So What? Auditing Your Content Marketing

And we are back. Uh happy Thursday, friends. Today, uh it was the uh marketing analytics insights life show. We are talking content optimization. I have John here.

Katie is off on uh well-earned vacation uh running around uh the lovely state of Massachusetts. So John, let's talk content optimization. Uh and I figured we would use marketing over coffee since it's you know the site is uh what a decade old, though over a decade old now, right? The perfect whipping boy, yeah. It's completely under uh utilized and neglected in many ways.

I don't know the last time we dug in, we were actually it was pleasantly surprising. Things were in fairly good shape considering the neglect for the site. So yeah, this will be fun. Um and so what we specifically wanted to look at are what are the pages and things that are on the site that could use some love and what are the pages that are doing reasonably well. So I figured we'd do this in a couple of different ways.

We do this uh with all free tools um and with just some basic stuff like spreadsheets. Even though that I think there are you know many, many, many more advanced, better ways to do uh content optimization, it also a lot of it requires programming, and we're not gonna touch on that today. Uh maybe we'll do that like as an advanced show at some point. So first things first, let's make sure that uh we go into our Google Analytics, and one of the most important things we can do with uh Google Analytics is make sure we have a goal set up. We can't do content optimization without goals and goal values.

Even if the goal doesn't have um a known monetary value, we can at least set up something. So I'm gonna go here to newsletter sign up. And when somebody gets to the thank you page, uh we gave them a value of eight dollars. Uh, do you do you remember how we came up with that value? I think that was yeah, that was poor back solving math.

I think at one point we were like, well, we make this much on sponsorships for this many listeners, and so we divide that out and we get to eight bucks. But yeah, that seems to be a very arbitrary uh number that we need to go back on. Exactly. But you know, it it's better than nothing, and and there at least is a somewhat semblance of logic uh behind it. So let's go to our behavior section first.

Let's go and go to site content and just start poking around to see what pages on the site are doing something, anything. I'm gonna look at the last 90-ish days here. Let's go ahead and put the side menu away. And what we see here is obviously the number of pages, the pages that get the highest views, the unique page use, entrances, bounce rates, exit rates, and page value. Now, it probably should clarify a couple of these things.

An entrance is when that's the first page in the session that uh somebody lands on. Obviously, the home page is a big one. A bounce rate is when somebody arrives on your site, goes to a page, and doesn't trigger any events and doesn't go to any other pages, they just leave right after that. That's that's considered a bounce. Now, there's a couple of challenges with bounce rate.

Uh, a lot of folks like to use that as a as a uh content uh marketing analytic uh metric. There's two challenges with it. One just because somebody bounces doesn't mean that it's a bad thing. If you think about how we use the internet, right? You're on Facebook, you see a blog post.

Well, maybe you know, nobody sees anything on Facebook anymore that aren't ads um and they're not seeing any of those either that's uh that's another show uh they go to read the blog post and and then they and they leave they go back to Facebook they go back to where they were so did your content do its job well yeah obviously if if it got somebody to read and hang out there um is it bad that it has a high bounce rate not necessarily the only place that we tell people that bounce rates are bad uh are on things like landing pages like if you're sending traffic to a page and it's a you know a page where you want them to fill out a form and do something and they leave and they don't go to the the thank you page then yeah that's that's bad. So the thing we want to take a look at there's two things one well look at this time on page time on page is a good indicator that somebody's actually hanging out so let's go ahead and sort through this is a podcast too so there's on each of these pages we have what do we have on these pages got uh the audio player we got the show notes and resources and stuff like that so there's enough there that you someone theoretically could hang out for 23 minutes if they didn't subscribe to the show some other way right yeah definitely you want to share this up too we're still in the two oh duh that would silly of me okay so here's the page um and you see there's the player etc uh and then the show notes so in Google Analytics uh we have our content, right? So behavior, site content and uh we have I've sorted here by time on page. Some of these pages, like this one here 29 uh 29 minute duration, right? So somebody went to that page and actually listened.

