Getting the Most Value from Google Analytics
During this webinar, Search Mojo’s Janet Driscoll Miller and Scott Garrett will show you a few tips on how you can get the most from everything Google Analytics has to offer, mining the rich insights it provides to help you make better marketing decisions.
Presenters: Janet Driscoll Miller, President and CEO, Search Mojo and Scott Garrett, Account Manager, Search Mojo
Presented on September 5, 2013
Welcome to today's webinar, "Getting the Most Value from Google Analytics." I'm Kari Rippetoe, Content Marketing Manager at Search Mojo, and I'll be serving as your moderator for today's webinar.
Before we get started, I just have a few reminders for you. Firstly, a recording of this webinar will be made available to everyone who registered and will be sent via email by Monday at the latest. Secondly, there will be a Q&A at the end of today's webinar, so if you have any questions for our presenters, please enter them in the GoToWebinar questions box at the right of your screen.
Finally, we encourage to you to tweet about today's presentation using the hashtag mojowebinar. Plus you can also follow on Twitter @searchmojo.
Today's presenters are Janet Driscoll Miller, President and CEO of Search Mojo, and Scott Garrett, Account Manager at Search Mojo.
Janet has nearly 20 years of marketing experience, and in addition to her work in search engine marketing, Janet has a background in marketing communications. She holds a degree in Public Relations and Communications from James Madison University. She is a frequent speaker at Marketing Conferences and writes for several blogs and print publications.
Scott is also a graduate of James Madison University with a degree in Marketing. He manages both SEO and pay per click campaigns for a variety of clients in sectors that include technology, legal and healthcare.
Scott is a Google AdWords Certified individual and Google Analytics Individual Qualification and Bing Ads Accredited Professional.
Search Mojo was founded in 2005 and specializes in all things search marketing, including SEO, pay per click, social media advertising, online reputation management and content marketing. Search Mojo is headquartered in Charlottesville, Virginia, and we also have an office in Charleston, South Carolina. We've been featured in several marketing publications and blogs and we also speak at several conferences, including SMX, MarketingProfs, B2B Forum and PubCon.
Our clients include a variety of B2B and consumer brands, nonprofit organizations and educational institutions. And with that, I will turn it over to Janet.
Thanks, Kari. So to start off today, let's talk about the way we used to measure analytics and the way we used to use analytics to measure our marketing goals and other types of goals we might have for the efforts we make on our website.
So in the early days, we used to measure through the elements like visitors, unique visitors, page views, unique page views, also called uniques, pages per visit, new visits, time on site and bounce rate. So, this is a description for each of these to understand what they really mean and what they try to measure. Let's go through each one of them.
For the first one on visitors, that really is representative of the total visitors of your website, and that can be repeat visitors. So it's not uniques, which is the next piece, unique visitors. That's the total unique visitors. So if I come more than once, I'm still counted as multiple visitors, even though I'm one person, whereas unique visitors only shows me as an individual and how many times I personally came to the site only counts one time, or one visitor, I should say. It does not count repeats at all.
Page views is the total page views on the website, again, regardless of visitor, regardless of type of visitor, type of page view, etc., whereas unique page views looks at unique page views to the site and does not count repeat visits to a page.
So, for instance, if I visit one page 30 times and it's just me visiting that page it's only going to count me as once.
Pages per visit were how many pages on average were visited on the site per visit. And so if I come to the site, how many pages do I look at on average? Do I look at three, do I look at 10? And so people use this typically as a gauge to understand what's the type of engagement I'm getting from my site? Are people staying there? Are they looking at lots of different content on the site?
New visits is the percentages of visits that were from visitors who had not been to the site before. Pretty self-explanatory. So, obviously many businesses want to increase the new visits to the site because they feel that we want to attract as many new people to our brand as possible and so that was what that measurement was often used for.
