 Building Blocks

To know and embrace analytics you need to consider a few building blocks of the whole system (these exist as options on most views). The following analytics elements are from Google.

Segments let you isolate and analyze subsets of sessions and users. In other words, these allow you to filter through the data and get the information you want. You use dimensions and metrics to do this.

Dimensions describe the data and are included on most reports. A dimension is a descriptive attribute or characteristic of an object that can be given different values. For example, a geographic location could have dimensions called Latitude, Longitude, or City Name. Values for the City Name dimension could be San Francisco, Berlin, or Singapore.

Metrics measure the data. Metrics are individual elements of a dimension that can be measured as a sum or a ratio. For example, the dimension City can be associated with a metric like Population, which would have a sum value of all the residents of the specific city.

What follows is a quick summary of the top features our partners find useful and keep in mind there is much more we can do, like setting custom alerts if something odd happens, specific custom reports based on multiple business goals, custom dashboards (we will set you up with a couple important ones at the start) targeting your most important performance indicators, dimension of behavior, mobile app analytics, Google+ integration, and much more. It’s deep. When we meet with you and get to know your business we will recommend some more involved Analytical properties or customization that are appropriate to your needs.

We will also adapt and modify as we work with you and see your business in action. Successful want to set it up so that you see the information you need to make intelligent decisions with your Analytics, but not so much data that you get buried in a sea of data, which is very possible. The following categories are listed in order of popularity and how you will see it in your dashboard.

 Acquisition

Acquisition is where we find out where and how you are getting your traffic. Analytics is going to tell you where it is coming from, usually in three different ways- organic, referral, or direct/none (user puts in your web address directly into the address bar is the most common direct). Then you can go deeper in each and find out specifically from whom it is coming. For example, imagine you had written a guest article on the Widget World and the Widget Town sites about your Yellow Widget product. We would dig deeper into referrals and be able to tell how many people came to our Yellow Widget page from your articles and which site sent more traffic. Importantly, we can start looking at the quality of traffic and how they behaved once here. Then you are going to know that it might be better to write more articles on Widget Town then Widget World because you got better returns there.

Acquisition goes deeper into campaigns, which is where we set up your goals to measure non-ecommerce conversions. Imagine we had multiple contact forms or even an event of viewing a video. We would set these up as goals (conversion points) and then be able to tell how many people reach the page, then converted. We could then optimize accordingly. It is powerful stuff. Keywords are also included with acquisition, though these are now becoming less useful as the browser moves to more anonymity and a securer web – so mostly you will see “none provided” here but you can get some useful information out of the ones that do and the paid ones.

The social analytics also lives here. If you are active in social networks, this is an important area for us since we can tell what social campaigns are performing well, in addition to which social networks are benefiting you and at what level.

Adwords is here as well. If you are using Adwords, we will be spending time here and in Adwords itself because they both have analytics components and work in sync. The adwords area really shows you how the campaigns are performing with the others in the acquisition tab’s sessions (similar to visits), bounce rate percentages, new users, pages per session, session duration, and goal(s) completion. This, plus the keyword / display placement / targeting, is critical to making informed choices about your Adwords campaigns and making adjustments to them. A must view.

The last of the popular options in acquisition is search engine optimization. Here you can tell (and like Adwords integration with Analytics so is Analytic SEO with Webmaster Tools – look at both) what organic search queries or search terms where put into Google as well how often they showed (impressions), how many clicks and click through rate percentage, as well as position on the search result page. You look at the most popular landing pages and like queries. You’ll then be able to get a snapshot of how they performed. Often times there is an art here. Through analytics you may get a hint from one part of analytics and then have to dig deeper in another to understand the complete picture. We love researching and can do it for you but we are happy to train as well. Finally, in SEO there is a geography summary that tells you what countries the visitors are coming from.

Here’s one quick note on direct traffic – we recommend using a tagging tool for any address you add to an email, pdf or word document- really anywhere outside the web so we can refine this information even further. You’d want to know if you were getting massive traffic from your pdf wouldn’t you? We’ll set you up with tagging. Also with email our custom email package will set up tagging for you automatically and most of the better email service providers (ESPs) like Constant Contact and Aweber can be set to have tagging enabled.

 Building Blocks

The starting point for any good SEM campaign is the research. If we have product A (in our example Yellow Widget) we then would need to research all the keyword combinations we can think of for Yellow Widget and find out what the level of traffic is for that. There are a number of independent tools to do that and both Google and Bing’s Ad engines have features that we would use to get this type of information and find out the popularity of the terms and find out the competitiveness / cost of marketing to these keywords. Some may be expensive and very popular, whereas others may be on the long tail or niche end of things (don’t ignore the niche especially when it takes little effort to market to them and they are usually very targeted motivated visitors).

 Audience

Another one winning the popularity contest with our partners is the Audience analytics. The first portion has a strong application to Adwords and is Google’s answer to Facebook’s targeting system. It is tied to your topics, interest, and audience settings in Adwords if you use them. You have demographics like age, gender, and parental status, then interest settings like Affinity (what their lifestyle is like for example – “Travel Buff”), In-Market Segments (People that are likely to purchase a product – someone consistently searching Real Estate for example) and Other Categories (this is your audience’s general interest and browsing habits – News would be an example). Clearly, you can imagine that understanding your user’s lifestyle, purchasing habits, and general interest could help you to tailor your general marketing campaign to them, write content for them, and make adjustment to your Adwords strategies.