Like they they listened to the page. Yeah, they run the player. Yeah, yeah. Yep. And you see a lot of this, you know, 287, 28, 29 minutes.

So people are actually do listen to the show that way. So those are pages that have I would say have a high amount of engagement, right? Um conversely, uh, if we sort that column the other way, these are the pages that have like nobody listening to them. And these look like bots, right? There's a lot of query parameters, a lot of other stuff going on.

So I'm not too terribly concerned that uh there's there's zero uh zero time on page. So let's go ahead and sort back by page use. We have page use, unique page use, entrances, bounce rate, exit. Exit rate, I think is is something somewhat useful. Um exit rate's useful because it tells us uh when somebody is this a page that people leave the site on a lot.

Now there's some pages that you definitely want a high exit rate on, right? Like the thank you page. They've done something, it's okay for them to leave now. Um blog posts, you know, it's okay. Obviously, we'd love to stick around more.

Um, but if it's a blog post, that you know that that's fine. Um and then there are pages you want low exit rate on, like you don't want the thank you, you have the the lead form pages to have a high exit rate. I'm gonna sort by page value. So this is where when you put that dollar amount in as a goal, even if it's arbitrary, it's like you know, this goal completion is worth a dollar, even though that has no necessary uh bearing, it starts to try and infer the value of a page, right? It starts to try and say, okay, well, this page showed up a lot in the path to conversion.

And so we're gonna assign it uh an arbitrary dollar rate. So let's go ahead and uh just slap a graph on that. Let's go with page views, and let's go with. Oh, page value is not in here. I can't do that as a as a bar.

Lots thanks. Well, you know what? Let's export to Excel. That's the easy one. To the Swiss Army.

Yes. Exactly. It's funny. There's a lot of people who like to uh to crap on Excel for uh not being you know super sophisticated, but it gets the job done an awful lot of the time. Yeah, no, there's no um easier way to do it.

I think Google Docs is starting to come onto it, but I don't know, there's just such a rich tradition of people digging themselves out of holes with Excel that exactly. Okay, let's slap uh a nice data bar on this and sort. So now these here are the pages where they've done the most to nudge people towards conversion. So John, what do we see here? What are the what are we looking at?

Yeah, well, I'm really excited. It's kind of like you know, all shows in the last year are are doing well on that with the huge time on page. As far as the Yeah, so what the value is just because it's holding the most traffic for the highest amount of time, right? I mean, these are just the most engaging stuff. Well, so though the page value is that that page did a lot to to nudge someone towards conversion.

Yeah, yeah. So it's not surprising me that it's you know the most recent and best shows. That's totally what you would love to see is that the great content is actually making things happen. Exactly. Now, one of the things that we can look at here is um so you have page value and then you have the page views, right?

And you can see there's there actually doesn't look like there's a heck of a lot of a relationship between the views and the page value. So those pages that have um good value, but not necessarily a ton of views, those are pages that from a content perspective would probably want to you know run some ads to or make sure that we're sharing them on social media or put them in in the newsletter. How often does the newsletter go out? Yeah, monthly, you know. So that these would be the ones we would want to feature for the month.

Because yeah, that's a classic problem is you know, every month I go and I look at okay, I gotta pick you know one of the four, which one should we feature? And this you know would clearly tell me, hey, you know, the more offensive stuff tagline is the one to go with for the June newsletter, obviously. Yep. Now the other thing that to think about here is um are there pages that um have like a low exit rate um and but have you know low page views. Could we send more traffic to those pages?

So let's do this. Let's um let's do some filtering here. And I want to let's do some filters. So I want page views, uh I want to filter greater than five, right? So start with start with that just to knock out like the ones use and twosies.

Um and then page and exit rate. Let's go ahead and put a filter on that. And I want that to be uh less than 50%. Okay, so now these are pages that have you know decently uh you know decent retention. Let's go ahead and sort this smallest to largest, decent retention, um, but relatively low views.

