RED1: Hello is Anyone Out There? Using Web Analytics to Understand your Audience

Kyle James 
Manager of Consulting Services, HubSpot


The audio for this podcast can be downloaded at


Kyle James: So let's get going. Who am I and why am I up here talking? I used to run a college website, so a lot of the data that I have is from Google Analytics back then. Some of that is two years old... but Google Analytics hasn't changed a lot since then... so you might notice a few discrepancies. I've updated some things, but not everything, because I'm not the Google Analytics power user I used to be.

What do we really want to accomplish here? We're going to build the argument of what is the importance of analytics, why do you want to track it, why Steve needed to do it two years ago.

We're going to talk about some key performance indicators, we're going to understand some key terminology... because there's a lot of misunderstanding about what certain things mean inside of analytics... all things Google Analytics, we're going to go through a lot of filters, a lot of things that you can do and go back and do now... and I think Seth is actually doing a workshop on Wednesday on Google Analytics to carry on with some of this... some additional analytics stuff and four rules for analytics.

So let's get going. A lot of stuff.

I always like starting with this one: if a tree falls in a forest and no one is there to hear it, does it make a sound? Does it matter?

 01:02

You could argue the same thing with a website. If you don't know if you're in a forest or you're in the middle of a city, the way that you find out those kinds of things is analytics.

And the difference between the Web and, say, print media or television or whatever kind of marketing you've done in the past is you can track everything for the most part. So that's why this is important because you want to make sure that you're not in the middle of a forest and that you are in the middle of a busy city.

The second point of this: are you listening to what your visitors are telling you? And that's really what we can get out of analytics. People are visiting certain pages and are doing certain things, but what are we doing with that? Are we being actionable? Are we making that experience better for them? Are we listening to them? Really, really important thing that you can do now that you couldn't easily do in the past.

How are you making Web decisions? You're taking this input that people are giving you, this data, by what they're doing and what they care about, and institutionalizing and making changes towards that.

 02:04

If you're deciding, 'We need to change a page in our website. You know what? We're going to change this page,' and if you don't look at something like your analytics, you might be updating and edit a page that gets one page view a month instead of something that's getting 1,000 a month. I mean, where's your time better spent there? How do you know unless you're looking at it?

So what is important? What do we really care about in our website? I'll just try to outline some possible goals here.

Obviously, from the Admission specific, 'apply to college'. That's a big deal. That's really, really important. We want to know that. We want a bit of tracking, we want to know where these people come from, we want to know why they're here.

And right below that, 'schedule a campus visit'. A lot of times this leads directly into applying to the school. Maybe request information, download a viewbook, take a virtual tour, sign up for email updates from the school, watch videos, subscribe to RSS, join a Facebook fan page, join a LinkedIn group, read blogs... so a lot of stuff that we want people to do. 

 03:04

But how do we know what they're doing or what they care about what they value? So really, go back and think about what are those action items that we really care about in our website and how are we tracking them?

A couple of key performance indicators, and it really does break up into these four different segments here.

I always like to start with success. Conversions. The single most important part of anything on a website is a conversion event. What is the point of a website... and I know there's a lot of different debates about this and that, but from a marketing standpoint, the most important thing on your website is if you're getting applicants.

If you're not getting new prospective students coming to apply to your school, you might not have a school anymore in a few years. And how do you measure that success? Are you measuring that success? Do you even care? I hope so.

Users. Segmenting down your audience. What kind of people are coming and visiting your website? What buckets do they follow? Who are your audiences you're targeting?

 04:01

We all know we've got a couple of different audiences: prospective students, alumni, faculty staff, current students, parents. Think about those and think about the different message you're giving to each of them. The content. What pages are they looking at? What specific things are they looking at?

And of course, sources. Where are these people coming from? Are they coming from Google? Are they coming from other sites? Are they direct traffic? Do they type in the URL? Are they coming from email? A lot of different places.

Understanding some analytical terms. We all think about, oh, we understand the basics and stuff like 'visits', 'visits by page', 'absolute unique visits', 'percent of new visits', 'traffic source', 'landing page'. I think we all have a pretty good idea about what these are. Anybody not... anybody want to... ?

But there's a couple other ones that can get very confusing, like, what exactly does a 'page view' mean? Is it important or not? 'Average time on site'. And it's just important to make sure that everybody really understands each of this.

 05:05

So something like page views... what does that mean? Is it good to have a lot of page views per visit or one page view per visit? It doesn't matter because it's completely dependent on what kind of engagement you have.

The example I always like to give is, all right, if somebody can come to my homepage, and I've already identified that the very most important thing to me is them applying to my school and they can do that in four clicks, if my average 'page views per visit' is 10, am I failing or not? How do you judge that? I mean, that's probably not a good thing if you've got that many pages per visit.

The flip side is, if you've got something like a blog, say, .eduGuru. If we've got something like 1.2 page views, that might be a really good thing because that means that everybody that comes to my blog and reads it, they don't need to go to other pages because they're already an advocate for it, they're already a die-hard reader and they've already read everything else in the site.

 06:04

So page views, throw it out. It's meaningless. It doesn't mean anything. You can't really gain a lot of insight from it.

