Crawling, Walking, Running Analytics: The 3 Stages of Social Media Data

Key Takeaways

up on YouTube. Cool. Hi everyone. Uh I'm Kevin from Buffer. I'm the director of marketing here and I'm really really excited to see you all here today. Um I'm doing a bit of a different experiment. So I'm chatting with this the folks here on YouTube. I'm chatting with folks up here on Facebook Live. Um you're more than welcome to follow along at either place. And I would love um any questions that you might have, any feedback that you have. You can leave comments on Facebook Live. You can leave comments on YouTube. Um, we'll do our best to to catch them as they come in. If we happen to miss anything, you can always email me at kevinbuffer.com. Um, or find me on Twitter, Kevin Lee. And great. Um, thank you again for choosing to join us on these webinars. We've had a really great response for these. Um, about 5,000 people signed up and really excited for mine today. We have one tomorrow with Brian who manages our social media at Buffer. um his is going to be on the Facebook newsfeed and then on Friday we have one with Haley and she's going to talk about social media PR and outreach strategies. So it should be some really really great stuff. Um for the webinar today I'm going to be doing a bit of um sharing with my screen. So I'll share you show you some spreadsheets. I'll also be stepping through some slides. So a bit of back and forth there. And uh for the audience on Facebook live, I'll do my best to to kind of describe what I'm showing on my screen and showing the folks on YouTube. Great. So with that, maybe we can we can jump right into it. I'll grab my slides here and we'll kick off from the beginning. Great. So the topic is crawling, walking, running analytics, and we're talking about the three different stages of social media data. Um this is a topic that's very near and dear to my heart. I'm not sure I'm really good at it, but it's something that I love to experiment with and explore. And what I mean when we're talking about crawling, walking, running, analytics, if you can imagine that across a spectrum, so in order to like children when they begin kind of that forward momentum and mobility, they start with crawling. And that's kind of the idea with social media analytics. You have to start somewhere and it's okay to start with crawling. So you go from crawling, next step is walking. You're doing a little more advanced social media data things. It's still not maybe to the the most extreme level just yet, but it's making progress. Then finally, you have the stuff in place to get to running analytics. And the running stage is when things, you know, they might feel complicated now. I'm hoping to make things a little less complicated for you all today, but it's stuff like growth rates and looking at impressions and these data points that aren't necessarily as easily seen just at a glance. So, today there's three things I love to help help us all learn. Um the first one is how to set benchmarks and calculate growth rates. The second one is three ways to analyze your social media stats. And that's from a top level things like clicks and follows and and those those stats that you hear a lot about. And then in the course of doing both of those things, I'm also excited to show some spreadsheet tricks for maximizing your data learnings. It's a few spreadsheets that I've been working on myself and excited to share with you all. They haven't been shared anywhere else just yet. So, I'm really really grateful that you joined that I have a chance to to share them with you today. Great. So, let's start with baselines and benchmarks. And I might I often use these two terms somewhat the same. Um they're they're rather interchangeable in my mind, but there is a slight difference in in in the way that I look at them. So, a baseline is the average of how you're performing at the moment. So if you're getting, you know, an average of 10 clicks on every social media post that you send, that's considered a baseline. And then a benchmark takes it to the next level. So if a baseline is where you're at at the moment, a benchmark is where you hope to get. And there's a few different ways to come up with these benchmarks and to measure the benchmarks. One of them is to set aspirational benchmarks. These are ones where you're looking at the people who are doing the absolute best on social media and seeing, wow, Nike maybe if you're a shoe store. Like I don't know that you want maybe to aspire that high, but let's say you choose something like that where they're doing, you know, thousands of likes per message. They're getting hundreds of comments. That would be a benchmark that you can set for yourself and try to get there. Trended benchmarks is another great one. And these trends could be industry industry trends. It could be um seasonal trends. It could be lots of different trends that you might look at to see, oh, around this time of year or given my industry or given where I'm at with our followers, like this is where I can get with the engagement or the clicks or the or the the growth with our social media. The next type of benchmark is an earned benchmark. Um, this one's a bit tricky for me to think through, but basically this is like what you're doing yourself. So, you can set a benchmark based on what your current baselines are. And again, this is where the terms get a bit tricky, but basically you take a baseline and then you would multiply this by a certain factor. So if your baseline is 10 clicks per social media update, an earned benchmark could be 10 times 1.2, so 12 clicks. And that would be, you know, a little higher than your baseline. Um, somewhat aspirational, perhaps somewhat trended if your trend lines are higher than your your baselines. Um, but it would be an earned benchmark where you're basing it on the stats that you're already achieving and receiving. And then the last way to benchmark is competitive benchmarks. And these ones are going to be based on your competition. Um, at