So there's a couple here that are are really old, like really, really old, um, and a few others that are in the in the back issues. So one of the things that we might want to do is look through some of these older ones, like this one from Seth Coden, which is always going to be popular. Um, there's even posts about that in here uh and say, let's go ahead and uh re-feature those. Maybe uh something like a time hop section in the newsletter on this day, you know, this month, five years ago, this was uh a popular episode. If it's especially if it's a an interview that holds up well, might be worth uh doing something like that.

Yeah, it's funny. It's just it's great too. I think you can't over um ex you can't talk highly enough about that filtering right there because you like now all the bots are gone. You've got a whole different view of the data set here, like that's critical to getting to the mean. Exactly.

So that's a good first step um to figuring out what content's working and what's not, looking at things that are popular, looking at things that have low or high exit rates, low or high page value. Um especially here, those most popular pages, uh most popular episodes page, decent traffic, low exit rate, and it's got some page value, right? So that is a page that definitely needs to be, you know, at the top and bottom of every newsletter, um, on the social feed, you know, every every other day or something like that. Um same for welcome message, you know, if the uh when you set up tweets and stuff, um, you know, have it the daily welcome message. Hey, if you're a first time listener, make sure you go to the first time visitors page and check it out.

You could send some traffic to that and uh and really crank it up. Yeah, and the good news is that stuff's all built too. There's no extra lift to go and getting that stuff clean. Exactly. So that's our first stop.

Our second stop uh is Google Search Console. So again, this is one of those things where you want to have uh Search Consoles set up for your site. Um, do all the authentication things that you need to. And what we're gonna pay attention to here is we're gonna go to the performance section, and there's a lot of fun stuff in here, right? There's queries and pages and countries and all that stuff.

I'm gonna go ahead and just do an export of this. And again, we're gonna send this to Excel because uh Excel is a really good tool for this particular application. And what we get is exactly the same thing, but in tabular format now. So it's it's a lot less uh mind searing. Let's go ahead and just crank up the visibility here a bit.

That's cool. I didn't know that that dumped straight to tabs by default. That's great. I thought it was just single screen. So you get a ton of info just by kicking that there.

Exactly, exactly. So it does it as you choose Excel, it comes out as tabs. If you choose CSV, then you get uh zip file of individual pages. Okay, so what are we looking at here? Well, we've got ourselves our site URLs.

Oops. Oh bad. That was great. There we go. Um we've got the clicks, the impressions, the click through rate, and the position in search.

Obviously, uh, let's go ahead and just put a some conditional formatting on this to make this a little easier to see. Stretch this out. And we've got a lot of pages that are ones and zeros. Right. We've got our impression numbers here.

So some definitions. An impression here means that this page showed up in a search result. Um, so when somebody Googled for something, anything, this was a page that showed up. The click is it showed up in the result, and somebody clicked on that and said, okay, that was clearly you know, uh uh something that seemed to meet their their needs. And then the position is shows what the average ranking of that page is in in search results.

Um so for example, uh marketing over coffee. This this page here ranks tens rank about uh around position four, this ranks on position seven, and so on and so forth. And of course, the click through rate is nothing more than uh clicks divided by impressions. So let's do a uh again, let's do like we were doing earlier. We're gonna do some filtering here.

I'm going to uh filter out on clicks. Well, I go at least at least uh greater than one, so more than one click. That's gonna chop away a lot of the table. And so what we're left with are all the pages have gotten, you know, two clicks or more. And then what we want to do is I'm gonna sort by a click-through rate.

So our yeah, go ahead, expand the selection. So some of these pages, you know, this one here, this one here, 3.78%. That's a CMO interview. Uh, we've got uh a discussion with David Muir and Scott from way back when we've got the welcome to 230 info. So again, those pages that have decent uh clicks, it's like, okay, you know, there's we we could send some more traffic to them, maybe see if those those pages are are meeting intent.

What I think would be more interesting is if we sort this the opposite way. So what's got the lowest click-through rates? Um, and when we look, I'm gonna go ahead and put a conditional format on this. There are pages like this one here, Seth Godon on the practice, high impressions, low clicks, and a very low click through rate. That's a page that could probably use some optimization.