'Average time on site'. And this is one that gets really tricky for people. The example here is if someone came to my homepage at 10 o'clock, one minute later they come and visit a second page, and then four minutes later they go to a third page and then leave, the total time on site there is five minutes because to a tracking code, there is no way of knowing how long someone was on that last page.

Because there's a timestamp that's taken every time someone renders a page, they pull it down to look at it. But on that last one, you don't go to a fourth page, so you don't know how long you were on that third page. So keep in mind that 'time on site' is always dependent on minus that last page and how long people were there.

'Bounces' and 'exit rates'. It's important to remember that a bounce is always an exit, but an exit is not always a bounce. It's inclusive.

 07:08

A bounce is if someone comes to your homepage and then they leave your site immediately. They zero out the window, they click on a link that takes them to another website, whatever... that's a bounce. One page, gone.

An exit is always the last page that they were on. So if they go and start out on your homepage, go to your admission page, and then go to your application to apply, we've got an exit because that last page here has a bounce on it, but they actually visited some pages on the site so that this first few pages wasn't necessarily an exit from your site from that.

Does that make sense? Remember, it's always an inclusive kind of thing that a bounce is always an exit but an exit is not always a bounce. So when you're looking at those, notice the difference there.

'Traffic types'. There are a couple of different traffic types you'll see here, and the ones that stand out are organic. Organic is what comes from search engines. It's people that are going and looking for stuff and then organically clicking on results to take them to your site.

 08:10

The other thing that comes from search engine is something called PPC or CPC depending on how it displays up here. If you're doing something like Google Adwords or any kind of ad campaigns, you'll see that. If you're not doing that... you're like, 'What are you talking about?'... don't worry about it.

Of course, 'referral sites' is someone else out in the Web that is linking to you. That's what that is. And of course, 'direct traffic'. And you can have other different sources in here, which we'll talk about later how you can go about tagging those things.

Benchmark. There's a lot of different ways that you can benchmark stuff. Three of the bigger data providers.

Why do you want to benchmark? Maybe you care like, well, I think we're getting about this much traffic. How do we stand up against our competition... our peer institutions, our rivalry schools, whatever it is? Here's three different sources you can look at.

 09:01

Alexa is pretty good. I think Quantcast is really good. They give you a lot of these demographic data you can see here, if you're just curious. And Compete is another one.

With all of these, it's not going to be 100% accurate. It's guesstimations. It's guesstimates. I love that word. You can get an idea... all right, what are we looking at? Because you have no way of knowing your competitors' traffic just like they have no way of knowing yours because they can't see your Google Analytics account or whatever you're using. So this is something you can do to cross-compare how you're doing. Three different services here.

This is a nice little tool that me and a couple of other guys at nuCloud built out. It's called eduRank. The link's there at the bottom. Just running through, we pull from some of these sites and give you a ranking based on 1,600 schools we've gotten the index. If you're just curious, 'How do I stand up against other people?' it pulls from some of those resource we just looked at.

And of course, even in Google Analytics, you can go opt into this. And it's buried, but you can see from the Law and Government Education Colleges and Universities, you can see how you stack up against all the other colleges and universities that opt into this.

 10:08

But it's always going to be very different because comparing a little private school that has 2,000 students to a big school like Florida that has 50,000, it's hard to really compare because that's not an apples-to-apples comparison. But it's another good place you can look.

So there are a lot of different analytics packages. There are a lot of different analytics solutions.

Avinash, who's one of these visionaries, one of these brilliant guys in the Web analytics space, his rule is interesting. It's the 90/10 rule that you play with analytics that basically says, of your budget, if you're going to spend $100,000 on analytics, you should spend $90,000 of that towards someone who actually interprets the data and $10,000 towards a solution. Don't go buy some big expensive solution if you have no one that actually knows how to read it.

 11:03

Good thing about Google Analytics and all these guys right here, they're free. So it's a great place to start. There are a lot of other analytics packages you can buy... yeah, we've even got one at HubSpot... but I think for going in this, start out with Google Analytics. It's free. You're not going to pay anything.

But a couple of things to understand about why it's free.

We always think of Google as this 'do no evil', big, gigantic entity. They do make some money in some ways. Just know that 98% of Google's revenue comes from advertising, comes from when you're on a Google search, those ads that go on the top. That's why they want to click more analytics on the Web. The more they know, the better able they are to serve up results.

They're giving you this data for free, but you bet your pretty penny they're also looking at it, too. The more they know, the better they can serve up ads. And also, because their bread and butter, 98% of the revenue comes from ad words, it has very, very tight integration with Google Analytics.

 12:03

So if they can get you using one tool for free, it's like, 'Well, you know what? I'd like to buy some traffic.' Great. Now they've got that tight integration and then upsell you from there. So be a little bit smart about why this is.

Another is a lot of security concerns in higher education about, 'Is Google Analytics OK?' It's like, 'I guess the rule of thumb I'd say is everybody is using it. It's Google. They're the biggest website, the biggest Web player out there. They're not going to do anything malicious. They're not going to spam me.' Because what's the first thing that's going to happen? We're all going to be Tweeting about it and we're going to quit using it. I mean, they've got too much on the line to not make this thing rock-solid.

