Let’s start at the beginning. Step one to any good plan of attack is organization, though so often it is overlooked in favor of acting quickly and producing speedy results. A large part of being organized is having proper infrastructure for the task you’re trying to accomplish.
Building a house requires blueprints and a strong foundation. Speed through those things and you’ll have a funky-looking structure that could crumble under a stiff breeze.
So let’s start off on the right foot with our metrics and analytics planning and give it the dedication it deserves this year.
(Also, don’t forget to take our survey for our new MetricShift section!)
1. Setting up goals
Every organization has different goals that they want to achieve with analytics. Take some time and list out your goals. Here are some questions you can ask yourself to get to started:
- Who are your funders and what do they expect from you?
- Which parts of your business are you trying to improve with analytics? Public-facing operations? Internal or organizational operations?
Once you’ve got a good list of goals together, share them with your peers, your higher-ups, and your colleagues. You might think you know what your company’s goals are, but including people in different departments in this process can help you brainstorm angles that perhaps never occurred to you.
- Increase readership on the site
- Figure out what types of stories engage readers the most
- Cut out things that aren’t working to save money and begin experimenting with new processes
2. Understanding different metrics types
As tracking analytics continues to grow more and more mainstream, the deluge of data available to you will not slow. On the contrary, it will continue to grow as we start caring more about qualitative measurements like engagement and real-world impact.
But don’t get discouraged. We’re here to help cut this confusion.
A good place to start is with your list of goals. See if you can understand your goals in terms of these main metrics types:
If this list looks familiar to you, it’s because this is how Google breaks down the metrics that they provide to you via the Google Analytics dashboard.
Audience metrics have to do with, you guessed it, your audience. Are your readers predominantly male or female, what other interests do they have, where are they coming from, how often do they visit your site, what devices and systems do they use to access your content.
Google Analytics is a good starting point for answering some of these questions, but other services can give you additional information.
Acquisition metrics have to do with how your audience gets to your content. Most of the big analytics platforms have at least three acquisition types, also known as referrers:
- direct: traffic that comes from someone typing your website directly into their browser, or from bookmarks
- search: traffic that comes from a search engine like Google, Yahoo or Bing
- social: traffic that comes from any number of social sites. Each platform will have their own list of sites that they consider “social”.
Other referrers available from various platforms include:
- email: traffic from links in emails
- internal: traffic from other pages on your own site
- paid: traffic from links that you have paid to promote on the internet
- other: and the elusive other bucket that will basically take all the traffic that does not fall into any other bucket.
Behavior metrics have to do with how your audience interacts with your content once they get there. If we’re talking about your website, this would be pageviews and time on page for specific content or content types. If we’re talking about social media, this would be impressions, likes, and retweets.
3. Analyzing and tracking your goals
Below, I’ve analyzed the goals we touched on above to illustrate how you can understand and define goals in terms of the main metrics types. Each goal requires its own processing, however you’ll notice that our goal tracking methods all follow a similar outline:
- define your terms
- identify your main metrics type
- compare different types of content to identify strengths and deficits
- make data-based decisions
Increase readership on the site
Right off the bat, “readership” tells me that I’m interested in my audience metrics. And “increase” tells me that I’m specifically looking for new users.
In order to attract new users to my site, I need to understand my existing users:
- Are they mobile or desktop users?
- Internet Explorer or Chrome
- Apple or Android
- How often do they visit the site?
- How many pages do they look at, and
- How long does my content hold their attention them?
In order to grow readership, I need to turn new users into returning users but continue to attract additional new users in the process. So the question to answer becomes: where along the way are my new users falling off?
I want to compare my two groups of users, new and returning. Where are the differences? Maybe most of my returning users are accessing the site via desktop. Maybe most of my new users are on Internet Explorer. Perhaps I notice that returning users come to my site via social over more than any other single referrer.
It is important to note that when we are comparing two different audience types’ behavior, we are interested not in totals, but percentages.
While it may look like we’re attracting an equal number of visitors from both categories, it’s extremely important to realize that 60 percent of our returning visitors, versus only 30 percent of our new visitors are coming from social. So once a user becomes a returning visitor, they are pretty likely to be pushed to your site via one social platform or another.
With this information, you can more accurately make statements like “We should invest in growing our social media presence, rather than direct marketing or online ads, in order to get a more dedicated site following.”
Figure out what types of stories engage readers the most
An initial look at this goal tells me that I’ll be looking to behavior metrics, since those are the ones that have to do with how people interact with stories.
An important first step in figuring out how to track this goal is going to be defining what “engagement” means to your company. Maybe to you engagement means that the people are reading the full story. Maybe it means that they are exploring the site. Maybe engagement requires commenting, signing up, or purchasing something on your site.
Defining engagement is a complex issue, one that we will certainly write more about on MetricShift. But we’ll save that for another time.
Our next step is to define our content types. Are you interested in knowing whether environmental stories engage readers better than political stories? Maybe you want to know whether people are consuming more audio or video stories?
Defining content types is something that is treated very differently by each analytics platform. Google Analytics doesn’t exactly make it easy, requiring some knowledge of code to implement the necessary structure. But it is free and as with most Google offerings, if you can learn how to use it, it will give you a lot of data.
Parse.ly makes things much easier. You assign keywords to your articles via your content management system. Parse.ly then takes those keywords and creates content types from them so all you need to do is login and navigate to your content type to start analyzing.
Once you know how you’re defining engagement and you have your content types established, you can begin to more effectively compare your content. Perhaps audio stories have much higher time on page than video stories. Or your environmental stories are showing a large exit rate, indicating that people are leaving your site from these pages, instead of staying on to browse more content.
Armed with this information, you can start drilling down into the content that is working and the content that isn’t. Maybe you’ll notice that content with more interactive elements is confusing your users and making them leave your site faster. Or maybe content that summarizes its findings in a sidebar keeps readers on the page for longer.
Cut out things that aren’t working to save money and begin experimenting with new processes
Looking at this goal, I notice again that I’m going to have to define some terms. What does it mean for content to “work”. Content that “works” is likely content that engages, as we stated above. It’s content that produces a desired result, such as a user signing up for your newsletter or someone commenting on a story that they read.
So at its core, this goal is very similar to the last one. We’ll be looking at our content types to see which are producing engagement and which aren’t. We are then in a position to cut or reduce time dedicated to the failing content types and instead try something new or something that has proven to get us our desired results.
4. Keeping up with what’s new and asking questions
I mentioned earlier that every organization has different goals as well as different ways they’re able to achieve those goals. Even with articles like this, it can be hard to understand how you can get your company where it needs to be with tracking analytics.
WE WANT TO HEAR FROM YOU. We have a survey for readers to get feedback on our new MetricShift section, launching in the coming weeks. Please take the quick 3 minute survey for a chance to win an Amazon Echo!
Also, join our Facebook group and let us know what you’re trying to do, what’s working for you, what isn’t, any success or failures that you think might help someone else develop their metrics plan.
In addition to our Facebook group, we will be starting bi-weekly Twitter chats aimed at facilitating discussion around working with analytics.
If you’re not a Facebook or Twitter native, drop me a line at email@example.com.
Alexandra Kanik (@act_rational) is the Metrics Editor/Curator for MediaShift.