Buffer, we like to to mention these folks as maybe the people who inspire you. So, if it's other companies within your industry or your niche, that's a great place to look. If it's social influencers um who can also fit in the aspirational category, if they're a bit more, you know, equal or on par with you in terms of your peers, then that's a great place to look. Also, a lot of those stats are available just by going to their profile pages and looking around and seeing what the stats are on the individual updates. And Facebook has a really neat section, too, where you can, you know, within your page insights, you can track and follow pages that you want to see. So could be competitors, could be places that you look to for inspiration and given the stats that you see there, you can set benchmarks based on those stats. Great. So I'd love to talk you through how I can come up with some of these baselines and benchmarks myself um through a spreadsheet. So I have a link to one here and it is mentioned in the YouTube comments below the video. I'll also add it right now into the stream at the side and I will add the link to Facebook live stream as well. Um, this is a link to a a free and open spreadsheet so you're all welcome to to hop in and join and use it. Um, it's got some cool stuff in there, I think. And I'd love to kind of step through it myself. So, I'm going to share my screen with you all so you can see what this looks like in practice. Let me make sure I'm opening the right tab here. I think it's this one. Great. So, this is on our Facebook page at the moment. Um, this is the spreadsheet itself and it's so fun. Um, for those of you watching on Facebook live, what we're seeing on the screen here is a bunch of people joining joining to look at the spreadsheet live here, which is awesome. So, a quick note on this on these spreadsheets. If you're opening it now, taking a peek at it. If you want to copy this and use it for yourself, we would absolutely love for you to be able to do that. If you go up to file, and I think since I'm the owner here, it looks a bit different, but there's a link for make a copy. So, you go to file, make a copy, you can copy the spreadsheet for yourself and customize it and edit it in whatever way you might like. But what we're looking at here is a basic export of your data from this. This particular example is from Twitter. So we have the date that the tweet was posted. We have the tweet text itself. We have the retweets, the likes, the mentions, the clicks. Then the potential, which is calculated based on your followers and the followers of the people who retweeted it. And so with this information, you can do some really cool stuff. And let me first show you where I got this information. I got this from Buffer. You can also get this in an export from Facebook or from Twitter itself. And if you happen to be on Facebook, if you go to your Facebook page, up across the top is a link for insights. And you can click this link. It's going to give you a page summary here. Up in that top top right corner is a link for export data. And it's going to give you some options here for downloading your page data or your post data. We're going to want post data for this example. Then you can select any date range that you'd like. And then the file format, an Excel file or a CSV. If you're using something like Google Sheets, which is what I've used in the example here, you're probably going to want to go with a CSV just for ease of use there. And then same thing with Twitter. Um, when you're logged into Twitter, if you go up to the top right, there is a profile picture up there and click that and go down to analytics. It's going to take you to your analytics home on Twitter. And from here you'll go up to the top and click on tweets in the top menu. And then there's a button in the top right corner for export data. And it's going to export whatever time frame you have set. So I just happen to click on it and 28 days was set. So it downloaded the 28 days of data. You can change that date range to be whatever you'd like. So once you have this data, you can import it into a clean sheet on on Google Sheets or on Excel or you can use this template that we have here. And basically we broken it down into each each stat has a different column. And we've added a little bit of custom custom color here. So um the first thing that we did was we found the benchmarks and the baselines. Um so our average and our median are two things that we calculated. And the reason why we calculate both of these ones is because sometimes we can get outliers that impact our average. And sometimes the median isn't necessarily fully accurate as well. So for instance, on an average, what you're looking at is um combining all those all those numbers together and then dividing by the number of tweets that you that you posted. So if you had one that did a 100 clicks, everything else did one or two clicks, your average is going to be really high, but it's not really going to be representative of where your tweets truly are. If you were to look at the median instead, this is going to look at every single step that you have here in your column, it's going to find the very middle value in that. So if you did have one that was 100 and the rest were twos and threes, um chances are it's going to pull out one of the probably the two or the three in that in that list and show that as the median. So, it's just kind of a checkpoint to see, you know, does this average make sense? Does this median make sense? And where to go from here. So, I added the formulas in for both of these and that kind of lets me see at a glance which ones are doing well and which ones are not. I went one layer further, added some color to the columns here. And this is what's called conditional formatting in Google Sheets. And the rule that we have here is based on the average cell. So if anything is greater than this average cell, it gets highlighted in green. Same thing for likes, same thing for mentions, same