Installing Google Analytics. It's pretty simple. They've actually updated the code, but this is the old one. It still works just fine. I underlined this little line here because it's something special you can do, and we'll look at this in some filters. I don't know if you still have to do this or not, so this is one area where this session might be a little bit dated. But it's still pretty awesome. You all are getting some good stuff out of this, right?

 13:08

All right, installing Google Analytics. There are a couple of tools. Once you get this tracking code on your site... and literally, installing Google Analytics is something that should take no more than five seconds. Grab this code that we just looked at, go into the template of your website... hopefully you've got a template in every page in static HTML or CMS or something... and you want to input this code right above the closing body tag.

And the reason you want to input it above the closing body tag is because you don't want this thing slowing down the page load time. You want to have that great user experience. Google is not going to load something slow, but still, put it at the bottom. Everything else gets rendered, then they load that.

But they have two free tools... actually, they're free to an extent... SiteScan and WASP, that will actually crawl your entire website to make sure that you've got the code installed on every page. Because, once again, the data is only as good as you set it up. So if you're missing pages that don't have Google Analytics installed, it might be a page that's getting a lot of traffic, and you don't know anything about it.

 14:08

Specifically a plug for WASP here. Anybody want to tweet me? I've got a free professional account with these guys to give away. I met some of their product owners. Send me a tweet, I'll dissect them and give somebody a free account. A really cool tool. They actually do a lot more than Google Analytics. They can see like 200 different analytics tracking codes that lets you know if it's on your pages. So send me a tweet and I'll give you a free account with them. One person.

All right. Data overload here. So we get Google Analytics installed. We've got it running on our site. We forget about it because we get dragged into a hundred different directions, and we come back a month later and look at it and like, 'Holy cow, what do I do with this? I don't know what to do with it. I'm overloaded. There's too much data in here. What does it mean? I can't really understand it.' So what's really important is you can actually come in here and start segmenting and filtering this data to get down to, what are the actionable steps I can take away?

 15:04

When you first set it up, it asks you for a Google Analytics profile. And what's important to remember is you can have up to 100 different profiles that you set up on one tracking code. That's good. That's big.

You can see a little bit of the specific things here. Make sure you define your site URL, with or without the www. Now this is branding purposes. I can get into a whole 'nother debate about which one's better. There is no right or wrong answer. It's more about how you want to brand and market.

Set up a default page. Let Google know what your default page is, excluding the URL queries. If you've got certain stuff, canonical issues that show up that, well, 'This page loads with this, but I'm doing that for some other reason whether it's a tracking token or something,' make sure you tell it to exclude those.

And definitely, definitely do this last one: set up site search. We're going to talk about a whole great wealth of information you can get just from that.

 16:00

Google Analytics profiles. This is what I had set up in Wofford back in the day. You'll see like there's 13 different profiles. But why this is really valuable is because you've got different people in your institution that care about different things.

Like, I've got a default one here that has all the data coming into and I can read it for when I want to. I've got one that's raw that I have none of these filters or any of these things set up, which I'll show you some of those. A sandbox, when if I'm going to set up a new filter, I'm going to play around with it. I'm not sure if it works or not, I'll set it up in the sandbox.

But then I've got one for Admission only, off-campus visitors. The Admission people don't care about all of this other stuff. They just care about their own subdirectory and maybe even their subdomain, so we can actually break it out and set up something so that they can get their data.

Same for Athletics, same for Alumni. GIFs whatever else, the blogs. Now you can set up all of these different ones and track them as individual segments but then also have a big one where it all comes back into.

 17:02

The first thing that will probably happen when you set up Google Analytics, you'll get something like this. What's wrong with this picture? Anybody want to guess?

Audience 1: They're all the same.

Kyle James: They are all the same. They're all the same. It's the same page, but we're seeing four different results. You can't read that! So what you've got to get in here and do is set up a couple of things so that you get this all down to one result.

And the first thing you can set up is 'all lowercase' filter, because you'll notice that, all right, there's a capital D here, there's a lowercase D here, but no, they're all the same page. So you can come in here really easily... and by the way, all of this is up on the blog, so there's a lot of stuff you can get to later... set up this nice little thing that makes all your results in displaying Google Analytics in lowercase. Everybody should just set this one up. It's the easy one that just always, always needs to be done.

The second one go into with the www or non-www. Unless you've got '301 redirect' set up one way or the other, to Google, these are the same pages in its duplicate content. But you don't want it to show up in your analytics as two different things, so make sure you set up one way or the other.

 18:16

It's all about branding. It depends on which way is better for you. Personally, I like without dubs because it's a lot less to type in. But some people might like it with it for branding reasons. It's your decision.

Set up subdomain traffic. Let's say that you've got a subdomain. In here, for example, back in the day... I think they did away with this, but Wofford had their whole Athletics on a subdomain in the main website. I wanted to set up a filter so that those guys could track that separately in their own profile, but we could also aggregate it in a bigger profile, too, so they could get their reporting without seeing everything else.

Here's a way you can track subdomains separately without tracking them inside the rest of their traffic using the same tracking code. And there's a lot more filters here, because you don't always care about everything, right?

 19:07

This is one that you would put on maybe your master account, that I want to set up and monitor everything, all of the subdomains, so that if you've got a blog subdomain... you've got a WordPress installer or TypePad or something... but you want to aggregate that in your big bucket, you can do that with this filter here. Great, great tool to use.

