S02E10: Making analytics work for you

This is a transcript of episode 10, season 2 of the boagworld podcast: Making analytics work for you

Paul Boag:  Hello and welcome to Boagworld.com, the podcast for those involved in designing, developing, and running websites on a daily basis.  My name’s Paul Boag.

Marcus Lillington:  And my name is Marcus Lillington.

Paul:  And I don’t want to be here.

Marcus:  Where do you want to be?  Jetting around the world like you were last week?

Paul:  No, I want to be curled up in bed with some hot chocolate and watching episodes of Fringe.  That’s what I want to be doing right now.

Marcus:  New series of Fringe; missed the first one.

Paul:  Ahh.

Marcus:  What an idiot.  No wonder you really need to see.

Paul:  I haven’t seen it, so don’t tell me anything about season four.

Marcus:  I haven’t seen any of them, I’ve recorded episode two, and it was the last one.

Paul:  Right.

Marcus:  I’m thinking, “I’m not going to miss episode one.”  I’m going to try – go back onto Sky Go, that’s a laugh on an iPad.

Paul:  Yeah, oh no.

Marcus:  I’d like to find this thing – nope, sorry.  There’s no way to search Sky, on your app.

Paul:  Oh, it’s useless, absolutely useless.

Marcus:  So what I need to do is pay for it.

Paul:  But I don’t know whether they have all of that stuff on, do they?

Marcus:  Yeah, you can go back and search on episodes and well, it says, “Watch now” and I click on it and then it says, “This is a subscribe only service” or whatever. And I think, “Well I have subscribed for this.”

Paul:  Why don’t you just BitTorrent it?

Marcus:  Well yeah.

Paul:  It would be a lot easier.

Marcus:  Yeah, but I’ll need to see episode one.  We’ve gone off on an instant tangent here; a Fringe tangent.

Paul:  That’s sad.  Yeah, well I think it’s a good series.

Marcus:  I do like Olivia.  What’s wrong with that?

Paul:  Nothing’s wrong with that, she’s a very nice person.  I’m sure it’s her personality you’re saying you like.

Marcus:  I think that she’s really hard, as well.

Paul:  Argh, don’t mess with her.

As you know I was in the States last week and on the way out I was watching those Fringe episodes because I had kind of given up halfway through season three.

Marcus:  Oh no, I’ve seen them all.

Paul:  And now I’ve caught up and last night I finished season three with the big cliffhanger.

Marcus:  Very exciting.

Paul:  So now I’m massively jetlagged and not in the mood to do a podcast at all and on top of which is a subject that I find very confusing, which is analytics. I don’t get analytics.  Isn’t it funny how you get some people who are just really good with numbers and figures and it’s like whenever we sit in our board meetings and Brian our executive director comes in and he opens up the spreadsheets and he just kind of picks out all these really obscure things.

Marcus:  But that’s just experience over the years.  He’s a director of many different companies, so he has to come in and point out one mistake, so he’s spending the whole of the first probably just going, “Where is it, where is it, where is it?”

Paul:  Right.

Marcus:  “There it is, I’ve got a mistake.”

Paul:  Right, now I know what he’s doing.  But people just can see numbers and stuff.  Now, I’m just so bad with analytics and that kind of thing.

Marcus:  You can make them do whatever you want though, can’t you?

Paul:  Well yeah, that’s the trouble.  So anyway, this is going to be an interesting show but I hope you’ll bare with us, because it is an important one and we are getting near, obviously, to the end of the season.

Marcus:  How many shows are we doing, Paul?

Paul:  We’ve got this one which is on analytics, we’ve got next week which is on bug testing and all that kind of stuff and release.  So in theory, after this show in two weeks’ time we’re going to be releasing the website.  Hence the e-mail I sent Rand yesterday saying, “Please don’t ask me to do anything else; I must finish the Boagworld website.”

Marcus:  So what’s episode 12 then, because I’ve got an invite to that?

Paul:  So then we’re going to launch, we’ll launch the site at 11 and then episode 12 is going to be kind of marketing community and leveraging your community.  Which essentially is me saying…

Marcus:  You didn’t say, “leveraging” did you?

Paul:  I did.

Marcus:  Because I might have to shoot you.

Paul:  Yeah, I’m sorry, that was painful.  Essentially that was me begging the audience.

Marcus:  We might be able to find some synergy with some other…

Paul:  Oh, shut your face.  So essentially the episode is going to be me begging the audience to tell everybody about how cool the site is.

Marcus:  Oh, okay.

Paul:  But I’ll dress it up and make it sounds more professional.

Marcus:  And then we’re going to do a Christmas special.

Paul:  Then we’ll do a Christmas special.

Marcus:  And then we’ll have a week off.

Paul:  And then I’m aiming if all be good to come back in April.

Marcus:  Okay.

Paul:  With season three, which I’m making this up as I go along.

Marcus:  It’s great isn’t it?  Yeah.  And you give us some titles of the first few episodes?

Paul:  What, of season three?

Marcus:  Yeah.

Paul:  The podcast itself is going to be on client-centric web design.  How to put the client first and working with clients, the whole season, which I think will be a really good one.  Because I think our approach – it was really interesting being out with our client in the States last week.  One of the things that she said to me is the reason that they chose Headscape is they can’t find any other agencies that work like we do with clients.

Marcus:  Really?