Time on site, or in Google Analytics it's called visit duration, is how long a visitor, on average, stays on the site. So when I came to the site, did I spend 10 minutes, did I spend three minutes? How long, on average, are visitors coming to the site and staying? And again, sort of a measure of engagement with the site.
And then bounce rate is a measure of the percentage of how many visitors immediately leave the site once they get to it. So I find a site through Google, I click through, I go to the results, and I see the page, and I click on that page to go to your website, and if I back-click immediately, that would be a bounce. So if they don't look at anything else on your site and they immediately leave, that's where bounce rate is calculated.
And of course, ideally, you want to have a low bounce rate because you want people to stay on your site. But these measurements really aren't enough, and that's really where the challenge lies. We've matured quite a bit with how we can use analytics, and so I want to explain why, for a lot of these types of measurements, you might want to consider moving away from them into different types of measurement, of really what counts on your site, because some of these things really don't' tell you what you really need to know about the success of your marketing endeavors.
So, for instance, with visitors, unless you can remove them, which you can remove, to some degree, certain people, certain IP addresses, in Google Analytics, this number can include your employees.
So, for instance, if your employees have their home screen on their browser set to your company page, every day when they come to work and they start up their browser, boom, that's a visit, and it's going to look like a bounce probably because they're not necessarily do a lot on the site.
But also, if you're testing your site and anything like that where you might have a lot of employees, or even vendors, coming into your site, doing a lot of work, they will be counted in the visitor number unless you exclude them.
I look at it is this is just a traffic stat. It doesn't measure true visitor value, and the visitors particular measurement also double counts. So, to me, it's not really as valuable. I don't really care about double-counting people's numbers. I want to know, specifically, what individuals are doing on the site.
So, unique visitors might be better for that, but just like total visitors it could include your employees and it is just a traffic stat. It doesn't say how valuable those people are. It doesn't tell you what kind of demographic came to the site. It really doesn't tell you much about them, other than this blank number. And what does it mean if they don't convert, if they don't actually yield sales or leads? So, visitors, in general, isn't necessarily, by itself, isn't a great statistic to look at.
Page views and unique page views are kind of in the same boat because they show the page that was served but doesn't indicate any action was taken. So great, they visited this page, then what? It really doesn't always tell you exactly what you need to know. Again, it's great that they looked at information about your products, but if they didn't request a quote, does it matter? It really doesn't matter if they don't come back and request a quote.
So page views, while it has some level of value, again, it doesn't get down to the nitty-gritty of what you really want to know about who's visiting your site and what they're doing on it.
Pages per visit, is this really in, does this really indicate what your goals are? What I would encourage you to do is know who your personas are, because, really, I can tell you for my site, as an example at Search Mojo, I can't say that people come to our site who are interested in buying from us and read 20 pages.
So what is the right number of pages? I mean, what should your average be? What looks like a good number? It doesn't necessarily mean that a higher number here is better than a lower number. It just depends on what your personas do and how they act and how they buy from you. And so you have to know what you're looking for there. You can't just generically say "Oh, 20 pages per visit is awesome, and we should have as many pages per visit as possible." That's not necessarily indicative of the buyer.
So in our case, as an example, many people come to our website and want to learn from the great resources we have here, and that's great, and we appreciate that those people want to learn from us, but if they visit 20 webinars, are they a buyer or are they somebody who's just trying to come to the site and glean information and learn? So really, it can be a double-edged sword in looking at pages per visit, be careful of how you judge that. More is not always better.
New visits is subject to cookie deletion. So, for instance, and a lot of these really are, honestly, but, like, uniques also are subject to cookie deletion. If I delete my Google Analytics cookie then you're not going to be able to tell that I've been to the site before, because I've deleted my cookie, and so Google Analytics, and really, almost any analytics package, really cannot put the data back together to understand who I am, and that's a problem, because then you can't always tell who is new and who is not. So it's not a really exact measurement. You've got to be careful about that.
