along with a variety of other things that then Journeys feeds off of. So it's like you're giving it the water and stuff, and then Journeys goes in there and is able to pull everything out and try to populate it.
Chadd Powell: So, had a few issues. It takes a really long time, at least it did for me, to try to even populate Journeys. So if you think you're going to set it up and click on it and it's just going to be there, it's not. It took ... I literally just set it and then I came back the next day to look at it because it took it a while to populate. Eventually what it shows is something like, well, what's on the screen here. It came from Facebook and it's one that you can see on the left-hand side where Journeys is at inside the analytics tool along with all these other things. You can see what we're talking about, and spoiler, we're literally tracking an order of how the things are listed. So we're just going through all the different tools for analytics in order of how they are listed here.
Chadd Powell: With Journeys, if you've used Google Analytics, Journeys makes sense. You understand what it's trying to do. If you look at conversion paths and analytics, but basically it's trying to show you where people start and where they end. So in this fictitious example, you can see that a huge majority is the web. Starts on the web, ends on the web. What's interesting would be all those pieces above. If this was real client data, what I'd be interested looking at, is how many people are starting on an Android and iOS, and even Messenger and things like that.
Chadd Powell: Then how many people start, particularly, and this has always been sort of the holy grail is, cross device. How many people start on one device, a mobile phone. So for me it would be, say, iOS, but then I go home and I get on my computer and I buy something, right? And I buy whatever I looked at earlier, so I did my research ... That's the whole thing, right? I research on my phone, I buy it later. We know that happens quite a bit. Maybe I'm a little old school and that I'm like, I'm not buying stuff from my phone. I've got to go home to my home network and feel safe in my little bubble.
Chadd Powell: That's what you're trying to understand. So you don't say, oh cut ... You would never do this, but don't target iOS or Android because nobody's buying anything. It's like, people do eventually. They just don't buy it on that device. So when you can see all the different lines here where people are dropping off and where they started and where they ended, you can see quite a few lines drop off and go into the web and things of that nature.
Chadd Powell: So it tells you numbers, it tells you how many people started where and where they ended up. It gives you an idea. I like the graph. There's also a table if you really like the spreadsheet way of showing you things. Another topic gives you sort of these large numbers, in terms of percentage of converted Journeys to how many people converted out the other side of going through what they're calling Journeys. You can call it a session. You can call it a conversion path. Everybody has their own name. How long it takes on average for somebody to do it. The average number of sessions. If somebody does before they complete, so they might come back, leave, come back before they convert. So you want to know how many times do they go back and forth before they convert and stuff like that.
Chadd Powell: The real crux of it is, is at the top right you can see converts on. Here it's purchases in this example, and it's saying the Journey timeout is seven days. You can tweak what events you're looking at. So if it's a newsletter, if it's a purchase, if it's any variety of things on your page, et cetera. Anything that you can pull in from Facebook can be put in here. So if you want to know that somebody comments, how often before somebody comments or reacts to something or things like that, you can go through and play with the different bits and pieces and kind of learn about it. You can look at the Journey timeout and things like that.
Chadd Powell: So Journeys is interesting from that perspective. It's Facebook's sort of way to compete with what we see in Google Analytics, using their own data and first-party trackers and things of that nature. Have you had any chance to play with Journeys yet, JD?
JD Prater: Yeah, I'm with you on that one on getting it set up and getting it optimized. So it does take some time, but this is huge. This is something that I think Google analytics just can't do. It can't do cross device, at least in their free version, which is what most of us are on, right? Whenever you can kind of think through cross device, but then you can actually really understand the value that you're bringing from maybe just a news feed ad on mobile, and then you're seeing them come back on desktop and converting. That's a very different story.
JD Prater: If you are just looking in Google Analytics, you might actually turn off Facebook. And you would say, well it's not working. I'm not getting conversions from this campaign, but now you actually have the data to say, whoa, whoa, whoa, wait. You do. These are working. They are doing really well, because again, it's people based measurement, not cookie device based measurement, which is really nice.
Chadd Powell: In Google Analytics, the way that I would look at assisted conversions and conversion paths to try to illustrate that somebody first interacted with you say on Facebook and then later it converted on your website. This is also a way of showing you what that looks like and where people are coming. Specifically, for things like Messenger and other placements where people may say, "Oh, I don't want to put an ad in Messenger. Why would I? It's people just chatting." But then you might find, well people are actually using it and actually converting on it.
Chadd Powell: So really useful in that regard. Really cool tool. Probably is going to go through some growing pains as its been released. At least that time frame it takes to load and populate data is pretty raw, so start early. Get your stuff set up. Get it going. Come back in a couple of days or day or whatever, and find out what it says.
Facebook Analytics - Percentiles
Chadd Powell: This one's super basic. This is just percentiles. Again, this is one of the most high level sort of just, okay. It's interesting. Maybe I can use it. The thing I highlighted here was on the bottom. This is a client that has fairly large average order because of what people are buying. So this is one that people spend quite a bit of money. Hundreds of dollars, even over a thousand dollars per purchase, so you can see here where it says the top 25% of your users spend an average of $1,400 more than the rest.
