The datapine Blog
News, Insights and Advice for Getting your Data in Shape

Top 6 Google Analytics KPIs: A Professional Guide to Modern Web Analytics

Women looking at a Google Analytics KPI dashboard on a laptop

Google Analytics (GA) has become the de-facto champion for free web analytics, as it monitors dozens of  Key Performance Indicators (KPIs) for your business. These web analytics KPIs (measuring the website traffic and visitors’ behavior), analyze all available web data with the purpose of understanding and improving web usage. With datapine’s Google Analytics Connector these KPIs can be measured, monitored and analyzed with ease. You’ll be able to ask dynamic questions, drill up or down and create beautiful custom Google Analytics dashboards that you could never build in Google Analytics alone.

Important web analytics metrics will vary from industry to industry depending on the type of company and goals. However, despite the discrepancy of what makes a KPI important for a business, there are some website KPIs that all businesses should be monitoring. Here are the top six categories of Google Analytics KPIs we think every company should focus on along with real world case studies of why they matter.

Exclusive Bonus Content: Advanced Google Analytics Features
Take Google Analytics to the next level and learn how you can create advanced data visualizations and dashboards with a few clicks!

1. Your Audience

Understanding who is visiting your site, if they converted, and where they came from is an incredibly valuable information in order to know which channels are showing results and consequently, where to put more resources. To do so, you need to see how many (unique) people are visiting your website and their demographics including age and gender, as well as geographic location. It can also be valuable to see how often your audience returns to your site.

Sessions and Users on a Weekly Basis

Number of Sessions and Users by week - datapine

A couple of years ago, Google specified their terminology: Visits are called Sessions and Unique Visitors are referred to as Users (which can be split into New vs Returning Visitors).

The number of sessions counts every user coming to your website in a certain timeframe. New users are the people who are coming for the first time during this timeframe, while returning are those who already connected to your website.

New Visitors vs Returning Visitors on a Weekly Basis

Graph: New and Returning Visitors on a weekly basis - datapine

GA provides you with two rows of data concerning your users: New and Returning Visitors. However, looking at the data, you may find it odd: our total number of users for the Week 25 is 11,047, while the new visitors account for 9,871 and the returning for 2,234. But 9,871+12,234= 12,105, so where does the 11,047 come from?

In that case, it is important to know how Google counts the users and sessions, and to be aware that the users might be doubled as both new and returning visitor, within a given timeframe. Let us take 2 use-cases:

– If a new visitor came during this timeframe and returned during this same timeframe, he/she would be counted twice, both as a new and as a returning visitor.

– If a visitor came before said timeframe, and returned during that timeframe, he/she would be counted as a returning visitor. However, even if it is a “first” during that date-range, it still wouldn’t be counted a new visitor because he/she already came earlier.

So we see that the the sum of returning and new visitors won’t always match the total number of users, but always will sum up to the total number of sessions.

Age Groups and Gender

Getting to know your users thanks to their demographics is also a great advantage you can take out of Google Analytics KPIs. It will help you in refining your strategy by targeting your audience better. GA will provide you with demographical data on your audience’s age, gender and households income (the latter is available only in the U.S.).

The age groups Google provides are the following: 18-24, 25-34, 35-44, 45-54, 55-64 and 65+, so basically ranges spanning over ten years. The data you will read will be enlightning to know which category is the most valuable to your business. Drilling down into the data will let you know how different age group interact with your website and convert. Then, splitting your audience by gender can help with content planning for your blog/website, or product choices.

Let’s say that your business had two kind of audience, working fathers and retired people. You will then create different advertisement targeting each of these two groups as their channels of communication and the approach are not the same.

  • Sessions per Age Groups

Pie Chart representing basic a Google Analytics KPI: Sessions per Age Group - datapine

  • Sessions per Gender and Age Groups

 Chart- Sessions per Gender and per Age group - datapine

By monitoring who is coming to a website and where they are coming from, companies can determine their strengths and weaknesses in attracting new users and develop strategies to convert those users into customers.