Paul:  Yeah.

Marcus:  That’s odd.  They cannot find other agencies that actually tell their perspective clients, that’s probably what it is.

Paul:  Yeah, it’s not that they don’t…

Marcus:  It’s communication.

Paul:  Yes, so that is what next season is going to be about.  It’s going to be about basically giving away our unique selling point

Marcus:  No, you can’t do that.

Paul:  Which is probably a bad idea, but anyway.

Marcus:  So, next season is going to be about coding.

Paul:  No, it’s so not.  Unless you want to run that season, which would be really interesting.  I’d like to see that, that would be good.

Anyway, Google Analytics.

Marcus:  I had something else to say.

Paul:  Oh, for crying out loud.

Marcus:  And I can’t remember what it is now.

Paul:  Is it really good?

Marcus:  No.

Paul:  Well you’ll probably think of it halfway through and interrupt me in mid flow like you normally do.

Marcus:  Okay.  Well then.  Analytics.

Paul:  Google Analytics, so yeah, so I’m rubbish at Google Analytics.  To be honest, I think I’m pretty much like most other website owners on the planet, which is I have Google Analytics installed but I don’t really have a clue of how to use it effectively.

Marcus:  It’s a bit like a bank account, I think, or the way I view my bank account.  I don’t really want to know where I’m at, you know what I mean?  I don’t, I just want to carry on, fa la la.

Paul:  Oh God Marcus, you scare the crap out of me.  I’m so glad it’s Chris who does the finances and not you.

Marcus:  Headscape accounts, that’s different.  Personal ones, that’s what I’m talking about.  I just want to know that I’m getting that kind of money and generally not spending more than that in the month and just carry on.  It’s depressing to keep looking at how much money you haven’t got.

Paul:  I’m really worried that you just compared this to Google Analytics.  Ignore Marcus.  If you’d listen to season one of Boagworld  you would know that my approach – I won’t say Headscape’s approach because obviously Marcus lives in a different world – approach to web design is built around the idea that websites should generate a return on investment.  Marcus, on the other hand, doesn’t seem to care.

Marcus:  No, no, no, no, you’re not understanding me.  I don’t want to see whether I’m overdrawn or not.

Paul:  So therefore, how do you know whether to stretch…

Marcus:  I have to make money.  That’s the point; it makes you make more money.

Paul:  Because you don’t know.

Marcus:  Because you don’t know.

Paul:  No, it makes you more money knowing, does it not?

Marcus:  No, no, no, see this is personal finances, very personal.

Paul:  Yeah, but you used this as an analogy for Google Analytics.  So it doesn’t stand up as an analogy.  I’m in such a grumpy mood tonight, don’t mess with me.

Marcus:  A lot of people have Google Analytics installed but don’t look at it.  That was the analogy I was trying to make.

Paul:  Oh, okay.

Marcus:  They’ve got all the tools there, but actually they don’t really want to know if they’re doing well or not.

Paul:  I don’t think it is that.  I don’t think that’s right.  It’s not that they don’t want to know.

Marcus:  They can’t be bothered.

Paul:  It’s that it’s too complicated.

Marcus:  Really?

Paul:  Yeah, okay.  Google Analytics isn’t complicated as an application; my point is they don’t know what to look at.

Marcus:  Sure.

Paul:  What’s the number you should be looking at?

Marcus:  It’s hits, isn’t it?  Homepage hits.

Paul:  Oh God.  Well that’s sorted then.  Thank you very much for listening and we’ll talk to you again in two weeks’ time.

Marcus:  Bye.  I suppose I have to carry on.  You can’t just walk away, even though you want to.  How many pages have we got?

Paul:  We’ve got bloody pages of it.

Marcus:  We haven’t done the first paragraph yet.

Paul:  No, we’ve got an interview halfway through as well, which we haven’t prerecorded.

Marcus:  I’m so looking forward to this.

Paul:  This is going to be great.  We’ve got Chris coming in.

Marcus:  Yeah.

Paul:  Chris who hates talking on the phone,

Marcus:  Every single person who’s joined us for our show has joined from start to finish, but Chris won’t.  So we can really set him up.

Paul:  Right.  So yeah, I was talking about season one, how it was all about generating a return on investment.  But to achieve that, if you want to generate a return on investment on your website, you need to be able to measure the successes or failures.  So kind of, I know in my head that Google Analytics and Web Analytics is important, but knowing it in your head is one thing and doing it is another.  I think the problem is, I’ve already said, is that most of us don’t really understand Analytics particularly well and we’re faced with a huge amount of data and we’re not sure what’s useful and what’s not.

When it comes to something like Google Analytics most of us never get past the kind of default reports on statistics such as page views, user sessions, and bounce rates.  And although this information is useful, I don’t think it gives much of an insight into the return our website is providing.  We need to dig deeper for Analytics to be truly useful, which is something we’ve covered on the show before.

We’ve discussed it before on a couple of occasions we’ve had a guy called Matt Curry who’s a real expert on analytics to explain all that Google Analytics has to offer.  And although I’ve found these really interesting and really enlightening when we did them, when I actually came down to think about analytics for Boagworld, I wasn’t really any wiser as to what to measure.  And it’s not because Matt’s advice was bad, it’s just that it’s probably too complicated for my site and probably for a lot of you listening to this.  It was perfect for an e-commerce site, which is his background, but I struggle to translate some of his ideas to a content oriented site such as Boagworld.