Let's go take a look at it. Okay. So what we've got here, people are going to this page for a reason. Um, however, the reason isn't necessarily clear other than it's about Seth Godin, right? Yeah, I mean, this is his latest, you know, book, so this is current stuff from him.

Okay. Um so what do you think, John? What should we do to fix this page? Yeah, I don't know. You know, it's funny as we're doing this too.

I was gonna pull it up in uh uh WordPress too, so I could see if um there was any feedback from Yostin there too. But I you know the big thing is probably to do a book summary right in the intro, you know, have more information about the book because that's obviously the poll for that, the practice. Right. Okay, let's see. Uh what we could do, I think, since we know this page has um high impressions, not a ton of clicks.

Maybe we can take uh a transcript, right, from from that episode and put it up here. So let's uh let's go ahead into our transcription software. This is clearly Seth himself, so let's tag him. Okay, and so we have now a nice transcript. I'm gonna go ahead and export this.

Let's export it as a text file. And now let's put in a little disclaimer here. Okay, well, it's machine generated, so typos are not the fault of any human. Typos are virtually guaranteed. All right, let's open up our text file here.

Well, let's take a look at our stats here so far. Uh, we've got a focus keyword, Seth Godin, right? Um readability is looks fine. Uh how many words is this page right now? So your key phrase is in general, the first paragraph, um, text length 258 words, minimum of 300 words.

Okay. I think we're gonna change, we're gonna bring in this transcript. And now let's hit update. So we went from 258 words on a page. So we can figure out where they are in the show.

Right, they can jump if they want to. And now it contains 9,124. And obviously, there's going to be a lot of uh additional words and phrases in this page now that will really help um with search. So if we go back to the post now, we've got a machine generated transcript. There's a lot in here, and because of the way we've uh done the transcript, you know, your name, but more importantly, Seth's name show up a lot throughout the transcript.

So if we're optimizing this page for this his name, we've now put a lot more uh uh opportunities for uh bot to look at this and try to understand what's in here. So that's a page where good impressions, not a lot of clicks. Um, hopefully we can get um we can boost its performance. So we want to take the metrics that come right out of Search Console and slice and dice them. Um again, this is not uh rocket surgery here, right?

This is literally just sorting and filtering in Excel. We haven't even done a V lookup yet. Um probably don't even need to. One of the questions that folks have is well, how do I know what I'm missing? Right?

How do I know what I might what kind of content gaps I have? That we answer in the queries tab here, which again is right out of what's on the website. We can look at this, right? These are the these are the clicks that we get. Uh so we see a ton of clicks, marketing over coffee podcasts, you know, doing well, marketing over coffee, doing well.

Uh then you start getting into obviously different guest names, um, things there. Marketing podcast is in there. So our clicks are are useful. What I'm gonna do is though, is I'm gonna sort this um by impressions, and then let's highlight our click through rates here. And what we've got now, these are the things that we get in we show up in search for a lot, um, but we're not necessarily getting great uh clicks.

In fact, let's go ahead and slap a filter in place here, and let's filter on click-through rate of less than five percent. And let's put a little data bar just to visualize a little more easily. So these are the terms now that we have very low click-through rates on, right? But we show up a lot. So again, these are terms that you know, influence marketing.

We got 1,100 impressions for the term influence marketing, but not a single person click through. If we're doing a podcast about marketing, we might want to do a show specific about influence marketing. Have we done one? I don't remember if we have. No, as a matter of fact, we've got um folks from Upfluence coming up uh in the next two weeks.

So that's like right on the on the docket. So we get lucky with that. But yeah, we should be doing more in that space. I'm just surprised to see some of these terms. And it's crazy too how positions they're they go as deep as a hundred.

I mean, nobody's ever looking at the 10th page, and yet you still you've got well, they are looking because there's 608 impressions there for newsletter marketing and 100 at the hundredth position. Exactly. Um yeah, in fact, let's let's um clear that and let's do color scale green to red. Some of these things, like this one here, obviously your name, uh, position five, which is good. Um is in there, uh coffee marketing, eating their lunch.