2. Traffic Sources / Channel Grouping

Traffic is one of the most important measure of your website’s success. Without traffic, you cannot analyze or promote anything!

However, there are different ways to focus on your traffic. Having the highest number of sessions might not be an ideal website KPI to fix, if the bottom-line is to convert those users. If you have a lot of traffic leading to just views of your website, then it does not matter if one or one thousand users came. If you want to spread a message solely, then high traffic is a good thing. To an enhanced traffic analysis, combine the number of sessions with other web analytics metrics like the average session duration or the bounce rate – we will see these measurements later in this article. For now, let’s focus on the origin of your traffic.

Number of Sessions by Source / Medium

How people find your website is an invaluable web analytics metric to measure. By understanding where your website traffic is coming from, it is easier to make informed decisions about where to focus your time and money.

As Google states itself, the Source is where users are before seeing your content (the origin: search engines and other), while the Medium is the general category of the source: “organic” for the users coming from unpaid search, “none” for the direct traffic, “cpc” for paid content, or “referral”. For both of Source and Medium, names are case sensitive: for instance, Facebook sessions appearing as facebook.com, m.facebook.com will be treated as a different source/medium.

Number of Sessions by Channel Groupings

This is why the use of the Default Channel Grouping of Google Analytics is easier. It groups several traffic sources with the same medium. They are broader and more “user-friendly”, avoiding having the same origin several times. The different Channels you will find are the following:

  • Organic Search – people arriving on your website after a search on a search engine like Google, Yahoo or Yandex.
  • PaidSearch (CPC) – people coming to your website after clicking on an ad you made with Google AdWords, Yahoo Search Marketing or Yandex Direct etc.
  • Referral – users that came via direct links on other websites.
  • Social – users who land on your website via Twitter, Facebook, Google+, Youtube, Linkedin, Instagram, Pinterest, Quora, Reddit etc.
  • Email – users who click on a link in one of the emails you sent.
  • Direct Traffic – users that typed in your URL directly: this is a good indicator of how well people remember your brand.

Hereafter you can see a stacked bar graph illustrating the number of sessions by channel on a monthly basis, and the proportion of each channel to the total.

Stacked Graph with Sessions by Channel - datapine

Exclusive Bonus Content: Advanced Google Analytics Features
Take Google Analytics to the next level and learn how you can create advanced data visualizations and dashboards with a few clicks!

3. Page Tracking

There are other Google Analytics KPIs available, that are tracking the behaviour of users on your website: if they found what they were looking for, how long they stayed, or if they left your site immediately. Before examining how these metrics can interact with each other, let’s go over their definition.

Bounce Rate: refers to the percentage of single-page sessions on your website: when users leave the site from the entrance page, ie. after viewing only one page. In Avinash Kaushik terms, bounce rate refers to customers following the motto, “I came, I puked, I left.” . Coming from the guru of analytics himself, that is a quite clear statement.

Average Pages per Session: pretty straightforward, it talks about the average number of pages a user sees during a session on your website.

Average Session Duration: refers to the average time users (including all types) spend on your site. To calculate it, Google divides the Sessions Duration by the number of Sessions.

Average Time on Page: refers to the average amount of time that all your visitors spent on a particular page. That’s a good metric to see if your visitors are reading your content. Google Analytics calculates it with each “hit” sent (a click to another page), and each “hit” carries a timestamp (day, hour, minute, second) that helps the calculation (mavenec.com explain it clearly in their blog post). However, Google’s calculation is flawed when it comes to the “exit” page, the last one of a session, that is not accounted for – but it tries to overcome this flaw by taking the number of exits into account. The formula they use is: Average Time on Page = Time on Page / (Page Views – Exits). For more details, you can read Analytics Edge, providing us with a very good article on these misunderstood metrics of Google Analytics.