Time on site or visit duration, as well as bounce rate, are really, I think, not very reliable at all. I would not use them for measurement for many reasons. They can be a guideline, but they're really not a great measurement. With tabbed browsing and search came around with Chrome and Firefox and now Internet Explorer as well, basically think about it like this, if I open a tab on Chrome to look at a site, but, "Oh," I get called to a meeting, "I've got to run," and I left that tab open, what happens to my time on site? It may look like I've been on that site for an hour and that page for an hour, but in reality I was not digesting any material on that page. I was in a meeting with someone else, so I wasn't even looking at the page.
So, tabbed browsing has kind of changed some issues with bounce rate and time on site, but specifically time on site. And then there's server time-outs. If you've ever been to an eCommerce site, most eCommerce sites have server time-outs, many times for the purpose of making sure that a cart, for instance, a shopping cart, gets emptied if you leave the site or over a certain amount of time, because the retailer does not want to leave, let's say, this one pair of shoes in the cart waiting, hoping you'll buy it, if there's a ready buyer who can buy them and you're not going to buy them, you're just going to leave them in your cart and abandon your cart. They want to make sure those shoes are available to somebody else.
So, server time-outs allow, especially with eCommerce, but in many different facets, log-ins for instance, and different sites, to basically define when you consider someone having left the site. Basically this is how long the period is.
The tricky part about this is, you can set a different server time- out on any server. There's no standard here. There's no default. It could be anything. A server time-out could be five minutes and it could be 10 days. It really just depends on the server and what you set it at.
So when you try and compare your bounce rate, what would happen with server time-outs is if you were on that page, let's say I'm on a home page of a site and I have my tab open and then I had a serve time-out because maybe it's set to five minutes and I just went to the bathroom and came back and it's, and the server's timed out, it will look like I bounced on that site. And so that's a problem, because if you're comparing your bounce rate to someone else's, they may be measuring and setting up things like server time-outs completely differently than you are. So you don't necessarily want to make that comparison and just judge yourself on that particular number. It can be a good guideline, again, but it's not something that I would use as necessarily a metric about the success of my program.
So, "Okay, Janet, you've just depressed me terribly, that all these great measurements I used to use are dying and I should not use them anymore. So, how can we use analytics for measurement today?" Well, I'm going to hand it off to Scott now who's going to talk to you a little bit about some of the initial things you can do, especially in Google Analytics, that can help you to measure what really is important to you, and that's things like conversion and goals that you might have for your site. So, Scott, take it away.
Thank you, Janet. Okay, so I'm going to go right into goals in analytics. So goals in analytics allow you to measure and track some specific actions that users take on your site. Each time a user completes one of the specific goals it counts as a conversion and is logged in your account. Goals are a great way to help you see past your traffic and really drill down to see and measure the true actions that drive success for your website.
Now there are four main types of goals: destination, duration, pages, and page per visits. So, the first type of goal is destination goals. Destination goals allow you to count as conversion every time a user reaches a specific page on your website. For example, you could place a destination goal on your "Thank you" page of your newsletter sign up or your webinar registration.
Second, in duration goals allow you to track the amount of time that a user has spent on your website. For example, you could create a goal that counts a conversion every time someone spends more than seven minutes on your site.
And then the next goal is pages per visit. So, this goal enables you to count a conversion every time a user has visited more than a certain amount of pages that you have set it as. So you could set the conversion to count for whenever someone has viewed more than three pages, or five pages, whatever you want.
This would allow you to track users who have visited more than one page and actually explore your site and not only the home page.
Lastly is goal tracking, and it counts a conversion whenever a user completes an action on a page. For example, an event goal could occur when someone clicks to watch a video on your page or clicks to download a PDF.
In addition, you are also able to assign a monetary value to each goal completion. So you can actually give a dollar amount to actions that are performed on your site.
Google Analytics allows you to see the channel information for each goal completion also. This information will help you see what channel drives the most conversions for each particular goal and it will allow you to better allocate your market and resources to top performing channels. By default, what Google Analytics does, however, assign each goal completion to the last click that immediately proceeded each conversion action. However, this can be worked around using attribution modeling, which Janet will discuss later in this webinar.