thing for clicks. Anytime that we're above average, it highlights it. And the great thing about that is you can see at a glance which ones are doing well. So for instance, this one here, the best copywriting and writer blogs to follow. It's all green all the way across because I know that that one has gone above and beyond for each of the different stats that we had. So again, I'll leave I'll leave you the link for this one. Um it'll be in the email that we sent out afterward. It's also in the YouTube description down below and in the comments on both YouTube and Facebook. Would love for you all to hop into this one um see if it if it works for you. Great. So that covers baselines and benchmarks. The next one I'd love to chat about is engagement. And engagement is a stat that we hear a lot about and we talk a lot about um you know all across social media. It's one of probably the next to followers and maybe clicks and likes and retweets. It's one of the bigger viral metrics that are out there. And it's not super clear what engagement means for us at Buffer. It means the sum of clicks, re-shares, comments, and likes. So those are the four ways that someone can interact with your post. you know, on Twitter, on Facebook, for the most part on any social network. The combination of all that is what we call engagement or total engagement. And I have a spreadsheet for that as well that I'd love to share with you all. So, this one is a rather simple calculation, but it it's still kind of cool to see it in spreadsheet form. I will add the link to all the notes here so that we all can follow along together. There it is for the the folks following on YouTube. And here it is for the folks following on Facebook. And let me walk you through how this one works here real quick, too. Great. I'm going to hop over to the total engagement one. It's awesome to see a bunch of folks popping in and out of this this document as well. So basically we have the same data that we exported before. We have it in column form. We have retweets, likes, mentions, clicks, even potential. And then all I did was add an engagement column which sums up those four columns. Retweets, likes, mentions, and clicks. There's some different ways that you can do this too where we found where people might weight certain stats more than others. So if you're tracking engagement, like let's say retweets matters more to you than likes, mentions, and clicks, what you can do is weight those a little bit more. So instead of um instead of combining the sum of the four of them, maybe you want to double the value of retweets and then add this individual value of likes, mentions, and clicks. Um, there's lots of different spreadsheet hacks that you can do in that way where you're waiting these in different amounts to show like the total engagement or the total value of the interactions that you have on your account. I personally think a total engagement number is a great number to set on its own. You can see not only the impact of the interactions overall, like if everything happens to be high, retweets, likes, mentions, and clicks, but you can see like given this post, this one had tons and tons of engagement, most of it came from clicks, or this one had lots of engagement, most of it came from likes. So, you can kind of better gauge what makes your post successful in this sense. And again, in the spreadsheet, we've highlighted anything that is above average with all of the different engagement stats that we have here. So, we can see one of them um at the mark for this one. I think we might have even put a benchmark stat on this one. Let's see. Looks like it's it's taking in the average here, which is great. Um in terms of the formatting that we did here, let's see. Yeah. So, the formatting is just an average of everything that we have here. Um, what you can do also is conditionally format it based on an aspirational benchmark. So you can set the specific number that you want to highlight or you can do it like what we did with the earned benchmarks where you multiply your typical average by you know 1.2 1.5 to kind of hope to achieve a bit more than what you could otherwise. Great. I'm going to hop back to the slides here. Thank you all for bearing with me as I go back and forth. Great. So, the next part we're going to talk about is rates. Um, we just got a great comment on on Facebook about impressions and rates. Um, this is what we kind of want to talk about in this next section here. So, so far we've done kind of the basic stats. That was kind of the crawling stats where you're looking at the generic stuff. Now, we're kind of moving up to the next step to to walking. Um, and rates is almost to the running stage where you're really getting some really keen insights into what you're doing on social media. So, the basic definition of a rate, you're taking into account all of your social data divided by the impressions that you have. And sometimes you can also look at rates according to the followers that you have. And I'll show you specifically what each of those look like here in just a second. So, we have a spreadsheet for this one as well. Um, this is one that I've personally used for my own social sharing at Buffer. And, uh, it does some really cool things in terms of showing you not only how many clicks that you're getting, but also how many clicks that is relative to everyone else who might be looking at that. So, let's say you do get a 100 clicks on a social media update. Well, that's awesome. Did you know a thousand people see that or did 10,000 people see it? if you got 100 clicks on this post and 10 clicks on this post, how can you compare the two without knowing how big of a reach or how big of an audience each of those received? So, that's really where these rates come into play, where the impression data helps a lot and where the follower data helps a lot. So, I have one more link to share with you all here. Um, another spreadsheet which you are more than welcome to have and copy and do whatever you want