'Exclude IP traffic' filter. This is especially for those departments that are facing externally like your Admission and your Alumni. If you're an admission department, you probably don't care people at the school that are visiting your admission page, right? 'Who cares about us? I want to know what is my audience doing.' So if you work with your network administrator, if he can give you the IP address range for all your internal traffic, guess what? Boom, filter it out. Clean up your data. Make a lot more sense of it.

Because if you've got a page and if you don't do this, and if you're looking at something like 'schedule a campus tour' and you've got 300 views but 100 of them are your Web guy because he's been working on that page and making some adjustments, that's not really what's going on there to your general audience. Great, great filter.

 20:18

'Directory' filter. This way, if you do have your admission department that's on a subdirectory of your site, you could set up a profile for them that just monitors that traffic. And this is the filter you would apply to that profile. You could say, all right, for the Admission people, I want to set up a profile that just looks at this subdirectory and excludes all external traffic and say, 'Here you go.' And they've got something segmented, they've got something meaningful to them without having to look at the bigger picture there.

'Country' filter. Now this is really... depending on what kind of audience... If you're a community college, you probably are not getting something like a lot of international students, right? Or maybe you are targeting internationally. You can come in here and set up different things for country, region. There's one for city.

 21:03

If you're a community college, maybe you don't care about people that are visiting your college that are outside of your own city. So set up a profile for that and just monitor that in there. A lot of different things you can do with this filter.

'Full referral URL' filter. Google Analytics does this little trick... and they might have changed this; correct me if I'm wrong... that if you go to referral sources, it doesn't always tell you the exact URL that visited the site. They'll tell you, 'Here's the domain, here's the website that sent some.' Well, I want to know exactly what page that link was on.

You can set this up on a profile. It's like, 'Oh, well, I got a link from some government site, but it was exactly this page,' and then you can go look at it. This is kind of cool if you're an SEO junkie, too. These two definitely go well together.

Tracking and tagging. There's a lot of different ways, as we talked about, you can segment your audience. By far and away the best way to do this.

 22:03

And this is one of those things that it's really complicated to wrap your mind around 'destination URL' tagging, but once you get it right, it's magical... especially for stuff like email marketing, offline campaigns. Because you might have a page, and you've got a billboard, and you want to know, 'I only want to know people that come visit this page.'

So you've got a nice little short URL that you give them and you set up a 301 redirect at some landing page. You can assign some of these tags, these tracking tokens, to the end of this URL. I'm going to say, 'Oh, these people came from this billboard instead of something else.'

Who out there has seen those TV commercials that say, "To learn more about our site, visit oursite.com/tv10?" You all have probably seen those. Those guys are telling you, 'We've got at least nine other commercials and we're measuring the effectiveness of each of them against each other.' All those tv10, tv9, they've all got these destination URLs set up in the backend that redirects them to the same landing page. They're not testing the landing page; they're testing which commercial's more effective.

 23:07

You can do the same thing if you're doing... I'd like to call it snail mail... if you're doing student mailings and you want to know, 'Was this mailing more effective than this one?' You're sending the same page, you can use this kind of stuff to track that.

When to tag URLs? There's three big buckets this falls into. Email marketing, 99% of all email providers will do this for you. You've just got to dig around the setting somewhere and try to find Google standard tokens. Turn those on, your analytics will be 100 times better.

Page search. Is anybody out there doing PPC? Do people do PPC? I see a few hands. So underutilized in higher ed. It's a great way to drive target traffic especially for enrollment purposes. But if you're doing that, you definitely want to be tagging that traffic and follow these offline event trackings. Offline event can be a lot of different things.

 24:03

And what I always recommend that when you're doing this, build a spreadsheet so that you can easily keep up with all these campaigns you're doing. Little URL down here in the bottom, bitly.com/urlbuilder. This is a Google spreadsheet that I've built. You can go get it, download it, reuse it for your own purposes. You basically give it the description, give it the URL, give the source, medium, campaign, and boom, it busts you out this destination URL.

So you don't have stuff like, 'All right, we call it 'e-mail' in one campaign and then 'email' in another one.' It's really important to be consistent with these tracking tokens because it's going to look so much weirder in your reports if you don't.

You can also track links. This is another cool thing that you can do. You can do this for a lot of things. You can actually track Flash inside of Google Analytics using these on-click events. You can track file downloads.

In this specific example, you can see this is a link on the Wofford website where we sent people to download our viewbook. Is that important to know if anybody's downloading your viewbook or not? I mean, that's cool to say, 'Oh, we're having 100 people download it a month, then we're having 500 people download it a month.'

 25:16

And these are some things I used to put up on Guru a while back, that we've got this little PDF directory and I can see, 'Oh, this was Rachel Reuben's social media report that she put up two years ago,' and in that timeframe... I think it was a month or whatever... we had 60 people download it. Kind of cool to be able to know that.

So it's really easy to track those things. Stuff like your course catalogue... you probably want to know how many people are downloading that... definitely the viewbook, and probably some other things out there.