Next I will further go into details about goals in analytics. Goals are separated into five sets of four goals each. These sets of goals allow to assign particular types of goals to each set. This will allow you to easily find and track similar goals. For example, you could set a set of goals to be dedicated exclusively to newsletter sign-ups or video views.
Once you've reached the limit of 20 goals in your profile you may either edit and change old goals or create an identical profile to allow you to access 20 new goals. I would always recommend going ahead and creating a new identical profile with access to 20 new goals, because if you edit and change old goals, it would make it very difficult going forward to find historical data based off your old goals.
Lastly, please remember that goals do not track retroactively. They will only track conversion actions after a goal has been successfully set up and activated. So next I will go into the basics of setting up a goal, actually, in Analytics.
So, setting up a goal in Analytics is quite simple. First you need to go to your profile and click on the admin button on the top right corner and then you click on the goals then click "Create a new goal." You will see a screen that looks very similar to the one right here on the slide. You will then want to click the edit button and you can name your goal and also pick a goal type.
Once again, there are four different goal types, that are destination, duration, pages per visits, and events. Since I've chosen destination goal as example I will now need to set a URL that will count as a goal completion. For a goal completion URL, you can make it equal to a specific URL, a beginning with, or a regular expression formula of your URL. In this example, I've simply chosen equals to for my destination goal.
Next, you can assign a monetary value for your goal if you think it necessary, and lastly, you can also create a conversion funnel for your goal tracking. So this will actually allow you to specify that a user needs to land on page A and click to page B, and then finally convert on page C for it to count as an example. But this is only required if you think it's necessary and you simply need to turn on the funnel feature and set it up that way. But for this example I've opted not to do that.
Finally, you simply need to click "Create goal," and congratulations, you have made your first goal.
Next I will go over the eCommerce module in Analytics. Apart from the standard goal tracking in Analytics, you also have the capability to track eCommerce metrics if you run an eCommerce site. The eCommerce module can be found under the conversions section drop-down and this side shows you an example of the eCommerce module. It gives you a summary of key eCommerce performance metrics, such as transactions, revenue, conversion rate, and also unique purchases.
This overview gives you the ability to view top revenue sources of your site by product, product SKU, product category, and also source medium. These quick views of top revenue sources will allow you to see what products or channels are driving the majority of the revenue for your site so you can easily allocate more resources to top performing sections.
The eCommerce module in Analytics offers various reports that give you more detailed information about metrics that drive the overall eCommerce performance of your site.
Next I will go into more detail about one of these reports. So here's an example of one of the eCommerce reports, and this interesting report is called the Time to Purchase Report. It allows you to see the overall time to purchase of your products through the metrics of days or visits. This allows you to determine the average duration from the first time a user visits your site to when they actually purchase and become a customer, giving you important insight on the buyer behavior of your customers. Knowing this information can help you fine-tune your marketing tactics.
For example, say, you saw the majority of your customers purchase a product within three days of initially visiting your site, you might want to consider re-marketing to users who have not purchased after three days so that you can encourage them to visit your site again and purchase to become customers.
Next I'm going to actually go into the details on how to set up eCommerce tracking properly in Analytics. So properly setting up eCommerce tracking in Analytics can be difficult, but overall is a very straightforward process. First you're going to want to ensure the Analytics code has properly been places on your receipt or "Thank you" page of your shopping cart of your eCommerce site.
This code is a little different than the standard Analytics code as it contains additional parameters that will allow Analytics to pull data such as revenue, product name and product category into the system. I will not go into the sort of details about how to specifically set up the parameters and the goal, and the code needed for the eCommerce tracking to work, but there is step by step information and it can be found on the Google Developers website.
So after the re-marketing code has been successfully placed on your site, you're going to want to ensure that you have actually enabled eCommerce tracking in Analytics, and to do so you simply need to check a box labeled "Yes, an eCommerce site," and this can be found under the view settings tab in the admin section of your profile.