I’ve therefore decided to approach the problem in a completely different way.  Instead of allowing my usage of Google Analytics to be tainted by the tool itself, I’ve stepped back and asked myself what exactly I want to know, right?

Marcus:  Well yeah, that’s got to be…which, it’s actually, you say it’s more complicated what Matt does, but what you want to know is easier.

Paul:  Yeah, yeah, yeah, yeah.  Yes, absolutely.

Marcus:  This is when it starts to become a little bit more difficult.

Paul:  Yeah.

Marcus:  What would be useful?

Paul:  Exactly.

Marcus:  Yeah, you can find out stuff about, or you might be able to push sort of more people looking at that particular, clicking on that button but does that really matter?

Paul:  Yeah, do I really care?  So what I did was a sat down and I looked at my success criteria from the very first episode of the season, which were to create an increase in number of people contacting…

Marcus:  That’s an accessibility issue here, Paul.

Paul:  He’s complaining there because it’s light gray text.

Marcus:  My eyes are going now that I’m getting old.

Paul:  Yes, but this was printed out by Biword and for some reason the stuff that it quotes it – oh, it doesn’t matter.  It won’t look like this on the website, is the point.

Yes, to create an increase in the number of people contacting me about work, to create an increase in the number of people visiting the Headscape website, to create an increase in the number of Tweets and links back to Boagworld content, to create an increase in the number of people following other Headscape employees via Twitter or their personal blogs, to create an increase in the number of people talking about Headscape via social media and direct links, to create an increase in the sales of supplementary material.

Marcus:  Yes.  Good place to start.

Paul:  So, yes, it seemed like a good place to start thinking about Analytics.  Some of those things, things like mentions on Twitter are going to need to be tracked outside of something like Google Analytics, but a lot of them can be monitored within the application if it’s set up correctly.  And I decided to focus on two things, right?  I could have focused on a lot of those but I decided if I narrowed it down then I’m more likely to monitor it regularly.  Do you see what I mean?

Because the trouble is if you try to monitor lots of things then it gets more complicated to look and you stop doing it and there are too many different variables, so I thought I’m going to narrow it down to absolute basics, which is people visiting the Headscape website from Boagworld.  I care about that, how many people are visiting that site.  And I care about people subscribing to the blog.  Those are the two core things that I want to achieve.

Marcus:  I agree, it would be good to keep an eye on mentions of Headscape.  We really ought to have a Headscape Twitter account, shouldn’t we?

Paul:   Well, I’ve got a plan for that and we do come on to that later, because this is what I’m going to start with and then when I get confident and I’m doing this regularly.  It’s like the old adage, isn’t it?  If you want to break a habit it takes 40 days or whatever they say to break the habit.  The worst thing you can do is try and break two habits at the same time.

Marcus:  Yeah.

Paul:  Or set two new habits at the same time, it’s just too difficult to do – one thing at a time.  So I’m setting really small, simple aims to begin with and then I’m going to hopefully, once I’ve got that as habit, then I’m hopefully going to build on that over time.  That’s the kind of logic that I’m going for.

Marcus:  Yeah, makes sense.

Paul:  Right.  But that said, I’m not just interested in the number of people visiting the Headscape website or the number of people subscribing to the blog, I also care about those people.  What is it about them, for example were users more likely to subscribe after reading certain types of blog posts?  Or were users coming for a particular referral source more likely to go on and visit the Headscape website?  So if I don’t understand why those users are completing the call to action then there isn’t a lot of use.  If I don’t understand why they’re visiting the Headscape website then I can’t focus more on that.  If I don’t understand why people are subscribing then I can’t give them more of what they want in order to subscribe.  Does that make sense?

Marcus:  Yes, it does.  I think we should just force them to fill in lots of form fields.

Paul:  In order to tell us in detail.

Marcus:  Yeah, yeah.

Paul:  Yeah, that always works well.  Yeah, before you can visit the Headscape website and buy our services, please tell us how dogs you have and your inside leg measurement.  Yeah, that’s works.  No, we can’t do that but I think there are things that we can glean, hopefully.  This is where we need to talk to Chris.

Marcus:  What would be interesting, what would be really interesting…

Paul:  Yeah.

Marcus:  I’m making this more complicated.  Is to track people, what they do once they get to the Headscape website.

Paul:  Yes.

Marcus:  These people who came from here end up looking at this on here.

Paul:  Yeah, and that is my ultimate intention because we can obviously set up stuff on the Headscape website as well.  And we can then track any refers coming from Boagworld all the way through.  But, little steps.  I want this to become part of my psyche and way of thinking about a blog, and I know if I do too much I’m not going to do it.

So, it’s not enough to watch certain metrics go down; we’ve got to understand why users are doing what they’re doing.  So hence we need some information on people that are completing these two primary calls to action.  Right, so setting up the basics.   Alright, so I’ve now got a clear idea on what I want to achieve, and I know that being involved, being able to track the clicks on links leading to the Headscape website or track clicks on buttons that allow users to subscribe.  I also knew that it’s going to involve setting up some custom reports within Google Analytics to allow me to better understand why users complete certain calls to action.

Marcus:  How do you do that, Paul?