with. I hope these links make sense. I used a Google shortener, a Google link shortener for them. Um, which kind of makes them look a bit funky, but they they do work. So, thanks for for trusting me on that one. Great. I just added that link into the YouTube comments and the Facebook comments. Um, it's also in the YouTube description that I'll email out to everyone afterward to make sure that you have these these in place. Great. So, let me pull this one up on my screen here. Excellent. This is the the one for social media rates and it's going to look really similar to the previous ones that we've just went through. It has all of your main stats. It also has the engagement column there. There are some additional columns off to the right which are going to be the ones we're going to focus on. In terms of this specific spreadsheet that we're working in, there's going to be one element of it that is going to need a bit of manual data input, and that's your follower count. So, there's a cell, it's cell M2, and what you do is go out and grab your follower count for whatever social network that you're you're comparing with the data here. For instance, on this example, I was on Twitter, so I grabbed my Twitter followers. and you place your follower account number up at the top. This is going to show you your in this particular spreadsheet your clicks per follower and your engagement per follower. So you can compare to see what these look like. One of the great things about follower rates, by the way, is that you can compare network to network. So, what you're doing is you're looking at um like let's say you had an Instagram account and you had a Twitter account and you have, you know, 50,000 followers on Twitter and you have 5,000 followers on Instagram. Those can sometimes be a bit hard to compare to see, oh, does my engagement on Instagram is it anywhere close to my engagement on Twitter? When you're showing per follower stats, this is where those networks really combine. you can see the power of, oh yeah, you know, we're getting 10 times the engagement on Instagram as we are on Twitter, even though our audience is smaller. That can help you make really well-informed decisions about where you spend your time and the networks that you want to really invest in. So, I'll quickly run through just how this is working here in the spreadsheet. The different fields that we have are clicks per impression, clicks per follower, and engagement per impression, and engagement per follower. What we're talking about when we talk about impressions is very similar to what we talk about when we talk about reach. So, how many people are actually seeing this content that you're putting out there. It's really, really helpful for putting the engagement stats in context to, you know, how many people actually did see this. You can pull these impression and reach stats from Buffer. We do give you reach numbers on anything that you share on Facebook. We give you a potential audience number for anything that you share on Twitter. Um, you can also get this from the networks themselves. Twitter has their own specific reach numbers if you export their data there. And then Facebook will also show you your specific impression or reach data if you export. So you can have a specific column for this in the spreadsheet that we have here. It's column column G which has reach. And then all we're doing is taking the clicks that we have. Um, in this case, you know, depending from post to post, how many you can look at this at the post level or at the overall profile level and seeing how many clicks we got per impression, how much engagement we got per impression. And then with all this, we're kind of setting an average and a benchmark for ourselves also. So, we're able to see, you know, we usually get 03% of people clicking on any posts that we share. So, in which cases, which which ones did we send out that got higher rates? where there are some that maybe didn't show up on our initial click report as having lots of clicks, but they had really high clicks per impression. One of the really cool things with that is that you can use that data to inform the ads that you put out there. If you find a post that might have low click numbers or low retweet, low engagement, but it has really high engagement per impression or clicks per impression, that's a signal to me that this content is something that people are really enjoying. and you might get lots and lots more if more people saw that. So, it's a bit of a data-informed hack to really find out which post might you want to boost on networks like Facebook and on Twitter. Great. I'm going to hop out of there and back into the slides here. All right. Excellent. So, so far we've touched on baselines and benchmarks. We've touched on total engagement and we've touched on rates. And again, this kind of comes back to that concept of crawling, walking, running analytics. I would love if you all are able to use these spreadsheets for yourselves and put your own data in there and start figuring out these things on your own. I think it's really neat once you start making a practice of this, you kind of build that sense of, oh yeah, I know this is a benchmark. I can look at a glance at my stats and see I got more than 20 clicks on this one. That's great. That one did really well. You can look at and see I did, you know, a thousand reach on this one, which is more than usual. It's kind of building that habit of knowing um knowing how well stuff is doing and being able to discover that really quickly. So, when we have this crawling, walking, running analytics, there's a couple different ways to look at it. Also, one way that I might do is looking at single stats. So when you're looking at a single stat, you're looking at something rather in isolation. And I say this too, not that this any of this is bad. I think crawling analytics is awesome and wonderful. Um walking is great. Running is great. So wherever you're at along the spectrum, like two thumbs up to you. This is