Finally, if you have an AdWords account that you use to advertise your eCommerce products through, I would highly recommend linking the AdWords account to your Analytics account. This will allow you to see keyword information and campaign data within Analytics from AdWords, and therefore you know which keywords are driving your top performances.
So linking your AdWords account to your Analytics account is quite simple and can be done through the account settings in the admin section. As you can tell, Analytics links quite easily with AdWords, however, it does not link well with other ad platforms, or does not link at all with other ad platforms such as Bing and LinkedIn. So you may be asking yourself "How exactly do you track campaign information from Bing and LinkedIn in Analytics?" and I'm going to show you.
So here's the Analytics builder from Google, and this builder allows you to create URL tagging, more specifically UTN tagging. UTN tagging in URLs is recognized by Analytics, and allows you to see the keyword and campaign data from other platforms such as Bing and LinkedIn as I mentioned previously. So you just do, all you need to do is fill in all the fields that you feel are necessary to fulfill your tracking needs, and then you click the "Submit" button, but I would always recommend filling in all the fields as more information is always better when it comes to tracking.
The tool will then spit out a URL that can be used and placed on your destination URL for your paid ads. In this example on the slide, you can see that my ad source is Bing and my medium is cost per click, and I've also filled in the campaign term, content, and the name, to allow me to see more detailed information when it comes to tracking in Analytics.
As you can tell, I am a fan of bacon, and if anyone knows of a summer bacon sale, please tell me, but next Janet will be going over exactly how to measure success in organic search results.
Thanks, Scott. So now that Scott's shared with you some of the ways that you can also measure your goals and conversion levels and so forth, your eCommerce and just the general goals you might have on your site, for, like, lead generation and so forth, how can you measure different channels?
So, here at Search Mojo, obviously to us it's very important to be able to measure the success of the work that we do with certain channels and today with all the craziness around how it's going to be more difficult to actually get generic rankings and understand how websites are ranking in organic search, really, the reason measurement has to be around the value we're getting out of organic search from a lead generation and a dollar perspective, actual sales.
And so one thing that you can do to measure your organic search is after you've set up these goals or you've set up eCommerce, you can go into the organic section under traffic sources as I've shown with that big red arrow there. So you can measure organic search value to a specific goal, be that eCommerce or be that a goal, a specific goal like lead generation, because as a marketer, what you really care about is conversions, not just the traffic. You care about those goals.
So, now on the other side, if you look at this particular area, you can track the various channel attributes. For instance, you can look at Google versus Bing or one website page versus another, like the different landing pages, because the beauty of that is, for instance, you may find while Bing doesn't have as much traffic as Google and everyone talks about Google all the time, what if you find that more people from organic results from Bing actually convert better? That's a huge piece of knowledge for you, because then you can put more of your effort into Bing and try and get more converters coming into your site, people who are going to convert on the site, as opposed to just people who might be browsing. So, even just things as simple as looking at the different, the search engines and the sources there in organic search can make a huge, huge difference in how you prioritize your marketing efforts.
You also need to remember that unfortunately keyword trafficking is essentially going away, which stinks, stinks, stink, stinks, but, you know what, maybe we don't need to be measuring them in the keyword. Maybe this is a lesson for all of us as marketers. And the reality is that you can measure down to the keyword to some degree, but it's not going to be very accurate measurement right now because of the fact that so many, and on our own site as many as 50% of the search queries that come in show "Not provided." You may see that in your organic search as well, and that's because those searches are encrypted, and it doesn't matter what you do to get around it, you will not get that data. So it is really challenging.
So try and avoid keyword tracking or tracking down to the keyword for your goals, because the reality is that's probably not very accurate tracking to begin with. So I would stick with some of the other pieces of the channel that can give you really great, informative data that can help guide your efforts, like, for instance, maybe which source or medium.