Going further with these metrics

  • Bounce Rate and Average Session Duration

By knowing your Bounce Rate and Average Session Duration you can identify where customers are coming from, what they are looking for, and whether they found what they wanted from your site. Jason Squardo, executive vice president of optimization at ZOG Digital explains to Mashable: “a high bounce rate may be an indication that your content isn’t engaging or that your advertisements are misleading. In order to reduce bounce rate, it’s important to set up users’ expectations through the content — whether that be the call-to-action in an ad or headlines that match blog content…You don’t want users to land on your site and feel deceived or cheated” . Once you know these Google Analytics KPIs you can focus on improving the content and perhaps revise your ads to save you money and get higher ROI.

Warning: Bounce Rate can be a misleading metric when considered alone. If a website is so well designed that a customer lands on the page, find what he/she wants, takes action, and then leaves without visiting any other part of the site, the bounce rate would be high. So a high bounce rate does not always mean customer dissatisfaction.

  • Time on Page vs. Session Duration

Something most website owners ignore is that Google actually can’t measure the time a user spent looking at the last page of their website before they left. That is due to the way Google measures the Time on Page, but also because at midnight (00:00 hour), every session is broken and a new one starts.

Average Session Duration can be a tricky website KPI to track. Indeed, if a session doesn’t have many pages visited, the loss of the last page timing can have a big impact on the total. If a lot of the users are “bouncing”, ie. visiting just one page and then leaving – even if the visit was of 30 minutes – the session count is 1 but the session duration would be 0. Secondly, quite often people will leave the browser window open without actually interacting with your site. This will drive up the time spent on your site artificially. It has to be noted, though, that the default duration of a session is of 30 minutes (and ends after this time, to start a new one). This default duration can also be customized.

For the Time on Page, Google uses the time of the next page view to calculate the time you spent looking at the current page. So when a user is on his last page before he exits (closing window or tab), that final Time on Page is unknown – same applies for single-page visits that are not counted, even if the user spent several minutes on it. We see that the problem to know the Time on Page is because of exit pages. However, Google tries to counter-balance it in counting only the users that had an interaction with the webiste – ie., the bounces are not taken into account. This is the big difference between Time on Page and Session Duration.

To clarify how Google calculates, let us take a concrete example applied to  each KPI:

A user looks at 2 pages: he spends 5 minutes on the landing page, and 2 minutes on the second one. 4 other users bounce directly on the landing page: we don’t have any data from them.

We know that both of the Time on Page and Average Session Duration are not able to track the 2 minutes on the second page read by the first user. Still: what would each Google Analytics KPI give as a result?

Average Time on Page = Time on Page / (Page Views – Exits), where 4 bounces are not taken into account.

Average Time on Page = 5 minutes / (5 page views – 4 exits) = 5 minutes.

Average Session Duration = Session Duration / number of Sessions, where 4 bounces are taken into account with zero.

Average Session Duration = (5 + 0 + 0 + 0 + 0) / 5 = 1 minute.

In that case, with 80% of bounces, the average session duration will be 1 minute. This means in the calculation they take into account 80% of data that they actually cannot measure!

This is why both of these metrics have to be understood in their context and cannot assess user engagement. Evaluating them in parallel with other metrics like the bounce rate is a good thing to do. With them, you can compare different traffic sources and validate the “quality of traffic” as a low bounce rate and high average session duration / average time on page will lead with a higher possibility to “desired actions” (conversions).

Tip 1: It is better to consider Time on Page in general as a valuable metric. If, on top of that, the percentage of bounces on your pages is low (under 50%) it will be all the more accurate.

Tip 2: For high bounce rates (>80%), both KPIs are pretty inaccurate (one uses just a really small percentage of all session to calculate it, and the other one takes a lot of unknown data into the calculation).

Tip 3: What you can do to overcome the uncertainty is to integrate an Event (see part #5 in this blog post), to measure the time on bounce pages with small time frames (0-10 sec, 11-30 sec, 31-60 sec. for instance).

To have a better way to calculate content engagement with Google Analytics, read this Analytics Ninja’s article – they even provide solutions to calculate the real time on page!