Paul:  I don’t know.  That is my problem, and so that’s where I wanted to get Chris on the show because he seems to be a numbers kind of guy and know about that kind of stuff.  Now I did consider tricking him and just calling him up and actually talking to him over the phone.

Marcus:  That probably would have been a better idea.

Paul:  It would have been a better idea in just getting him to talk through stuff.  In the end I decided that was a bit deceitful and that I shouldn’t record him without his knowledge, so we’re going to get him on the show to talk about it.  So, here is Chris.

Okay Chris, we were just saying there are a couple of things that we want to achieve, right?  I’ve kind of narrowed it right down in order to make my life easier because I’m worried that if I try and look at too many things I won’t look at anything, you know what I’m like.

Chris:  Yeah.

Paul:  So, the two things that I want to look at are people visiting the Headscape website, which is fairly obvious, okay?  And people subscribing to the blog.  So on the new design obviously in various places we’ve got links to cross to the Headscape website and we’ve also got some big, strong calls to action which is the subscribe buttons.

Chris:  Yeah, okay.

Paul:  So, I’m guessing we need event triggers or whatever they are.  You put triggers on particular elements, is that right?

Chris:  You can, but just start off simple with the things that Google Analytics does out of the box.

Paul:  Yeah, this is the problem.  This is what we’ve just been saying, I tend to get distracted by page views and bounce rates and all these default things.

Chris:  Yeah, and we should come on to those, but I think just start off with looking at the visitors you’ve got for the site and looking at the behavior of different types of visitor, yeah.  And looking at visitors to the blog and what they’re doing, and then work on to where you might want to use event triggers later.

Paul:  Right, because obviously I’m kind of working the other way around.  You’re coming in from the type of users coming in, I’m working backwards.  The people I care about are the people who subscribe to the blog and visit Headscape website, I want to work backwards and understand those people.

Chris:  Right, okay.

Paul:  But you could do it the other way around, that’s fine, it doesn’t make any difference.

Chris:  If you want to look at people who subscribe to the blog.

Paul:  Yes, why do they subscribe, what do they read?

Chris:  You can pick those up if you’ve got a thank you page on the subscription form.

Paul:  Yes.

Chris:  You can pick them up at the point that they see the thank you page.  And you don’t need an event trigger to do that.

Paul:  But I don’t have a thank you page.

Chris:  Right.

Paul:  I have an event trigger is what I’m going to have to have, because it’s all Javascript driven.

Chris:  Okay, then you’ll need an event trigger.

Paul:  Yes, I thought I did.

Chris:  Why haven’t you gotten a thank you page?  Thank you pages are great for analytics because it gives you a nice, clean signal, you’re not messing around.

Paul: yeah but they’re rubbish for user experience. Open up a new browser.  So if you click on that subscribe button at the top.

Marcus:  That gently throbbing subscribe button.

Paul:  Shut up.  It pulls down a list of all the different things, right?

Chris:  Yes.

Paul:  Now, you click on any of those – well, they won’t do anything at the moment, but it’s not a thank you page scenario because the e-mail will open up a mail window, the RSS will go to an RSS feed, it will go to iTunes or whatever else.  So I can’t do thank you pages.

Chris:  Okay.

Paul:  Does that completely screw things up?

Chris:  No, no it doesn’t.  It just makes it a bit messier in Analytics.

Paul:  Right, yes that’s the story of my life.

Chris:  And it means that someone will need to put the code in the right place.

Paul:  Well that’s okay.

Chris:  That’s okay.  And then wire things up in Google Analytics, which is fairly straightforward.

Paul:  Right, that’s the big thing.  Can you talk me through how to do that, because that’s what I don’t understand?  I get the principle, I know event triggers exist but I’ve never actually added them, so how do I go about adding them?

Chris:  Okay, so we look in content.

Paul:  You’re making this up as you go along, aren’t you?

Chris:  Yes.

Marcus:  That’s okay, makes great media.

Paul:  They just don’t change interface, is that the problem?

Chris:  They have just changed the interface, and it’s actually easier to find things.  Because I can see – when I click on content, there it is – events, you see?

Paul:  Oh yes, there it is.  Very pretty.

Chris:  And we’ve got an overview of events on this particular site that I’m looking at, and there we have number of events and here for example we’ve got an event which is something to do with a basket on a shopping site.

Paul:  Oh, you’re looking at another website, right?

Chris:  Yeah,  yeah.

Paul:  Okay, so they’ve added an event.  So it must be, with the events trigger, basically you must in the code you put and give the event a name, which then I presume makes it appear in Analytics?

Chris:  Correct, yeah you do

Paul:  So I’ve got to copy and paste some code across, I can do that.  Except I’m going to have inline Javascript, that’s not nice.

Chris:   You are.

Paul:  Oh well.

Chris:  Yep, no way of avoiding that.

Paul:  There might be.  I’ll have a look.

Chris:  Something else that’s good to do is to get all your search results into Google, so you’ve got those there as well.

Paul:  You mean the results from your internal search engine?

Chris:  Yes.

Paul:  Yeah, that’s one of the things I wanted to do, because the good thing about that is that then people are suggesting topics for you to write about.  If they search for something and it brings up no results, then you think, “Oh, that’s worth writing about.”

Chris:  I’ll come back to events.

Paul:  Yeah, yeah, fine.

Chris:  So look, site search, it’s all wired up for this site.

Paul:  Cool.