So here's a look at what the report would look like or the, I guess it's a report, I guess it's what they call it in the Google Analytics, and what you'd see if you had set, for instance the goal conversion rates. So you see here with this red arrow where you can change, instead of just site usage and visits, you can change this view to be the goal conversion rate. So how high or low was the conversion rate on a given day? And if you look at the bottom you can see how the different landing pages are doing from organic search. So the one, so here's what the traffic from organic search looks like and how they perform in the website, which is really helpful to know, again, from a channel perspective, do you prioritize organic search?
And so you see here in this box below I mentioned how you can also instead of just looking at, say, landing page, you can look at the source, right? Was it Google, was it Bing? How does one perform versus another? And you can see in this particular case we have eCommerce showing, and you'll see that the average order value from Bing is $171.88. The average order value from Google organic is only $127.59. So your better buyers are coming from Bing, even though there aren't a whole lot of them.
So think about that and use this data to your advantage, because that can be a competitive advantage for you in trying to tackle how to prioritize your marketing efforts and where to put the, for the time and effort.
Now, yeah, now that we've talked about organic, what about paid search? How can you measure paid search more effectively? Well, as Scott had mentioned, you want to hook your Google Analytics to Google AdWords. It enables the Google Analytics to show up in AdWords, which, frankly, is just a big time saver. You don't have to keep going back from AdWords to Analytics to Analytics to AdWords. It's very handy.
Also you want to make sure the auto-tagging is enabled in Google AdWords. It is on by default, which is fantastic, because as Scott said, you want to capture everything you can because you can't go back later and fix things. Well, you can go back later and fix things, but you can't get the data back. Once the data has happened, it's gone, it's not captured. So you want to make sure that you're capturing everything at the moment that you can capture it, and even if you don't use it today at least you've got it there in case you need it later.
It also allows for Analytics code, again, to be automatically inserted into destination URLs by turning on that auto-tagging. So what Scott was showing you with the URL builder, you don't have to do that for AdWords, which is fantastic. You can just have it auto-tagged for you and it makes it a lot easier and faster.
But again, you can use, you can still use Google Analytics tags on destination URLs, on ads in other paid search platforms, like Bing, if you use the URL builder that Scott demonstrated, because that way you can measure things like different campaign names and so forth from, that are residing in other platforms like Bing or Yahoo! or what have you. You can start measuring from those different areas.
So here's an example of a paid search traffic report, and again, looking at, this one's looking at visits, but what I wanted to show you us that similar to organic search, the paid search measurements can go far beyond just conversion tracking. For instance, Google Analytics allows you to look at orders and order value.
So as an example, I mentioned earlier how maybe Bing might be more valuable, here's an example again where you can see that the conversion rate here, the eCommerce conversion rate on Bing from paid search was actually higher conversion rate, .11% I believe is what that says, over Google, which has .08%.
So, again, converting more people. And the average order value, in this example, isn't as high, and certainly the transactions aren't as high of a quality, but the thing that you need to know here is that Bing deserves some respect, and maybe you should think about it. I know we all talk about Google, but when you look at your own data and you start to see some of these trends, it will help you define where you should be putting your money and how you should be focusing your efforts.
Same thing is really true for display ad measurement. It's very similar to tagging for paid search ads and other platforms other than Google AdWords, and again, you want to use the URL builder to build your destination URL with the tags that you want, with the different sources and mediums and so forth.
So here's an example of a report you can run with paid search traffic or any kind of paid traffic, and you can see here that I've got Google, LinkedIn, and Facebook there. So if you're doing Facebook advertising, you're doing LinkedIn advertising, here's a great way for you to tag those URLs so that you can see on the back end how they compare with other channels.
Which takes us into our next area, which is attribution modeling. One of the biggest challenges I think marketers face today is attributing the value to a particular channel. Which channel was the thing that pushed these people over the edge and made them leads or sales? And it's really hard to tell.