You can tag audience segments, too. Here's a couple of ways that you can set up these page trackers to actually track. If you've got a campaign that you're only driving prospective students to it, you might want to segment by that. So you can say, 'Oh, here's the campaign. We always drove prospective students to this page. If they come through this channel, we can track them that way.'

 26:09

I've never really used this a lot. I'm assuming it works. Right? If somebody's got a better way to do it, let me know.

All right. Reporting. There's a lot of reports in Google Analytics. I'm just going to run through a couple of my favorites and why, how you should really be using them.

Site search. I mentioned this way back. When we're setting up, it's important to set up site search. And the reason for that is this is where you find out where you're failing. That if people are having to use your internal site search, it means they can't find something on your site that's really valuable to them.

In this case, the Number 1 thing that people can't find on the website is bookstore. Well, we probably want to make that a little bit more obvious, because there's two kinds of people that use your site search: those people that are desperate and can't find what they're looking for and people that are just plain lazy.

There's a lot of us, we get to a website, we know what we want to look for. We don't want to try to find it; we just want to search for it. So help those people out.

 27:03

You'll notice in the top 10 results here, it's interesting that 2, 5, and 6, employment, human resources, jobs, well, we've obviously got a lot of people here looking to get a job. You might want to make it a little bit more obvious for those kind of people to find that so they're not searching our website.

So this is telling us where we're failing, and then some statistics about that. How many people are searching? How many times do they have to reiterate on that search? How deep are they going?

'Keyword' report. One of the big things about SEO. Everybody's probably heard it. Keywords. The buzz around keywords. Keywords are what people look for.

Not surprisingly, this is taken from Wofford. The first nine items all have 'Wofford' in it. Not surprising at all. But then notice number 10 here, 'Betsy Cox novels'. Betsy is a professor at Wofford who is a famous novelist. Well, knowing that we're getting search traffic, people are going to Google and typing that in, there's a lot of low-hanging fruits that you can get there.

 28:06

If you want to learn a lot more about SEO, I'll be in some other room right after this presentation to give you 75 more slides on that.

But the big thing to notice, this is your low-hanging fruit. If I want people are looking for on your site and they're trying to find, and make those pages more obvious, do some optimization around them. You will get a lot more traffic from them.

'Content by title' report. Once again, this one ties directly into SEO. These are the page titles for those pages. If it's not really clear and obvious what each one of these is, you've got some clean-up work to do, because this becomes especially important for Google. Because when someone goes and searches Google, this is the results that show up.

So if every single page on the site were to say "Wofford College - Quintessential Liberal Arts Education", well, that doesn't help me if I'm looking for the Biology section, does it? So you want to make sure that every one is specific and unique and it really describes what the page is about.

 29:01

And the best way to do that is this report right here, because you can easily scan through things like, 'What does that title mean?' and go look at the page and say, 'Oh. We need to clean that up a little bit.'

'Referring site' report. Another good one. Where are we getting traffic from that's not necessarily Google or direct traffic? Pretty interesting here. Of course this screenshot's two years old now, but wofford.fyi is a recruitment portal we had. Terrier Fans is a forum where sports fans could go talk about what's going on. Facebook's right up there. GoUpstate is a local newspaper.

It's important to know who's linking to you. Who is sending traffic to me? It can be especially important if there's a certain kind of community out here who are really talking about you to go out and engage with them more to get additional links, nurture them along. They're obviously rabid fans, supportive. Just see where you're getting your traffic from, because there's a lot of times you'd be like, 'Wow, we're getting a lot of traffic from that. Where is it going? What are we doing with it? Are we getting any value out of it?'

 30:07

'404 Error page' reporting. Does anybody know what a 404 page is here? So it sucks, right? It's frustrating when you get to a 404 page.

Well, this is a big usability thing. You can actually set up a filtering Google Analytics... and there's a whole blog post about how to do this; the very top is links... so that I can see these are pages that people are trying to find on my site, it's sending them to the 404 page. So they're looking for Athletics, football, default, something, something, something. You can drill down and look at it.

Well, I'm failing right there. All 12 of these people... I think, in the last two weeks that tried to visit this page... had a sucky experience. They tried to go see something, it wasn't there. We failed them.

So go get those kinds of things cleaned up. You'll have a better user experience. It's great for SEO because they're clicked on something. And it's really, really good to set up these things.

 31:00

A perfect one is... I remember that people would go type in 'wofford.edu/math' looking for the Mathematics department. We called it 'Mathematics'. Why not go ahead and set up a redirect form right there? You could clean it up for them.

There's always a debate, well, do we call it Admission or Admissions with an S? You're using one, you're not using the other, I guarantee you people are typing in the other one. Go set up a redirect and then make that experience better for them. This is how you learn about where those low-hanging fruits are.

Goal tracking is probably the most important, biggest thing that you can get out of Google Analytics. There's nothing more important to be able to say that... Seth and I were talking about this earlier and he's got a great workshop that is really focusing on this. Unfortunately I'm not focusing on this in this presentation.

Let's say that you do have a 'schedule a campus tour' or 'to come by, schedule a campus visit', and you know that 25% of people that schedule a campus visit actually apply to your school, and you know that you accept 25% of applications, and you know that the average student stays three years and is worth X dollars to me.