awesome. Awesome stuff that you're even getting into the social media data part of it. Single stats are going to be the crawling part where you're looking at, you know, an individual tweet. How many clicks did it receive? um you're looking at it somewhat void of context or at least void of the context that you might have if you were walking or running. Um it's still really important to know those numbers of course, but in terms of crawling crawling analytics, those single stats is kind of the the baseline, the beginning of all of it. The next level up is aggregate stats. And so these are ones where you're combining different stats together. This can sometimes be like a total engagement stat for instance where you don't necessarily receive this stat automatically from a place like Facebook or Twitter. You have to put some work in to actually find out what this number is. Then this also informs your single stats. So they kind of build build on top of it. And then the running portion of it is rates. So with the rates, this is when you're getting into the really powerful stuff. So not only do you know the individual level of how this particular post is doing, you know the aggregate of how this post is doing given all the different stats that you can track on it, but then you also know the rates for it. So you know what do these stats mean in comparison to other posts? What do these stats mean um in context with how many people actually saw it? So crawling, walking, running, this is one way to look at it. Another way to look at it, you have stats in different, I guess, from different perspectives is maybe a good way of saying it. Um, you have stats per profile, stats per post, and stats per impression. And these also line up with that crawling, walking, running um, perspective that you could have. So, stats per profile. When you're looking at these numbers, you might be seeing something like, yeah, on Twitter this month, we got, you know, 500 likes. On Facebook, we reached 10,000 people. These numbers on their own are really great to have. Oftentimes these numbers will be the ones that you might present in a report to a boss or a client. Um they might be numbers that you that you personally look at as kind of benchmarks overall. Um setting for like a long-term goal or a bigger goal that you might have. When it comes to the stats proposed, this is digging a level deeper. So, not only do you know the higher level stats, which are great for kind of knowing the overall health of your profile, maybe not as helpful for knowing the specific ways to iterate and improve on the post that you're doing. This is where the stats per post come in really handy. So, in addition to knowing the big picture, you also know the specific picture. You know, which post did, you know, XYZ in terms of clicks or likes or re-shares. You know, which Instagram photo got the most comments, which one got the most the most likes. um what's your Snapchat engagement for this particular story. There's lots of different ways that you can go with this. So, it's going from the profile level down to the post level. That's crawling to walking. Then, one step further would be running. So, you're not only looking at the big picture, you're looking at the individual post and then you're looking at that individual post in context to everything else that you're doing on social media. So, you're seeing, oh, well, this post did this, you know, XYZ in terms of clicks or likes, and it did this in relation to how many people saw it. Then looking at this stat, you can compare it to this other post. This post did XYZ and it did it in relation to this many people seeing it. And you can compare those two. You can see which one did better, which one might be worth boosting and kind of go from there in terms of in terms of the actionable steps that you take based on the data that you're learning. And awesome. That brings us to the end of things. Um, thank you so much for following along here. I did want to mention a couple of resources that we have on the on the Buffer blog. Um, one of them is a blog post all about social media benchmarks. Um, it was one that I had a lot of fun researching and writing and I find there's some really cool stuff out there about it. So, we collect all that information. It's in a blog post on the blog. I will email that blog post out afterward to everyone here and also drop it into the YouTube show notes so you can see what it was all about. And we also put one together on uh social media reports and sending these out to your boss or your client. And these are great as well. A lot of them come from Google Analytics, which is a wonderful source of seeing how your social media ties in to everything that you're doing on your website. And that's something we didn't quite cover today. That would, you know, definitely fall into the running side of things. There's so much that you can do and know there. Um, I would highly encourage you all if you're interested in knowing kind of the way to to tie those two together, your social media and your Google Analytics, your site traffic. Um, I'll send that post around too via email. You can check that out and see if there's any anything that catches your eye there. But thank you all so much for for joining us today. I know I ran through a lot of things here really quickly. If you have any follow-up questions afterward, I'm keenbuffer.com. Um you can definitely find me on Twitter also at Kevin Lee. I'll do my best to reply in both places. Um and the links for each of those spreadsheets also are down in the YouTube comments. They're in the chat streams on both Facebook and on YouTube. I'd love for you all to to check those out and let me know if you have any questions there as well. Well, cool. I think we're a couple minutes ahead of time even. So, I'll I'll leave things here and let you all jump to to your next activities. Um, definitely hit me up with any questions and thank you all so so much for joining me today for this. All right, talk to you soon. Bye.