Uh we've got some entertaining ones in here, but there are a decent number of terms where you're right, we are showing up for like chat marketing. So, you know, who do we know anybody who'd want to come on and talk about that particular topic? Um, again, 900 impressions, not a single click on it. Yeah, no, it's it's crazy to see how some of those topics just don't pull at all. Yep.

And that's probably because we don't necessarily have a show for them. Um, some of them that we do. So, like for example, we have Simon Sinek, um, you know, he's been on the show a couple of times now, right? Yeah, yeah. It's you know, he's deep in the library and pulls.

I you know, the challenge with all that is we are competing against you know his own content and somebody like him who's putting out you know tons of videos and audio stuff all the time. It's a huge challenge to try and crack through. Exactly. But we've got the content, so as we did with Seth, we might want to take uh the the top names here uh and make sure that at least that the their pages, their episodes have transcripts to go with them, just uh to to give the engines a little something more to grab on to. More to chew on.

He's like, What's a quorum marketing podcast? Hmm, what that means. Let's uh weird. I didn't know they did anything. Well, they may not.

Let's take a look and see what comes up in search for that. Put an incognito tab here. What are some good podcasts on marketing? So a question. Marketing over coffee, number three.

That would be why we're showing up and uh the growth with core podcast and stuff. So there are uh some things in here. How do you score for marketing? Um it makes sense that if we're being mentioned in there, uh, we might want to go necro some of those threads, uh you know, comment on them 10 years later. Hey, just seeing this, thanks for mentioning.

Yeah, right. Thanks. I'll have to check it out. It's great. Um but from a content gap perspective, this is a great way to feed to find your content gaps.

To say, okay, what do we what do we what does Google semantically think uh we might be relevant for? And people just aren't clicking uh on us. They're just they're just not seeing anything of relevant irrelevant in the search results. It could be, you know, in some cases, because yeah, you're on page eight of the search results, uh and you know, the likelihood of you doing well on this is pretty low, or it might just be that there isn't matching content. And for us, uh something like you know, modern marketing and newsletter marketing.

We should definitely do a show on newsletter marketing. Adding to the to-do as we go here. Exactly. Um, those are some of the the basic ways that you can start tying some of this uh data together. And again, just going through and looking, like what are the pages where you're getting clicks, you're getting impressions and things like that.

Uh and you can, if you want, if you were so inclined, uh, take this data, you'd have to do some manipulation, but you could do um V lookups, for example, and match it. So you have Google Analytics data right next to your Google Search Console data. See what are the pages they're getting impressions, but you know, you're not seeing actual traffic for it, or you're getting traffic for it, but you know it's not coming from search. So, where else are those pages getting traffic from? Um start optimizing it.

So the there's two types of optimization. There's um traffic optimization and then there's content optimization. Um for content optimization, those are the pages that like we saw that have have very good conversion ratios, like they they they have good click-through rates and things, but bad traffic numbers. We want to send more traffic to those pages. There's the reverse, which are pages which have good impression data, but bad traffic, which means that um we uh we may not have done a good enough job optimizing the content itself.

So we need to spend some more time transcripts, tuning things up. Maybe even, you know, I don't know, change the dates on the page or something just to liven things up. Freshen them up. Um let's take a look here and the other software and see what other recommendations are are made for this page as being a relatively new page. So we don't have a keyword phrase uh set, uh meta description all image alt attributes.

So some of these things are things that you could fix with plugins too. Just go in and say, okay, I want to make sure that every every image at least has a you know a default alt attribute of, you know, this is an image. Um making sure that uh the red things, key phrase length of not having one fix those things pretty easily. Okay, so this one would be creative. Yep.

I don't need I don't need related for you so I just need to set that phrase. There we go. Okay, so we've got our phrase set and look at our analysis. Key phrase is not in the first paragraph. Uh density is a little on the low side, title width is too long, description is too short, phrase is not in the title.