 32:05

Well, you can actually compute an ROI for anybody that completes that event. Well, get in there and put up a goal conversion event, especially if you're trying to get credibility with your boss. And you can say, 'Well, we know that it's worth X.' Do the math backwards. And if you can say, 'Well, this form is generating X dollars worth of revenue,' you really build a lot of credibility for yourself and you really help buy the authority and the support to get more things pushed towards that direction.

Goal tracking is really, really important. You can see where conversions happen. Seth is doing a workshop on Wednesday. There's your plug.

[Laughter]

There's a lot more. You've got all these other channels that people are coming from visiting your website: newsletters, Facebook, TV, LinkedIn, direct mail, email, identity management.

And we'll start with these, with offline campaigns. I mentioned these earlier, that you can track this kind of stuff. You can tell if someone comes and visits my website from a link in the newspaper if you set it up right. 

 33:08

Because the problem with all of these forms of traditional media that we don't have a problem with with online is we can track it. If you do a radio commercial and you put it on when your team's out playing a basketball game and you promote something on your website, you have no way of knowing how effective that was unless you set up one of these kinds of campaigns. The only way you can do is if you drop them online.

In this example, these are a couple of newsletters that we used to do. Terrier Top 5, building up some stuff towards homecoming, and then the news where we send out news releases. Well, each of these links has these tracking tokens I was talking about earlier so that afterwards, I can actually come in here... and this is Bronto for this specific homecoming event... and I can look in Bronto by my email marketing thing and get an idea about how many people click through things, how many people came and looked at stuff.

 34:03

But that only tells me half the picture. That tells me they clicked through the email. I want to know what happened to them once they clicked through it. And by setting up these tracking tokens that we talked about earlier, you can actually see that in Google Analytics.

You can see that, wow, all right, not only did we get traffic to the site... this is a homecoming one... but what's really interesting is look how many people actually kept that email and was still continuing to read it two months later. It's the kind of insight you wouldn't have otherwise if you didn't have these two things set up and linked together.

So you could see the other side of the event. Especially if you've got the goal conversion set up on it, then you know, well, this email actually drove this much revenue to our website. Big, big picture stuff. Big, big insight.

How many of you today have any idea whether an email that you sent out actually... how it impacted some kind of campaign that you're running? That you could actually trace the steps all the way through? A few of you. A few of you. How important would you say that is to, like, decision-making? Yeah.

 35:06

Audience 2: [35:06 Unintelligible].

Kyle James: Right. Yeah, it does. You're right. It does take some work to really take a look at everything. But if you're talking about a 'get big' campaign, you want to know how effective it was, especially if you're trying it through multiple channels. Email is just one of those channels. Snail mail is not necessarily dead, especially for something like homecoming here that goes to Alumni. Your Alumni members that are 60 years old, they're probably not as active on email. But they do know how to use the website to an extent.

Blogs. A couple of things to know about blogs. Two big tools: FeedBurner, ShareThis. That's really it. be on the track, that kind of stuff.

FeedBurner gives you a lot of picture like how many people are... You've probably seen this. Any blog that you look at now shows you the subscriber numbers. A lot of them have that publicized.

 36:

FeedBurner is really the only good way to get any kind of idea about how many people are reading a blog, or an RSS feed for that matter. And the way that it works is it looks at the daily pings to that feed. So that feed sitting out there somewhere, anytime someone pings it for information, it looks at unique pings in a day. Something like .eduGuru, we've got 2,000 readers if you look at the thing. That means in the last 24 hours or yesterday, 2,000 people went out pinging that feed to look for it.

It's also why you'll notice that if you ever look in RSS subscriber counts, on the weekends it always jumps way down and then jumps back up, because no one opens up their reader on the weekends. Or fewer people do. It gets volatile through the week. But setting up FeedBurner is really, really important if you care about those kinds of numbers. A very good way to share engagement.

And then that ShareThis widget, you've probably seen it. It's getting really popular for people that like to share content through social media channels or email. Really good. They've got some pretty basic analytics you can get on the backend.

 37:03

But what's really interesting is even now as we talk about how important social media... Twitter, Facebook, all this stuff... is, email is still the prominent way that people share content through the Web. So email is not dead as I argued two years ago.

Web Analytics 2.0. Now that you understand all the basics, there's a lot more that goes into this. And Avinash really talks about... you understand the data, but there's still another piece to it, and that's where stuff like doing surveys... SurveyMonkey is a great free tool. What's really become popular is just Google Docs. You could set up a Google form and request information, dump it in a nice spreadsheet.

Survey your audience. You could see that a lot of people are coming, looking at a certain page, but that only tells you half the story. Find out how they're interacting with a page. Find out why it's valuable to them. It's OK to survey your audience every time, especially if those audiences are really important to you.

 38:00

Actually, this is a slide that's probably outdated. There are not all these different social media sites anymore. We've really condensed down to two winners and a Tier 1, A1. You've got Facebook and Twitter, and then LinkedIn's kind of a tier down. Those are kind of the big boys. Because you can monitor something like the growth of Facebook over time. Facebook's got these nice little insights.

I'm sure everybody has a fan page for your school. You can also do things like set up groups. It's good to see that engagement growing over time. But that's about all you can really do with it... I think. I hope I'm not shooting myself.