Track Your Bounce Rate, Average Session Duration and Time on Page per Channel

Graph visualizing some Google Analytics KPIs: Bounce Rate, Average Session Duration and Time on Page by Channel - datapineOn this graph we parallelized 3 different web analytics kpis: the Average Session Duration, the Bounce Rate and the Time on Page according to the traffic channels. We see that the discrepencies between Time on Page and Session Duration are not as bad when the bounce rate is low, like for the Organic Channel.

4. Goal Conversion Tracking

The conversion rate corresponds to the percentage of visits that resulted in the desired action (buying, selling, signing up, etc).

However, conversion means different things for different parts of your website. Signing up for an email newsletter or making a purchase can both be counted as a conversion. Each company needs to set up their own “goals conversions” for Google Analytics to track. Goal conversions are the primary metric for measuring how well your site fulfils business objectives. Segmenting your conversions will help you hone the results and get a better understanding of your audience and its behaviour.

With these goals in place, you can see which channels brought the most conversions. Again, you can compare your channel specific conversion rates with the bounce rates / average session durations and you should see a correlation between low bounce rates, high average session durations and high conversion rates.

Number of Conversions by Channel last month

Graph representing some website kpis: Number of Conversions and Conversion Rate by Channel - datapine

On the chart above, we see that the channel performing the best in terms of conversions is the Organic one, as it has indeed the highest conversion rate and the highest total number of conversions from the three channels.

Ariat, an equestrian footwear and apparel brand, used Google Analytics to discover how often shoppers used the tools on their website and whether this was related to sales increases. Ariat used Event Tracking tags to find out what tools were resulting in increased sales. Over time, this data revealed trends in conversion/sales assistance. With this data, Ariat found the weak spots in their conversion funnel and made adjustments. Their conversion rate grew by 14% resulting in increased profit and less customer dissatisfaction. Conversion rates can show you when your business in healthy and when you need a reboot.

Conversion by Landing Page

Another way to assess your website’s efficiency is to track which landing page converts the most, if they convert at all. These pages are extremely important because they have a big impact on the general conversion rate of your website. Thanks to these insights, you can fine-tune your landing pages and monitor their performance. By knowing which are your “star” landing pages and your worst ones, you can learn best and worst practices from them and adjust the rest of your website accordingly. Be careful, however, to compare the same kind of conversions: if “clicking on a video” works very well on one landing page, “register for a trial” might not work as well on another one. Consequently, the conversion rate will be dramatically different.

Google is plentiful of articles and testimonies from companies/webmasters/bloggers working on their landing page and seeing an incredible increase in their conversion rate that you can read. Wishpond gathers five examples on redesign and content improvement that helped businesses invert the curb. For instance, WikiJob – a website aiming at helping interviewees – was struggling with paid conversions. From the original landing page, they added at the bottom in the place of tests to download, simply some “customers testimonials” on how it did help them carry out interviews and provided good practice. The result was of 34% improvement over the original landing page. The reason behind it is the “trust factor”: the most trusted source of information is increasingly becoming the “people like yourself”, so customer testimonials or reviews are a big advantages in today’s internet economy.

Exclusive Bonus Content: Advanced Google Analytics Features
Take Google Analytics to the next level and learn how you can create advanced data visualizations and dashboards with a few clicks!

5. Event Tracking

In Google Analytics, an event refers to the user’s interaction with a web page element that is tracked. For instance, an element could be a video, a podcast, a form to fill, a button, an image, etc. An interaction could be the loading of a video, the viewing of it, clicking on an image, abandoning a shopping cart or a form to fill, scrolling down, etc.

By default, GA only tracks events that generate page views (eg. clicking on an internal link). The other events generating nothing (like loading Flash Player or clicking on an external link), are not accounted for.

As we saw in the previous section, Event Tracking is also used to optimize conversion rates. For instance, you can track how often people click on a certain color of button with a certain text inside, and see which one works the best. On our table hereafter, we calculated the average conversion rate for all the texts and colors, and then compared the percentage difference of each text-color to the average.