Chris:  So you can see here that on this ecommerce site, the users of the site are using product codes as their main search criteria in search.  So if we look at full report then we can see that the top 10 are all just using product codes, not product names.

Paul:  It tells you loads about the way that people operate on that particular site.

Chris:  Yeah.

Paul:  Jeremy McGovern wrote some really good stuff about what you can glean from internal search engines.  Wow, loads of it are codes.

Chris:  So on this site we have to get down to 153 in order to get a text search.

Marcus:  That tells you an awful lot.

Paul:  That does tell you a heck of a lot.

Marcus:  It tells you that basically new visitors to the site do not use search, it’s only repeat visitors.

Paul:  Who know their code.

Chris:  They know their codes.

Paul:  But from my point of view it would be brilliant because I’m hoping that it’ll suggest topics to write about.

Chris:  Yeah, yeah.

Paul:  Which is always the problem when you’re blogging, is trying to think of what to write next.

Chris:   So, setting up event tracking.

Paul:  Right, back to event tracking.

Chris:  So you can label it in various ways.

Paul:  Ah, I see.  So you’ve got variables, basically that you’re putting into the Javascript.

Chris:  You’ve got all sorts of variables.  And so you can code those obviously to catch the name of the page or whatever it is, you can see categorized events in the Analytics user interface.

Paul:  Because what it would be quite useful for me to do, because obviously I link to the Headscape website quite often.  There probably is a way of me running a bit of Javascript that will, whenever it sees the word Headscape in a blog post it automatically attaches the event handler to it.  I should be able to do that.

Chris:  Yeah.

Paul:  Which would be really useful because it means I won’t have to add every single time, which would get really annoying.

Chris:  Yep, yep.

Paul:  Cool.  So that’s good.  So once you’ve got your event handlers in, how do those that appear within the Anayltics?

Chris:  So, they just automatically appear under the…

Paul:  Event tracking?

Chris:  Under the event tracking and if you’ve got any events they’ll show you categorized lists.  I mean, on this side we’re looking at here there’s only one event type that’s being tracked, but those would all appear there and you can just drill down into the categories.

Paul:  Okay.  So, obviously the next thing which is what you were saying right at the beginning is that it’s all well and good knowing how many people have completed an event, like go to the Headscape website or subscribe to the blog, but if you want to encourage more people to do it you need to know a bit about why the people that did do it, did it.  Is that even a sentence?

Chris:  Yes.

Paul:  Yes, it is a sentence.  But you know what I mean?  So I need a bit more information on those people and why they do what they do.  So are you able to kind of create a custom report – I’m making this up now – custom reports or something like that that enables you to find out a bit more information about the people who completed those events?

Chris:  Yes, the really, really good thing about Google Analytics is something called Advanced Segments .

Paul:  Right, that sounds advanced.

Marcus:  Matt Curry spoke about segments last time.

Paul:  He did, but I didn’t…

Marcus:  You weren’t listening.

Paul:  I was listening, but I didn’t fully get it.  The great thing about this time around is I’m actually applying it to a real project on a real site which kind of helps you get your head around it a bit more.

Chris:  Yeah.  What an advanced segment enables you to do is segment your users.

Paul:  No!

Chris:  According to things they do or places they’ve come from or technology they’re using in all sorts of ways.  So there may be a combination of pages that you’re interested in for the user session, and then you want to constrain your analyses to just users who, for example, load up a basket and don’t check out.

Paul:  Right.  So, my segment is actually really simple – just people who have clicked on a certain link or whatever.

Chris:  Correct, correct.  And so you can build, and the tools are really easy to use.  They’re drag and drop tools.

Paul:  Are they?  Show me.  I know nobody else can see this, I will describe for the audio podcast what is happening as it happens.

Marcus:  You can even have a segment of people who’ve gotten a new MacBook Air and still want to use IE.

Paul:  Are we talking about Chris?

Chris:  So, if you create a new customized segment…

Paul:  I can work that bit out, that’s good.  It really is drag and drop, isn’t it?

Chris:  It really is drag and drop.

Paul:  Ooh, look at that, like that.

Chris:  So we might want to look at users who come from a particular country.  The condition is “matches exactly” or “does not match exactly” or “contains”, you know there are all sorts of possible values.  Let’s say we’re interested in users who come from the United Kingdom.

Paul:  Oh, okay.  So essentially what you’re doing, just to describe this to everybody else – you open up Google Analytics, go into Advanced Segments, was it?

Chris:  Advanced Segments.

Paul:  Which was there on the bottom of the left hand menu.  Then you create a new segment.  What it gives you is essentially a list of all the different dimensions which are things like visitor information, content information, all this kind of stuff.  Then you just drag them into a right hand panel, almost like the kind of search you’d get, I’m trying to think where else you’d get search like that.  It’s like creating advanced search, isn’t it really?

Chris:  It is.

Paul:  And you can just drag them on.  Cool.

Chris:  Yeah, so we’ve specified country/territory matches exactly, United Kingdom.  And we might want people who come from that country.

Paul:  And you can do it with either an and or an or statement there.  So you’ve gone people from the United Kingdom and page…

Chris:  Page matches exactly basket, for example.

Paul:  Right, so that’s a particular page on the website.  Yeah, I get the idea.

Chris:  Yeah.  And then you give that segment a name.  UK Baskets.

