A Web Analytics Primer

Imagine, if you will, a letterpress company. You are the owner, and you are very excited for your new website.

After months of watching the progress of this new website, launch day has finally arrived. You have donned your frock coat and spats. You are prepared to pull a Scrooge McDuck and dive headfirst into the giant vault of money that will surely follow. You announce your new website to your existing customers and family, and you’re actually getting great feedback from your followers on Facebook. Slap on the goggles. Let’s get ready to dive.

But no. Because there has been no actual action from your users. The goggles come off. There will be no vault diving today.

Clearly this is not what we envisioned. We were expecting a flood of attention, not this inactivity. But, we have one thing going for us: this is the internet, where we can see the pages our users visit and can quickly make assumptions to respond to their actions. Or, in our case, their lack of action.

This is the internet. And we have analytics on our side.

Determining user outcomes.

The good thing about this story is that our website can adapt to the needs of the customer and our business. Yet before we can worry about specific analytics, we have to understand what we expect out of our users. We must dive into our web strategy.

(And no, the strategy can’t simply be “Have a website.”)

Web strategy is shorthand for the decisions that guide web creation and promotion. It is based on business objectives, customer demographics, marketing efforts, and potential opportunities, and it serves as the basis for our content strategy, analytics strategy, website design, development and marketing decisions.

At minimum, a usable web strategy must answer two important things:

• What does the user want out of their visit?
• What does my business want out of a user’s visit?

We are going to assume that for our letterpress company, our goals are to allow users to:

• trust us as a viable letterpress option
• request a quote or ask a question
• interact with our company to create a long-standing relationship
• follow us through other means for easier updates

While still tying to our business goals of:

• becoming a leader in the printing space
• selling more letterpress projects

Where did these goals come from? They came from tons of interviews and research and important meetings – meetings that gave us a long list of web demands and features.

We want our users to view our services, browse our portfolio, and request quotes for future projects. We believe in being thought leaders, so we have a blog, a Facebook page, a Pinterest page, an email newsletter, and a Twitter account. We want repeat customers to have the best price possible, so we run a loyalty program, a referral program, and run specials on certain types of paper stocks.

We have made thoughtful and logical strategic decisions about what goes on the site. Yet, we don’t know if these tactics are going to work. We must start measuring their effectiveness.

The analytics toolbox.

Before we can get started tracking our site, we need to know what we’re looking for – and what will help us get those answers.

The main things we want to know about our website are:

• How many people are visiting the site?
• What do visitors do when they get here?
• How did they get to my site?
• Are they taking actions toward our goals – or theirs?
• How is our marketing changing this traffic?

Although there are a number of options out there for website analytics we recommend using Google Analytics because it is very easy to install, robust, has great documentation, and is free.

(If you are not already using Google Analytics, installing it is pretty easy. You simply use [or create] a Google account and then add the Google Analytics tracking code to the source code of your website. We suggest using the new Universal Analytics tracking code so that you will have all of the newest available features. Deeper instructions can be found via Google Analytics Support.)

Along with Google Analytics, we rely on an add-on product: Webmaster Tools, which is useful to learn more about what people search to find you and where you rank. Webmaster Tools also has a number of other features for monitoring and managing your website, but for the purposes of this post we are mostly going to be focused on the search queries and ranking.

For more advanced SEO analysis and competitor comparison, we like to use a service called Moz, which allows us to view competitor information, social information, keyword information across Bing and Yahoo, and get on-page optimization analysis for selected pages.

Website strategy + analytics setup = analytics strategy.

The key to using analytics is understanding which are worthwhile and which are a waste of time. Our letterpress business already has a solid web strategy, so our next step is to develop an analytics strategy that will dictate which metrics receive the bulk of our attention – from key performance indicators (KPIs) and conversion values to goal and event tracking.

Defining success.

Goals. Goals. Goals. Goals, as mentioned before, are different for nearly every website, and are always dictated by the overall web strategy – and different goals mean different measures of success. In other words, we can’t all rely on page views or bounce rate to tell us the whole story.

For example, an e-commerce site can use raw transaction data to determine which product pages are most popular, when users are abandoning the shopping cart, and which promotional efforts are leading to the most transactions.