Conversion Rates Tracked by Event - datapine

We see that the Text 1 with a red coloured button is outperforming the rest of the texts and colours with a conversion rate of 6% – that is, compared to the total average of 4%, a +50% difference in percentage. The least performant is obviously matching Text 2 with a green button, that has the lowest conversion rate of 2,5%.

La Tienda, an online food retailer, used Google Analytics to gauge the sales from two regions: Region A was places near enough the company’s warehouse to always get reasonable shipping costs while Region B was everywhere else. By installing Event Tracking onto the “Add to Cart” page and then using an Advanced Segments Custom Reports to separate visitors in Region A from Region B, La Tienda was able to track where their customers where buying from. The data proved that visitors from Region B were 48% less likely to buy if the shipping cost was too high. To combat this, La Tienda implemented a low, flat rate shipping rate and then monitored sales from Region B. After the test, the rate at which Region B visitors completed the shopping cart increased by nearly 70%. 

6. Cost Analysis

Last but not least, the cost analysis is the most important of the tasks that you will have to do, as it is what brings the cash in (or out).

Cost per Conversion has been dubbed “the only metric that matters” because you can calculate the profit and “true” return on investment when putting cost per conversions in relationship with your actual revenues per conversion. These web analytics KPIs make sure you aren’t paying more for a customer than they are worth. A poor ROI or Cost per Conversion can signify ineffective marketing activities. Tracking which are the channels the conversions are coming from helps narrow down problem areas and proves what channels are yielding the best results.

HomeAway Inc., an online vacation rental marketplace, identified keywords that lead to inquiries. Homeaway decided to increase spending for those keywords and used Google Analytics to track if the campaign was successful. The results: HomeAway saw a 23% increase in conversions for those keywords after implementing changes, compared to a prior period. By utilizing the Cost per Conversion KPI and identifying what channels the conversion were coming from, businesses can optimize their marketing budgets and bring in more revenue.

Conversions and Costs per Conversion in Q1 2017

Graph visualizing important web analytics kpis: Number of Conversions and Cost per Conversion by Channel - datapine

Costs, Profit and ROI by Channel in Q1 2017

Bar Chart: Costs, Profit and ROI for the Q1 of 2017 - datapine

Both of the graphs above provide us with important web analytics metrics, as they compare the performance of two channels -CPC and Organic- in terms of Return on Investment, of conversions and of costs related to these conversions.

Bring Your Google Analytics KPIs One Step Further

Once you have set and tracked all the metrics you need, interacting with them, visualizing and communicating them are the next important steps. But GA’s interface is not so user-friendly when it comes to tracking specific KPIs you need and combining them with others that are not in their default/standard masterview. This is why at datapine we developed an easier and more flexible way of working with our Google Analytics connector. You can create advanced interactive dashboards that will provide you with a wide range of possibilities to display the data you need to work on or the information you want to convey.

Here is an example that gives an overview of daily, weekly and monthly web analytics KPIs: Sessions, New Users, Bounce Rates, Sessions Duration, Goal Completions… Everything you need is tracked on this dashboard. For more dashboard like this one, you should read our post on custom Google Analytics dashboard examples and let yourself inspired!

Google Analytics Daily, Weekly & Monthly Performance Dashboard

The Google Analytics Daily, Weekly & Monthly Performance dashboard provides the perfect overview of your KPIs, and enables you to discover early-on if you are on track to meet your targets

** click to enlarge demo dashboard **

Exclusive Bonus Content: Advanced Google Analytics Features
Take Google Analytics to the next level and learn how you can create advanced data visualizations and dashboards with a few clicks!

With the wealth of information Google Analytics offers you, it can be overwhelming to choose which web analytics metrics are crucial for your bottom line. By using the top six Google Analytics KPIs above, you can start to identify areas that need attention because nothing measured, nothing gained. Learn how datapine can help you to turn your web data into insights – start your free trial today!

 

1 Star2 Stars3 Stars4 Stars5 Stars (109 votes, average: 4.84 out of 5)