Paul:  This is so easy.  Why was I scared about this?

Chris:  Create the segment.  And then we can view the stats for that segment.

Paul:  Oh, ok.

Chris: by clicking on the name of the segment.

Paul:  And it will give you all the same stats you get, yeah it’s giving you the same data.

Chris:  The same dashboard, everything.  But it’s just a strain.

Paul:  Just narrowed down.

Marcus:  It’s a segment.

Paul:  It’s a segment!

Paul:  It’s an advanced segment.

Chris:  And then you can compare.

Paul:  Ah, so you can compare that to all visitors.

Marcus:  Can you compare with other segments?

Chris:  You can compare with other segments – up to four at any one time.

Marcus:  Okay.

Paul:  That’s brilliant.  That pretty much – I’m finished now.

Chris:  It’s the answer to everything.  You need to think logically about what you want to do and play around with it.  But you can have your ands and your ors all nested and you can exclude things, and so it’s incredibly powerful.  Incredibly powerful.

Paul:  Yeah.  Let me go through, I wrote down a wish list.  Let’s see whether we can do all of these in advanced segments.  Do people who visit the Get Started Guide stay longer on the site?  So yeah, I can find that out because the Get Started Guide is a page, so I can basically slice the advanced segments on people who visited the Get Started Guide.

Chris:  So what I would do there is create two segments.  One for people who view the Get Started Guide and one for people who don’t.

Paul:  Right, so then you can compare.

Chris:  Compare those two segments.

Paul:  Oh, easy.  Cool.

Marcus:  It’s a little bit more complicated than that though.  With the Get Started Guide because it appears in a big way on the first visit and then it only appears a little tiny bit.

Paul:  Yes, yes.

Chris:  Well then you can also add criteria to select new visitors versus repeat visitors.

Marcus:  There you go.

Paul:  See, he’s clever.  I like this.  What else have I got on my list?  How many people visit the Headscape website?  Well that’s easy.  What can we learn from those people who have visited the Headscape website or go on to?  Well we can do that because we’ve got our event trigger.  You don’t know the exit URLs, do you?  Does that make sense?  So, instead of having an event trigger can I just say anybody that goes to headscape.co.uk after visiting Boagworld, you can’t do that, you need that event trigger.

Chris:  No, you can’t do that.  Yeah.

Paul:  That’s fine, okay.  But then I can do all kinds of slicing and dicing of that.

Chris:  Because if you think about it, Google Analytics is only capturing information from your site.

Paul:  Yes, I know.  I know, it makes sense, but I just thought it does magic things, so it might do more magic things, I don’t know.

Chris:  It does magic things.

Paul:  What can we learn about those people who subscribe?  Well again, we can do an event trigger for people who subscribe and then we can learn all kinds of stuff.

Chris:  Advanced segment.

Paul:  Well, we need the event trigger and then advanced segment.

Chris:  Yeah, yeah.

Paul:  Articles which lead to view – ah, this is quite an interesting one.  Articles that lead users on to go to on to the “Hire Me” page or visit the Headscape website.  So, that’s okay because I can do the event trigger, the visit Heascape, or go to the “Hire Me” page, that’s an advanced segment, correct?

Chris:  Yeah.

Paul:  And then I can look at the most popular posts, could I not?  Within that segment.

Chris:  Yep, yep.

Paul:  So that would tell me what kind of posts are most likely to encourage people to go to the Headscape website or to the “Hire Me” page.

Chris:  Correct, yeah.

Paul:  I’m getting the hang of this then.  Okay, what else have I got on my list?

Marcus:  It’s all completely random anyway, so it’s pointless.

Paul:  Thank you Marcus, that’s really helpful.  Okay, so equally I could look at what articles cause people to subscribe because I can segment them by people who are subscribed and then look at the top articles.

Chris:  And then look at the top articles, yeah.

Paul:  Where do people go after reading an initial post?  That’s an interesting one.  So that’s not a segment, is it?  You can get that straight out of the normal Analytics.

Chris:  You can look at each post and then look at where people navigate after that, if you like.

Paul:  What else?  What am I trying to get at there?

Chris:  Why do you want to know that?

Paul:  I want to know I guess whether, after reading one post they A) drop out of the site completely or B) go on and read a second post or C) go on and look at an alternative page like “About Me” or “Hire Me” or something like that.  So essentially what I’m looking for is people who have viewed one post and what they do after viewing that initial post.  Does that make sense?

Chris:  Yeah.  There’s a nice view that shows navigation paths.

Paul:  Oh, that’ll do.

Chris:  So if we look at the navigation summary.

Paul:  Oh, ok.  So this is off the content dashboard.

Chris:  This is off of content.

Paul:  Yeah.  We wait for a while for the really fast internet access that we have here at the Barn, Two meg up, two meg down.  That’s all you get in the country.

Marcus:  I think it’s been doubled, hasn’t it lately?

Chris:  No, I don’t think it has.

Marcus:  Oh, it was going to be.

Paul:  Yeah, well a lot of things were going to happen.

Chris:  So we’re here, we’re looking at the homepage and we can see that from the homepage there are 32% exits.

Paul:  Right.

Chris:  And 68% move on to another page.

Paul:  And then you can see what those next pages are.  Yeah, that does it perfectly.  Cool.

Ah, that’s another thing I wanted to do.  Track how many people clicked through at the bottom of the new Boagworld website there is a list of Headscape people.