Same for LinkedIn. But LinkedIn doesn't give you any data so that if you want to track something like maybe an alumni group on LinkedIn, you really have to do it through something like set up a Google doc and remember once a week or once a month to get in there and look at the growth of it.

This is two years old, but Wofford set up their LinkedIn alumni network. I set a personal goal that said, 'You know what, I want to get to 200 members of this group in the next month and a half,' and worked with the Alumni and worked with things through the websites. We made different places that are on the website that drove people to that.

 39:14

We did certain email campaigns, and as you can see, once a week I went into a spreadsheet that I was keeping track of and kept track of it. I don't think we hit our goal in my timeline, but we did hit it. But you've got to do this extra legwork on your own to keep up with LinkedIn. It's surprising they still haven't fixed that.

YouTube gives you all really cool analytics. Once again, this goes down to keywords, too. Think about relevant keywords. Think about what people are looking at. Also think about how you're promoting it, where people are watching it.

And then finally, monitoring your online identity. How many of you all have a Google Alerts account that just sends you stuff regularly? Yeah. I mean, that's the big one. It's become the de facto. There are a lot of RSS feeds you can... This is my insanity of trying to keep up with all the mentions of Wofford a while back, but I realized that one email a day from Google Alerts did everything.

 40:06

There is a takeaway: don't try to kill yourself like I used to. Monitor what people out there are saying, and especially monitor what links you're getting.

And finally, viral campaigns. Viral videos, it's like magic in a bottle. You never know when you're going to find a good one, and you're just going to keep trying. There's been a lot of interesting discussion of blogs recently.

This is a cool story how our president three years ago now went and did a TED talk. And then usually you think of a viral video, it has to be something that's like five minutes long or less to be viral. This is a 20-minute talk that just went all over the place. I think Chris Brogan blogged about it before anybody even knew who Chris Brogan was. I was like, 'Oh, that guy! I knew him!'

All right. Final thoughts here. How are we doing with time? We've got a few minutes left. Wow, I made it through all that.

Number one... so the four takeaways if you take anything away... always be testing. Set up campaigns. There's a lot of different tools you can use here. Think ahead of time. 'What do I want to accomplish with this?' And just be testing it, testing different links to different places. Set up different little testing environments that you can do, because you never really know what's going to happen until you just throw it out there and see what the results bring back.

 41:24

Don't get caught up looking at all the numbers. Look at the trends. This is a big one. Looking at something like, 'Well, we had X number of visits last month and this number of visits this month. What does that really mean?' Look at the trends over time.

Most likely during the summer months, all of you are going to have a drop in traffic. That's just the nature of it. So you can't really compare that month to month. But you can look at the trends. And once you've got over a year's worth of data, well, how does May of 2009 compare to May of 2010? Are we continuing to get more traffic? And trends also from campaign to campaign. How are we growing those?

 42:03

Set up a reporting schedule and track key metrics. Find out what those metrics you really care about are and track them. I've got a couple of old reports I could probably send people. They're buried on the Guru blog of monthly reports that I used to put together for Wofford.

Figure out what these things are. You want one way to get credibility? Put together a monthly report that you can present to... I sent this to the whole president and cabinet. They all didn't care about the Web all the time, but once a month I could drop some in their inbox to get them a little bit excited about it, and then give them a big footprint. Really, really important to do this.

It takes time, but you don't have time to spend every single day in analytics. We just don't. But if you can block off half a day once a month, it's a lot more realistic, and you're going to get a lot more actionable things. And it goes right into these set goals. I love the idea of setting monthly goals. What do you want to accomplish this month?

 43:01

Knowing the cycle of the year and looking at things like analytics, when it's a certain time, you'll notice that certain things are happening. Like the beginning of the school year, certain things are more hot and people are looking for them.

Knowing that ahead of time or knowing that from year to year, you can set goals. 'All right, before September 1, we need to make sure that all our content for stuff like Greek Life is ready, that all our stuff for incoming students is ready.' And making sure that those kinds of things are set up, they're optimized the right way, and they're easy for people to find.

And finally, just to leave you with some takeaways. Set some goals. I challenge each of you to take one of these, take none of these, make up your own, but I want to say that 'Our campus gets X campus visits scheduled a month.' If that's something important to you from a recruitment standpoint, track it and be able to say, 'All right, we've got 150 visits scheduled last month. We want to shoot for 160 next month.' Track it. See how you push it, because that will get your decisions about how you set up your websites and how you target different things really make you think about those.

 44:08

You might say, well, we don't have a link on this page. It's related to that content that links people to our 'schedule of visit' form. But you might go back and do it because now you've got an actionable goal you are shooting for.

'Number of viewbook downloads a month'. If that's important, shoot for a goal. You'll probably do the same sort of thing. It's like, 'Well, is this a page that makes sense to link to that?'

'Lower bounce rate on specific landing page X over next month'. So if you've got a conversion page and you know that people are bouncing from it, maybe you shorten up the form. Maybe you try to make the copy stand out a little bit more to increase the conversion rate. Maybe conversion rate is not as important to you as actual number of conversions, so you drive more traffic to it. You get a lower conversion rate, but you're getting more people converting. So what are your goals, and work towards them.