On the other hand, non-e-commerce sites require a deeper view of success. Here, we look back at our definition of a successful visit and choose metrics that help us measure that definition. Examples include visits to the contact confirmation page, call-to-action clicks, document downloads or promotional video views.

Choosing key performance indicators (KPIs).

Using our definitions of success as a guide, we then choose KPIs that are helpful in determining our progress towards achieving our goals and the overall improvement of website traffic.

(For a great (large and free) resource on KPIs and their various forms, check out The Big Book of Key Performance Indicators by Eric T. Peterson, available at Web Analytics Demystified’s website. [Click “Support” and scroll about halfway down the page; you’ll have to give your email address, but it’s worth it.])

An example would be if one of our website goals was to get more visits to our “Contact” page, which presumes an intent to contact or visit our physical location. In this example, our KPI would track visits to the “Contact” page over time.

Some KPIs we commonly use are:

• Non-brand keyword queries – Keyword lists are more useful when we omit variations of company, product or service names. We can do this with an advanced segment through Google Analytics. Overit’s post “Using Advanced Segments in Google Analytics” gives a good overview.
• Visit acquisition distribution – We use this to monitor changes in how users are getting to our site. Google has built in channel groupings of Organic Search, Paid Search, Referrals, Direct, Social, and Email. You can reorganize the channel groupings in your Analytics account as you see fit.
• Top-20 referrers – What 20 websites are referring the most people to our website? How good is the traffic? What are their conversion rates?
• Contact confirmations – Often times a website is used to generate contacts to your company and setting this as a goal will allow us to gauge website and marketing effectiveness.
• Goal completions – Setting goals in analytics is key to understanding what content, marketing, keywords, and design elements are most effective in achieving our goals.

Tracking events.

Sometimes our site has content or actions that don’t trigger anything meaningful in analytics, but they are very meaningful to our conversion process. Examples include simple steps like clicking a call-to-action button, in-site ad or promotion, or whether they watched our video.

Google Analytics provides us an opportunity to track these activities through Event Tracking and Virtual Pageviews. Using these tools, we can identify “non-conversion” actions that might be useful in our analysis of site performance.

• Event tracking: The exact syntax and method for this is different if you are relying on standard Google Analytics or if you are using the newer Universal Analytics tracking code. Google has provided documentation for starting with or upgrade to Universal Analytics. If you’re tracking an action as part of a larger story through a Visitors Flow, Reverse Goal Path or Goal Flow, you’ll want to rely more on Virtual Pageviews, which are very similar to setting up a Event Tracking but with a slightly different javascript syntax.
• Virtual pageviews: If you do choose to use Virtual Pageviews it is important to know that your site pageviews will be inflated. If you are actively tracking pageviews, you may want to use an Advanced Segment to filter out any of Virtual Pageviews. Grouping all Virtual Pageviews into a subdirectory (e.g., /virtual/) makes it easier to filter those within your Advanced Segment.

If you’re tracking one event, Event Tracking is the most logical way to go. Via a snippet of javascript, you attribute a category, action, label and value to a movement someone takes on your site – whether it’s clicking on a link or opening a file.

Applying strategy.

So you’ve set up your analtyics strategy. Now what?

The first step is monitoring. If you don’t actively pay attention to your analytics (or at least assign or hire someone to do so) then you might as well not even bother.

This doesn’t mean you need someone pouring over analytics for eight hours a day. Unless you are aggressively marketing your site, or constantly updating its content, we feel that starting with a simple monthly analysis is perfectly fine. At Blend, most of our Analytics reporting is done monthly, with occassional weekly monitoring for clients with additional AdWords campaigns to determine which ads and keywords are providing the best traffic and conversions.

We begin our monitoring with a quick glance-through of Google Analytics four main predefined categories: Audience, Acquisition, Behavior, and Conversions. If anything stands out, we compare that data to the previous period, looking for improvements or big surprises.

Now that we have a general idea of what was going on with our website over the previous period we can move into a more detailed look into the specific categories.

Audience.

The Audience category focuses on who is arriving on the website, and answers questions such as:

• How many people came to our website?
• How long did they stay?
• How many pages did they visit?
• From where did they visit?
• How many were mobile users?
• How did they progress through our site (visitors flow)?
• What are my users demographics?

Note: It is important to cross-check any significant changes of Audience metrics with Acquisition metrics so you know how (and maybe why) the changes happened and what type of users contributed to the change.