Marcus: I’m going to pick you up on that.  You confused the hell out of me on Friday.  He phoned me up in a tizz.

Paul:  Why is that?

Marcus:  Yeah, not really.  All of your questions to him were referring to the Headscape website and he says, “So, I can’t see any links to anyone else.”

Paul:  I wrote the entirely wrong thing.  I use them interchangeably.  So, at the bottom of the Boagworld website there’s a list of other Headscape people and I want to track how many people go to those, so they’re all going to need event triggers on them.

Chris:  They’re going to need event triggers because they’re separate sites.

Paul:  And they’re leaving the site.  Yeah, that’s cool.  You know what?  I think that’s it.  So all I needed to know was advanced segments and event triggers.

Chris:  Advanced segments are magic.

Paul:  They are magic!

Chris:  They’re magic.

Paul:  That’s cool.  I get it.

Marcus:  Yeah.

Paul:  See, that was worthwhile, Chris.  I’ve learned something new.  Thank you very much.  That’s brilliant.

So here I was thinking Chris was clever, but it’s just all advanced segments from the sounds of it.  I can do that.

Marcus:  You also have to use your brain and you add new rules and all that.

Paul:  Yeah, it’s thinking that logic through, isn’t it?  Yeah, absolutely.  But I mean, that’s brilliant.  So event tracking, advanced segments sorted.  So I feel like I’ve now got a firmer understanding and a firmer foundation for monitoring the successes and failures of the website.  But information provided by these kind of custom reports, advanced segments, whatever you want to call them, it gives me an indication of what subject matter would be good.  That was one thing that I liked about what you said about search results and stuff like that.  And which subject matter encourages people to complete calls to action, etcetera, etcetera.

So once I’ve got the basics up and running, I’m hoping I’ll get that taking over.  And then what I want to start doing is playing around with multi variant testing so I can actually try and influence these results and improve them and stuff like that.  So hopefully that will allow me to tweak the design and the content in order to maximize the number of users visiting the Headscape website and subscribing to the blog.

Marcus:  Put the link in the middle of the page, that’ll do it.

Paul:  Click this, yes.  That kind of thing, we’ll test that and see how it works out.

Marcus:  Spinning – they already are doing that, aren’t they?

Paul:  Yes, they are, maybe.  So I can use Google Website Optimizer for that, which is really simple.  If you haven’t checked out Google Website Optimizer do so, it is very good.  And if you are, like me, running a WordPress blog there’s also a very good WordPress plug-in that makes integrating Google Website Optimizer just a little bit easier, which is worth checking out as well.

So, I think what I’m going to do is once I’m up and running, everything’s going alright, I’m going to start with the Google Website Optimizer.  I’m going to look at in post calls to action, trying some different calls to action within a post to get people to subscribe or visit Headscape.  I’ll mess around with post titles and maybe try different titles, see which ones perform better.  Post lengths as well, I think length is quite a key issue.  Sometimes I wonder whether some of my posts are too long, a bit like this podcast.  The use of imagery post structure, all that kind of stuff.  So I can play around with those things using Google Website Optimizer.

And I’m hopeful that that will actually help increase the conversion rate.  I think the key here is that it’s really important to have an ongoing plan for improving your site analytics.  That’s why not only am I going to do some multi variant testing, but I’ve also got a plan of how I’m going to make my analytics more complex as I become  more confident over time.  I want to be eventually testing a lot more than just how many people visit the Headscape website and how many people are subscribed to my blog.  Although I don’t want to do that initially for fear of overwhelming myself – I want to kind of have that in my head.

Marcus:  Sure.

Paul:  So, eventually I’m doing more than looking at those two criteria.  Particularly, I’m also going to want to look at what you mentioned earlier, the profile of the Headscape brand.  In other words, how much it’s mentioned online, and also the visibility of other Headscape employees, which is a part of the same thing.

Marcus:  I don’t care about that.

Paul:  You don’t care about it?  Well that is one way of increasing the Headscape brand.

Marcus:  Oh, I’m joking, honestly.

Paul:  So, as I’ve already said the new Boagworld site is going to have that filter which highlights Headscape employees, so I’m going to want to track those.  And also we’ve got all the profiles, so when somebody else writes a blog post on Boagworld I highlight them a lot more and I want to know whether those make a difference, in particular I want to know is worth wasting other Headscape employee’ time by getting them to write a blog post.  Does it make that much difference to their profile if they write regularly, for example, so I need to be able to track that.

Marcus:  It should do, shouldn’t it?  But yeah.

Paul:  I would have thought so, but you want to be able to see how big a return on investment, you know if it’s interrupting them doing other work, how much are we really getting back from that?

I mean, ultimately though Google Analytics I think can only take me so far, because I’m not just interested in the traffic that Boagworld drives to the sites of other Headscape employees.  I’m also interested in whether Boagworld can increase their online profile and that of the Headscape brand.  And to do that I’m going to need something other than Google Analytics to kind of detect what people are saying out there.

You know, I know that we win work for Headscape through the Headscape and Boagworld sites, but it’s not just…

Marcus:  Which is great for me.

Paul:  Which is great, yeah.  But it’s not just that.  That’s the trouble.  Some people come directly to us without ever visiting either site and so I don’t think that we can rely purely on Google Analytics to track the success or otherwise of our brand building exercises.