And this is a report that came out about a year ago showing that most marketers are trying to attribute some level of credit to a particular channel, however, what you'll see is that many, many people, either 1/3 of them are not tracking it at all, any kind of attribution, or 1/3 of them are tracking last attribution. The ones that just say "Yes, it doesn't define exactly how they're tracking attribution."
And so last attribution isn't necessarily the most accurate, but it is, unfortunately, what many marketers are sort of tagged with today and have to deal with because of the limitations of many of the pieces of software that we use.
So, for instance, if you're trying to use, if you use salesforce.com as your CRM tool one of it's many challenge is that is only has one spot for lead source. Well, you have a choice. You can either use the first interaction as the lead source or the last interaction, but you have to choose. You can't show every interaction along a path, and as well all know as marketers, it's not typically just one thing that made someone convert, it could be multiple things, multiple interactions. So how do we attribute value to each of those pieces so we can fully understand how to budget our time and our resources?
So, lo and behold, a beautiful tool in the Google Analytics called attribution modeling. It's, as I mentioned it's one of the greatest challenges to marketing, but you can find this in Google Analytics and you can use it to compare different attribution models to see which one really most accurately tracks attribution as you see that it is happening for your personas and your leads that are coming through.
And so if you go under the conversions area, it's under attribution, it's called Model Comparison Tool, as I've shown with the red arrow here. Now, there are currently one, two, three, four, five, six, seven, seven types of attribution models that Google automatically gives you. These are the most common modules that we would see. They include last interaction, which I just talked about, which is the last thing someone did before they converted, the first interaction, which is the first thing that they did before they converted, last non-direct click, which is the last thing they interacted with before they click on an actual link, and then last AdWords click. So what was the last AdWords campaign, etc. that they responded to?
Linear is giving equal attribution to multiple things over a period of time. So, for instance, they may have responded to an email, then they looked it up through SEO or through organic search, and then they clicked on a paid search ad another time. So all of those things would be given equal weight.
Time decay, however, says give the most value to the thing they did most recently, but still give some level of value to the first thing also and all the things along the path, just reduce the level of importance over time. That's what time decay is. So the most recent thing is the most important.
And position based basically says the first and last are most important and everything in between gets a little bit, but not as much as first and last. So you can try these different models and see how they work for you and see how they really reflect what you can see as the true measure of how people are coming into your site and how they're interacting with it.
The other thing that's great is you can actually do a custom model if you want, so if none of these fits your model and you don't feel this is the way people are interacting with the site and converting, then you can create your own model. And there's a lot of different setting you can set in attribution modeling. It's really a fantastic tool, and I'm going to take, show you a quick screenshot here of how this works.
So here's an example of the Model Comparison Tool, and what I've done here is I've compared first, last interaction, to first interaction. And what you can see is I've highlighted here in this red box all of the different channels, paid search, direct is when they just type in your URL to your website, referral, display ads, organic search, and social networks. And what you can look up here is where the grey bar there it says last interaction. It shows conversions. And then adjacent to that to the right you'll see first interaction conversions, and then it even gives you, because this is a set-up for eCommerce, it even gives you the values that paid search, on average, the cost per acquisition was for those particular channels. And you can see that paid search, as an example, look at the difference there, as a first interaction, had a lower cost per acquisition than it did if we counted it as the last interaction as being the most valuable thing.
And same thing with display advertising. So, you want to, again, play around with this, see how it works for you, and if it's accurately reflecting your attribution model for your company.
So now I'm going to hand it back to Scott, who is going to talk a little bit more about how you can continue to improve your conversions and your conversion options through Google Analytics as well.
Thank you, Janet. So, how do you improve your website even more? So one way to do this is using AB testing. So, in AB testing in Analytics, it's accomplished through something, what Google calls Experiments. So these experiments allow you to test some variations of the webpages to determine which version is best at driving specific goal completions. Experiments can be found and set up in the content drop-down section of your analytics interface.
So, each experiment can test up to nine variations of one webpage. So it is quite a good many variations. So these variations are used to determine which is most successful at driving your conversions.