Acquisition.

This is one of my favorite categories on which to do analysis because it hits on something that has always interested me – and our clients: how did my users get to my website?

For example, we commonly track both top-25 non-brand keywords and top-20 referrers. (Be sure to also toggle “Landing Page” as a secondary dimension to see where your top-20 referrers are entering.) These two metrics give us a pretty good idea of how people are finding us and which methods are most effective.

Acquisitions includes a Campaigns channel that helps us determine whether digital marketing efforts (PPC, email, banner ads, etc) are working (or not working), while the Social channel helps us determine which networks and types of posts are most effective.

By tracking our conversions or events, we can see which referrer, keyword, medium, campaign, or ad is outperforming the site average. This gives us a good idea of where we stand, where to focus our resources and where we need to improve – and that’s the whole reason we’re doing these analytics in the first place.

Finally, you can also evaluate the entire scope of your digital marketing through the Paid Search, Social and Email channels. Judging these metrics based on conversion rate is very helpful in determining where a business spends its digital marketing budget.

Behavior.

The Behavior section of metrics contains important information about the content of our site and how users are interacting with it, giving us a look into what content and actions are most important and are engaging users the most effectively. Some of the common KPIs we like to use are: top pages, top-20 landing pages, top events, top-10 exit pages and site search terms.

Using that data, we can begin to make educated decisions on writing content, emulating the more effective content areas and content types or eliminating content that is performing poorly.

This is where Events can test the effectiveness of the paths a user might take to get to a certain chunk of content. For example, if you have multiple call-to-action buttons and promotional graphics pointing to the same page, you can use event tracking to find out which works best.

What’s more, you can actually test content against itself, thanks to Google Analytics’ Content Experiments – something you might know as A/B Testing. (I won’t get into the technical details on setting up A/B testing because Google provides a lot of documentation on how to do it, including a YouTube video on Content Experiments.) This is a great way to test improvements while retaining your original content, randomly subjecting users to one of several paths in order to determine the most effective or successful content.

Conversions.

Finally, Conversions are the most important set of metrics to help us accurately gauge the effectiveness of our website and digital marketing.

Our letterpress business might want to track “users who contact the company via a contact submission form” as one of its goals. If we want to see which of our marketing efforts are most cost-effective, we can compare the number of contacts we receive from each marketing avenue to how often we actually convert this person into a customer. If we combine this with what we know to be our average sale to be we can attribute an actual perceived value of each contact we receive. For instance: if 100 users fill out our form and, of those 100 leads we convert 30 people, our conversion rate is 30%. If our average sale amount is $2,500, we multiply that by our conversion rate (30%) and we get a conversion value of $750.

Using our conversion value we can determine which pay-per-click campaigns, organic keywords, email promotions or website referral sites are producing the most value. This helps us make more educated decisions on where we spend our marketing dollars.

From a content perspective, we can look at the Pages/Visit and Average Visit Duration to see if there is any correlation to how active someone is on our site to the rate of conversion. If more active users tend to convert, we can give them more relevant reasons to stay on the site longer. The amount of time we spend generating content now has a quantifiable value instead of just a gut feeling.

The discussion about conversions can go down many different paths, from Reverse Goal Path and Goal Flow tools to integration with marketing automation system like Hubspot and Marketo or CRMs like Salesforce. But they all end on one thing: placing real value on the often-fuzzy world of web content and marketing.

Putting it all together.

It’s true – when we create a new site, we may not know exactly why things aren’t working. Our thoughtful strategic ideas are just that – ideas – until we start putting actual numbers behind their effectiveness.

Thanks to our data, we can learn that conversion rates increase as people become more engaged in our brand. This proves the importance of our newsletter, which allows us to share specials and inspire our existing customer base.

Thanks to our data, we know that our conversion rate through the quote page is 30%. This gives us a real number to work with.

Thanks to our data, we also get a lot of website traffic from shares by our followers on social networks which lead to conversions. This justifies time spent on Facebook and Pinterest.

Our business goals help us generate our web strategy, and our web strategy influences our analytics strategy. Our analytics strategy guides us in monitoring our site, with which we can make improvements and adjust our web strategy.

This cycle of constant improvement based on actual data makes it easier to justify the time and expense of improving our website by showing us actual results instead of just making random assumptions.