Further down the list are categories that are pretty self-explanatory. They include: Geo (location and language of your users), Technology (Browser / OS and Internet Provider), Mobile (type of devices viewing your site), Custom Variables (recording specific interaction on site), and Benchmarking (which allows you to compare yourself to others in your industry that have opted in – are getting more powerful lately and we are looking at this one more closely). These are all quite useful if you are creative and are looking for it. For one client we saw a drop off in traffic that was suspicious. It turns out that by using the technology report, we were able to see that people using Internet Exploder 11 (19%) were dropping off consistently. The error was that one of our plugins was causing IE 11 users to not see the page at all, so we fixed the plugin and fixed the error. Traffic returned. You can count on analytics to help you out.

We also want to mention that Audience is pretty consistently useful. User flow is a graphical representation of how users go through your site. This place offers visual details and we can really segment your users’ behavior among the different dimensions in a visual manner. Some of our partners have found this to be easier to understand and make decisions with. There is also a version of this customized in the behavior section under behavior flow and it works in much the same way. You may find yourself spending most of your time in these areas depending on your style.

 Behavior

This is the area that defines how people move through your site and their interactions. It is important to pay attention here as it can inform you on what is working for the user and what is not. Reviewing Site Content is one way to get insight. There you will see all your pages, how many people visited them, how long, who bounced and who exited. You can go deeper through the landing pages, exit pages and the little used (by us at least), content drilldown to really find what is causing the issue or the success. It is more than just a quick snapshot of your pages in a professionals hand, like Analytics as a whole there is much actionable knowledge here.

Site Speed gives you an overview on how your pages are performing speed- wise. As mentioned in the SEO section, this is critical and will affect your ranking as well as your user’s enjoyment. You can use this to make judgments on how your server is performing (maybe you need to upgrade), find those problematic pages that are slower than others, and perhaps the most useful are speed suggestions. With speed suggestions, Google Insights will load your page and then give you specific action to speed it up, which can be very useful.

Events show how people use interactive elements in your site/app. For example, many websites contain video players, games, and other interactive experiences. Simply tracking page views doesn’t allow you to capture these kinds of interactions. When a user interacts with a video player, no page view is generated. Other examples of interactions that don’t generate page views include Ajax-based activities, file downloads, and clicks on links that take the user to another site. This definition from Google itself pretty much sums it up and Successful uses this mostly with video.

Out of the remaining options, we are huge fans of Experiments – you’ll want to read what we have to say in our A|B testing section. The bottom line is that this is where you view and analyze your testing data. Adsense is set for those serving ads on their site and gives them a number of useful metrics to judge their success. In-Page analytics will give a visual representation of how your users interact with your page including click areas; it is very useful for the visual user. Think of it as an Analytics overlay for your pages. You will need to be using the Chrome or Firefox browser and have access to your Analytics account.

 Conversions

The last aspect of Analytics we will cover is conversion, and it is one of the most important ones. Again, a conversion is action that a user takes which you define as important to your marketing goals (like signing up for your newsletter). We will need to set up these goals and add a bit of code to those pages. Once done, then we can see the conversion rates and how often it completes. Compare that with the traffic on the page and we can determine if the page is behaving successfully or if we should adjust it. Part of conversions is Ecommerce.

Imagine setting a goal for each of your products then having it track how many were purchased, the purchase amount, and the billing location information (not Credit Card) of the visitors. Time of purchase is also key since it lets you know the average number of visits (sessions) and average number of days it takes for your users to make a purchase. One might use this data to decide that if it takes too long for the users to make the purchase we might need to set up something that compares your product to your competitors to get possible delay causing questions out of the way.

We can also use parts of e-commerce data to determine what product might be needing a discount or if given time (and liberty) see how the consumer behaves when we adjust the price of a particular product. Do they purchase at the same rate when it cost five dollars more? A Successful mind imagines clear ways to predict profit with this type of data.

 A|B Testing / Experiments

A|B Testing, or Experiments, is critical to measuring and refining pages on a site. Successful considers this a must- have for the proactive partner. Essential are the two pages we set up: the control and the experiment page. We then set the same goal or user action that we want to see accomplished. Once the experiment has begun we select what percentage of users we want to participate in the experiment. Then fifty percent of them will see the experiment page and fifty percent the control (these percentages can change), the page that gets the most goal conversions is the winner. We then, just like ads, set the old experiment page as the new control and make a new experiment page off that: rinse, wash, and repeat, constantly growing the effectiveness of those pages.
We can experiment with any number of items on the page. It could be some layout element, a picture, headlines, body text, call to action or even something as simple as a button color. There was one famous experiment where changing the color of a button from red to green created an eight percent increase in conversions. Something that simple can make a big difference, though we recommend making it a big obvious change – nothing so subtle that the user would not really notice the change if they had seen both. It needs to matter. Generally, it is about goals. We do want to mention that we also use it for other more subtle things like bounce rate, time on page, and so on. If you have a low performing page, A|B testing is a good place to start to fix it. As a side note, we have used experiments to settle arguments in a marketing department – one I remember right off concerns voice used in the text of the site. The director wanted a very clinical, professional textual voice on the page, whereas the rest of the team wanted a very informal, friendly voice. The rest of the team won with an eighteen percent increase in conversion boost. The team got a free lunch party from the director. Testing can do a lot to eliminate guess- work.

There are a couple of key rules when you think about making a test. We would want to test one of your more important and preferably higher traffic pages (it is difficult to test all your pages if your site is medium to large and you need the higher traffic pages to gets test results faster and more accurately). Again, the changes need to be loud. We would want to test only one or a couple things on the page to not distill the results too much. This, of course, applies generally. We have had situations where we just built the page and made radical layout and image changes to the experiment just to find the best general look and feel. After that, when we kept testing we kept the changes small and bold. There can be more involved experiments depending on how we set them and the attribution channel we are aiming for. We hope, however, that you see how critical experiments can be to your Success, no matter how big or small you are. Successful has a package just for this so let us know if you would like to enable it.