So we got some work to do. Um yeah, and you know that as I'm thinking of this, and we we see this, it's like it's the transcript should just become the default. I should just get a VA and give them you know 10 bucks a week to make sure that the show is done every week and just get that in there because the the volume of text is just you know, it makes a show notes are nothing compared to transcripts. Exactly and with the timestamps in them I mean you can even do some light formatting on them. So for example here are the last 600 some episodes of the show and just stuff them all into transcription software um otters transcription software and now we can go through and again either clean things up but ideally export it in such a way that it has um that that has some decent formatting to it.

So there are ways you can write scripts that will take this text and then you know for example make all the names you know like uh uh heading to uh headings within the HTML again gives this the struct the page some more structure and helps um the bots understand what the page is about all right so to recap um we've taken a look at uh how to find optimizations in your Google analytics data and your Google Search Consultator these are two very very easy uh simple sources to use they're free um and showed um the some ways to manipulate that in terms of timing um this kind of optimization ideally you do it um quarterly would be good um if you're if you start seeing substantial declines in your traffic particularly from organic search uh you may want to increase the cadence maybe do it monthly uh I know some folks who are like just hardcore will do it weekly but at the very least a quarterly audit go in see what's working see what's not you can later on uh once you've done the basics start looking at more uh complex uh ways to do this analysis. So one of the ways that we talk about uh doing this analysis a lot is using machine learning uh and doing uh uh what's called Markov chain modeling, which allows you to take a uh essentially a look at what are the pages that convert the most. Um so now instead of looking at traffic data, we're looking at conversion data. What are the pages that you get the highest number of conversions? Uh, this is back from February.

I'll probably should re redo this at some point. Um, but again, pages that have a very uh high conversion rate. We want to send a lot of traffic to them uh as much as we can. Pages have uh a high conversion efficiency, meaning that it takes very few visits to convert. We want to make sure that those pages are well optimized.

Um question from Brian: Do we have export and backup or Search Console data, just pull historical data when we need it? Search Console now gives you 13 months of data look back. Um so uh I find that it used to be 90 days, and yeah, you'd be scraping it every 90 days to try and store it. But now it's a 13-month look back. I think that's enough.

I mean, because these days uh in the world we live in now, anything more than 13 months is practically a different planet. So I don't know that I need to go that much further back. It's not a bad idea to have the historical data archived somewhere, maybe. Uh, if you want to do that, I would suggest doing it using um uh the Search Console API and having a piece of code do it for you automatically. That way you don't have to remember to do that.

Um, any any uh parting thoughts, John, about uh ways we should be optimizing our our poor uh shoeless child. Yeah, more transcripts is the biggest thing. Especially I noticed Aaron Cliff's interview there, CMO of Waze. That has been a really popular interview. So yeah, that's another one I gotta slap that transcript on and tune that up to go.

Exactly. And one other thing that to think about is if you've got regular like blog content that doesn't have rich media, um, you may want to start doing rich media just to give you stuff um to feed your pages. I know um it was it's funny. About a uh six months ago I stopped doing a daily YouTube show, and I'm thinking about going back to it, not because my YouTube channel is doing particularly well, it wasn't, um, but because it's easier for me to talk for 10 minutes than it is to sit there and agonize over like word choice for for 45 minutes to write the same amount of content. And uh I am starting to see some of those older shows actually getting uh attention.

Um, so I'm thinking I may go back to doing a daily show because every time I open my mouth for 10 minutes, it's about 1,500 words. Yeah, yeah, yeah. That's a huge amount of writing for anybody that's trying to sweat through a manuscript. That's just crazy. Exactly.

So all right. Well, so that is gonna do it for this week's uh show. Um, if you got additional questions and things and you we didn't get a chance to chat about it uh here, go on over to the Slack group, uh Analytics for Marketers. We can uh we can have a discussion there. But uh, we will see you next week.

And thanks for watching today. Be sure to subscribe to our show wherever you're watching it for more resources and to learn more, check out the Trust Insights Podcast at TrustInsights.ai slash T I podcast and a weekly email newsletter at trustInsights.ai slash newsletter. Got questions about what you saw in today's episode? Join our free Analytics for Marketers Slack group at trust insights.ai/slash analytics for marketers. See you next time.