Chadd Powell: Maybe if you understand who those top 25% of users are, okay. So I might see this and say, oh, top 25% of users spending a lot of money. Okay, who are those users? Then use analytics to try to figure out, okay, who. Is it under a certain age? Is it other demographic information? Is it certain regions, certain areas. I can tell you this client has various places around the country that people can convert from. So in terms of buying the product and we can try to understand what areas might be the most useful to target from a regional perspective if certain areas are doing better than others. That could be a top 25% and things of that nature.
Chadd Powell: Again, top level perspective, but helps us work our way, sort of peel the layers back on the onion. How to work our way down and figure out more and more insights, get more and more detailed.
JD Prater: Yeah, I would agree with that one. I think this is really good for, you were talking about width, but segments. I think it's going to be great for like big marketing and kind of understanding how many times people need to come back before they purchase. So I think you combine this with Journeys and you'll understand like, hey, four to five, maybe six page views in those top 25% before they actually purchase. So I think this was pretty cool to kind of understand. Then again, building lookalike audiences off your top 25% that have visited the website. So there's some cool things you can do.
JD Prater: You can't do it right in here. Let me clarify. You can't build audiences right within analytics. That went away with the whole Cambridge analytica stuff, but before that you could and that was really cool. But now you have to go into your actual audience section to do it, but again, just like Chad was saying, I'm with you on that. It's cool. It's good to understand. Where the power comes in is you creating those segments and understanding the breakdowns of everything.
Facebook Analytics - Events
Chadd Powell: Kind of quickly just skip through this one. I know we're getting a little bit close on time here if you want to do some Q and A. This is probably more, including this, I get more excited about events and properties probably than most. Just because I spend a ridiculous amount of time in these interfaces troubleshooting the events that people have set up or that I've set up and trying to figure out are we tracking the right things. How much are they tracking accurately. So are we actually getting data and traffic that's coming in for those events, et cetera.
Chadd Powell: So if you go to the events and properties, it just gives you a really nice view of all your events. Everything you're currently tracking and what's coming in. So you can see here, there's predefined ones on the bottom, which are the ones that Facebook prepopulates for you. Then on the top are custom ones that we could custom. We customize it fine, and that's really where the power comes in. If you have custom events that you want to specifically track that are specific to your business and you can do that.
Chadd Powell: For me, this is just, especially when we get new clients and things as an agency, I go in here and I want to see what are all the things that are set up. What is populating data? What appears to just be something that was set up at one point as an idea that was never implemented or was broken? What do I need to reach out and talk about, et cetera, or ask questions about?
Chadd Powell: So it's just a good from that perspective, and it goes back to the beginning when we talked about that your events, and your pixels, and all your data needs to be set up correctly for any of this to be useful. So if you're going to pick one of these events for the other tools that we've talked about, we hope that then is working correctly if we're going to start trying to make decisions on it.
Chadd Powell: To speak even more, I used PPC in an example here, but if we think the event is not ... Particularly within my view, if we think the event is not working correctly or we think that the pixel was placed somewhere it shouldn't be, this event debugging is really important. So I think of events and event debugging as ways of fixing the rest of the analytics in terms of looking at what's going on.
Chadd Powell: So here you get actually time stamps and URLs and the event and when it happened and when it's coming in. I used this to troubleshoot purchase events that either aren't coming in at all and we're trying to figure out why that is. We're getting way more purchases than seems accurate, so we have to come in and see what is firing that pixel so we can get it fixed. Stuff like that.
Chadd Powell: This is a great way because you can go through and tag different events or sort by different events, et cetera, look at different device Oss, do a lot of different things to try to troubleshoot your tracking. Like I said, part of the analytics is just getting it set up in the first place. That's half the battle before you even do anything, and this can be really useful for that.
Facebook Analytics - Lifetime Value
Chadd Powell: I think we might be almost done our last slide. Lifetime values. So the cohort analysis or the cohort tool that JD talked about earlier, this is actually really very similar. It just layers kind of an extra thing on top of it, in terms of assigning value to that cohort and what it means.
Chadd Powell: So in this case, I wanted to make it really easy for everyone to see, so I removed a bunch of lines. What we're looking at here is three age brackets. So you see 18 to 24, 25 to 34, and 45 to 54. The reason I did that was because there was a clear line that over time you can see that on the left side they all start out pretty equal, pretty close, as far as that initial value, but by the time you get to week four, they separate quite a bit. In that older age group, that 45 to 54 becomes worth a lot more money than those younger age brackets.
Chadd Powell: Now that makes sense given what I know about the client and what they sell and how expensive the product can be, but that helps us to understand that hey, if we only look at week zero, we might bid the same on all of these age brackets. But if we understand that four weeks later they sort of diverge in a pretty substantial way, we might say, oh, I'm willing to spend more now on 45 to 54 because I know that the return is going to be there and it may not be there on those younger age brackets.
Chadd Powell: And you can see that I sorted this by all users versus paying users. You can on the top, there's a tab there, you can hit a little button. You can hit paying users, but I thought it was interesting because I don't know which users are going to pay and which aren't. I just know I'm going after the people that I think are in the right grouping that should pay. So what I want to understand is how much should I bid on every user knowing that a certain amount of them will end up converting, but I need to know what I need per every person I'm trying to reach and what I should be paying and so that's what this shows here. It shows that view of how their value translates over time. So it's pretty cool.