Marcus:  It’s weird though, isn’t it?  Because yes, some, a few, and this is all anecdotal, but I would say a few come via not being Boagworld readers or listeners.

Paul:  Yes.

Marcus:  But of those few, they’ve had personal recommendations.

Paul:  Yeah.

Marcus:  And you’ll get one in a blue moon that says, “I found you on Google.”

Paul:  Yeah, most of it comes through Boagworld.

Marcus:  Yeah.

Paul:  But that doesn’t necessarily mean that it could have been a contact with Boagworld ages ago or it could be that they’ve just subscribed to the podcast without really ever visiting the website.  Or it could be that they follow me on Twitter but don’t look at the website a lot.  So there are a lot of other factors going on here and we need a tool that’s going to be able to track not just the influence of the Boagworld site but of the brand generally, and of other Headscape employees.

I mean there’s no shortage of tools like that.  I mean, one end of the spectrum is something as basic as Google Analytics which tracks mentions of your brand on Google and on the other there’s like enterprise level solutions, there’s one called Spredfast that’s got this whole host of very fancy tools and costs a small fortune.

One of the most popular tools for this kind of stuff, brand awareness stuff, is something called Clamped.  And although Clamped gives a reasonable indication of your influence on social networks like Twitter or Facebook it doesn’t really provide the wider picture.  For example, it only monitors Twitter or Facebook IDs, so it fails to pick up on mentions of other keywords, such as Paul Boag.  You’ll get Boagworld because that’s my Twitter name, but not Paul Boag.  So it can only go so far.

Because my objective is to monitor all mentions of Headscape and its employees irrespective of where or how they’re mentioned, Clout isn’t going to do the job, but I have found a tool that I’m pretty happy with and it’s one called Social Mention.  And this tool impressed me because you can search on any keyword, irrespective of whether it’s associated with a Twitter or Facebook account.  So earlier you said perhaps we need a Headscape Twitter account, well actually this way we don’t.  We can just monitor the word Headscape.

Marcus:  Cool.

Paul:  Which is great – it also is able to monitor across a range of sources, from blog posts to forums, to social networks, everything, which is brilliant.  Social Mention will allow me to see how effective our online marketing efforts are at increasing brand awareness for Headsape, myself, and the other employees.  My hope is that having this information will allow me to tweak my approach to maximize my return.  For example, which is more effective – writing a post for Boagworld or me writing for another website like Smashing magazine?  You know, because Smashing magazine has got a much broader readership, is my effort better placed writing for them than it is actually for Boagworld?  I suspect a mixture of both is the right way, but it would be good to get some figures on that.

Marcus:  My problem with stats in general is making decisions based upon very small amounts of numbers, and you don’t know you’re doing that.  Say for example you’re comparing just the example you’ve just given about is it more successful if I write for Boagworld versus Smashing magazine.  And it may be that you only get two people that in reality, physically two people that you can measure that they went to the Headscape website after reading this, blah, blah, blah.

Paul:  Right.

Marcus:  But all you’re getting is 100% of people did this.  And you’re making decisions based on two people, is always my concern.

Paul:  Yeah, well you have to look at the physical numbers rather than just the percentages, absolutely.  You’re entirely right.  I mean, I guess my desire to do this is that having this kind of information helps you to target the limited time and resources that you’ve got, because I don’t have that much time for this stuff amongst everything else I’m doing.  So I want to spend my time as effectively as possible.  And I think ultimately that’s what all analytics is about, we all have limited time and resources to put into our sites and our marketing and we need to know what gives the best possible return.

Without analytics we have no way of knowing that we’re achieving this or not, and even a limited knowledge is better than none at all.  I accept what you’re saying Marcus, but some figures are better than no figures is my attitude, as long as you haven’t put too much weight on them.

Marcus:  Yeah.

Paul:  I think the trick is striking a balance between meaningful analytics that help decision making and something that’s overly complicated and ends up confusing you or taking too much time to understand, and that’s kind of what this exercise has been about for me is creating a set of analytics that I know I will actually look at and use.  And I think I’ve got that now.  I think I know what I’m going to look at and I know what I’m going to do, which is great.

Marcus:  Fantastic.

Paul:  So, I think I’m pretty much done for this show, just your witticisms to end it with.

Marcus:  Yeah, I have got a joke.

Paul:  But…

Marcus:  But first of all, there’s something I need to tell you.

Paul:  Oh, dear.

Marcus:  But it’s really hard to say.

Paul:  I’m right about something, that’s what it is, isn’t it?

Marcus:  Ken Dodd’s dad’s dog’s dead.

Paul:  Ken Dodd’s dad’s dog’s dead.

Marcus:  Someone told me this and I thought, “Yes, that’s my kind of joke.”

Paul:  Ken Dodd’s dad’s dog’s dead.  Brilliant.

Marcus:  For all those people who aren’t my age…

Paul:  I doubt they’ll know who…

Marcus:  Ken Dodd was an old comedian.

Paul:  Yes, but really that’s somewhat irrelevant.

Marcus:  Yes, it’s just the silliness.

Paul:  The tongue twister, yes.  Very well done, congratulations.  We shall talk again in two weeks’ time when apparently I’ll be launching the Headscape website, the Boagworld website.  Get it right, Paul!  Bye.

Marcus:  Bye.