'Generate donor email campaign to generate X amount over next month'. All right, we're going to send out an email campaign and we're going to try to drive this amount of money. Well, you've got the tools now to track that. Now put it all together and give yourself a target. Set a goal. And once again, once you have it set up, you can test it for next time. So you've got incremental things as you go.

 45:19

'Increase our video exposure X% for next month'. Because if you're trying to drive people to your videos, you're going to be doing other things through social media, throughout the website, to drive people to those things.

'Promote and increase presence on a social network by X%'. If you want to do a big campaign to grow your Facebook reach, you're probably going to do things to work towards that.

Finally, a couple of other closing-out things. If you haven't seen Karine's Higher Ed Analytics Revolution, I definitely recommend checking out this site. She's got here these 12 metrics that every website, every college should at least benchmark themselves by. Great little survey. Fill it out. There's sharing data, it's benchmarking against other institutions. Fill out the quick little survey. It's a great little resource to check out.

 46:08

Let's get everybody on analytics standards so that we can compare it and we're all pushing forward and know what's going on. So, so important to what you're doing. And it buys you a lot of credibility.

So a couple of readings to close that. As I mentioned, again, doteduguru.com/webanalytics has got everything I just talked about in blog format. There's posts on every single one of them.  Occam's Razor here, that's Avinash's blog. This is his book on the right-hand side. I've mentioned him a couple of times. Dude's a genius. He works for Google now. He's like their Google Analytics tsar of the world.

Shelby Thayer. You probably know who she is. Great higher education blogger on web analytics. And then of course the Google Analytics blog. Great, great resource for when new things come out. These guys are the one to talk about it.

So if I'm understanding, you're asking me... because something like WordPress or any of these blogs generate the feed, you're asking, do you need to set all of those up with FeedBurner links? Sure.

 47:08

What you need to do is set up FeedBurner as like the intermediary step so that you've got your feed that your journal's pushing out, and we push it through FeedBurner. And then what you pull on this dashboard or whatever else is using that FeedBurner feed. That way, it gets all the backend analytics. And it also does stuff like it can standardize the feed.

So the question was, how do you know that FeedBurner is counting people? I guess you don't. So it could be that you could have 10 subscribers to a blog, and one of them could be a Twitter account that you've got set up that's automatically pinging that feed to post to a Twitter blog, one could be some search engine that's crawling that feed and getting it.

I guess it's not necessarily people. It's direct interactions with it. They're unique. And as one individual, if I'm doing something like I've got a feed set up with something like a Google Reader but I've also got something set up in my inbox, I'm technically hitting that feed twice. So, yeah, it is that kind of thing. But it's two unique ways of interacting with it.

 48:09

It's definitely not a perfect metric, but it's better than nothing, because the alternate of not using FeedBurner is nothing. I'm still waiting for... I'm surprised that Google has not integrated FeedBurner with Google Analytics. They bought FeedBurner three years ago and it's still not integrated. I don't know why, though.

So the question was about custom reporting. When I first did this presentation, I didn't really have custom reporting. I've done a little bit with it. What I've found was fun was custom reporting off of stuff like social media services and seeing which way you can pull that. If you're doing stuff like the tagging that I was talking about earlier, that's gorgeous for custom reporting because you can really segment these different campaigns that you're driving stuff in. And it's the only way to really look at one format that's really clean.

Yeah, but I haven't used a lot of it. I'm sure Seth could probably go a lot deeper than that. Yes.

 49:05

So the question is, if you change a URL, is there a way to know that they're the same page? Not really. Yeah, I mean, the only thing you can do is look at that data going forward because if you set up 301 redirects right, there won't ever be any analytics associated with the old page because, a lot of times, that redirect is rendered before it even gets down to the tracking code.

But there's no real good way to say, 'Hey, I know all this traffic's associated with this page. Now make it this page.' Unless you were to go write custom filters for every single one of those, and that's probably a lot more work than you want. I don't think there's a way of just mass-importing that in.

I haven't. I haven't. This is where Google Analytics and me separated when I went to HubSpot. We've got our own analytics and I'm just focusing on that. Once again I'm going to dump it to Seth because he could probably go into that a lot deeper than I can.

 50:04

Yeah. No, that's true. If you couldn't hear Mike... anything that's like Ajax or Flash, things like that that are not necessarily rendering a whole new page but specific events on a page, I think that does go with that whole on-click that I was talking about with downloading files. It's 'be as creative as you want'.

I think there is a post on Guru that Nick DeNardis wrote a while back about how he actually uses that same event to track external links. You could actually set it up that if I know there's any links that are leaving my site, I can track them using that same event. Tracking, and you have some sort of special way, a special profile filter of reporting on them. But that gets a little bit... there's some cool advanced stuff, too, because you need some special Javascript.

I know there is a WordPress one. So the question was, do you know of any plugins or tools that allow you to track that off-site? I know there is one for WordPress.

 51:01

The last time I looked at it, it was like a year ago, it was very much still in development. I think it worked, but you had to go do some custom things to your template. I'm sure it's a lot farther along than that now.

As far as specific other CMSs, I don't really know.

That's probably why you could do that with CSS.

All right. Well, guys, thanks for coming. I've got a few more of these. If you want some free schwag, come up and get it. And come to SEO and some other room in a few minutes and I'll keep talking.

[Applause]