So set up an experiment, you need to ensure that your goals are properly set up in your accounts because experiment's based off goal conversions. For true AB testing I would recommend using only the original page and one variation of the page in the experiments because to do this it allows you to just test one or two variations in your design. For example, you could test the changes in button color, and also changes in call to action for the button, and see how this affects your rate of driving new goal completions on your site.
So in the example in the slide you can see that variation page, the orange dot, and the original page, the blue dot, are almost in a dead heat in terms of conversion rate for goal completions. You will notice that the experiment has been running for 35 days now, which isn't out of the ordinary, as some experiments will run well into two months, however, I always recommend running these experiments for a minimum of two weeks to ensure that a large enough audience size has been sued for these experiments.
And then you'll also notice on this slide there's a great graph that shows you the conversion rates over time. So you can see whether the original page is winning in the beginning, and then it dropped off, then it's more like in a dead heat, and then, so over time you can see the changes in conversion rate, and also in goal completions. You can change that.
So next I will go into a user feature in Analytics that allows you to see key performance metrics in one single view. So this is a dashboard. So this dashboard is a useful feature, and that allows you to organize and easily see all your key performance metrics in just one single dashboard view. It's pretty self- explanatory. So this slide example shows you dashboard center for eCommerce sites with key performance metrics included, such as cost, revenue, ROI, average order value, top products and average cost per order.
And dashboards can also be easily set up for other non-eCommerce sites to track goal completions, organic traffic numbers, and pretty much any other metric found in Analytics.
So you may be asking yourself, "How exactly do you set up one of these dashboards?" So next I'm going to go into exactly how you set up one of these dashboards. It's fairly simple. All you need to do is click on the dashboard drop-down, which is located in the top left corner of the Analytics interface and click on "new dashboard."
The dashboard uses customizable widgets to display your data. These widgets allow you to select specific metrics and then define them with various dimensions and filters to create snapshots of your performance data. Currently there's a limit of 12 widgets per dashboard you create, so choosing and creating the correct widgets for your specific needs is very important.
You always want to make sure that you create the widgets that display the most important performance metrics for your site, whether that be revenue for an eCommerce site or webinar sign-up goal completions for a service-based company.
So setting up a widget's fairly simple. In this example you can see that I went ahead and labeled the widget "Summer Webinar Registration Sign-Ups" because clearly it's for a webinar that we'll [inaudible 37:38] here in the summer, and these are kind of the goal completions for every time someone has completed the sign-up and landed on a "Thank you" page.
So, when you go ahead and do this you want to choose a specific metric you want to measure, and as you can see here, I've chosen the goal 19, which is the webinar sign-up goal. Then you also want to pick how you want your data displayed. So you can display it in various forms. You can pick a timeline and GEO map, a pie chart, or even a bar graph. So, viewing this performance set in various forms can help you, give you more powerful understanding and help you visualize what your performance of your sites.
So always keep that in mind. So then you can click "Save" and there you have it. You've created your first widget, and you can create 11 more widgets if you'd like to and fill up that dashboard with great information that you can pull at any time.
Thank you very much. I'm going to go ahead and pass it off to Kari now.
Thanks Scott, and before we get into some Q&A I just wanted to let you know about our upcoming webinar, which will be two weeks from now, on September 19th at 2:00 p.m., and Tad Miller and Blaine Anderson will be telling you a little bit more about how you can leverage YouTube for online marketing success. So you can register for that today at search-mojo.com/youtube.
And if you're looking for a little bit of help with your search marketing, then you can get in touch today with Sean McCusty and there is his contact information there.
And if you'd like to connect with either of our presenters, Janet Driscoll Miller or Scott Garrett through social media or with Search Mojo, then you can do so with the information shown here.
Search Mojo offers Google Analytics Consulting Services to help you set up and get the most from Google Analytics, as well as Search Engine Optimization Services and Pay-Per-Click Management Services.