When it comes to online analytics and more importantly, social marketing. What exactly constitutes success? Is it the number of people that visit your website? The number of people that visit your social profile every week? What exactly do these numbers mean? Are the number of people in your friends list even important? What about page followers?
What’s important, is to figure out what exactly is important. for example, a guest at an online store simply looking at products isn’t what we are hoping for. Adding an item to their basket is definitely better and paying for a transaction is the best outcome possible for this scenario.
Once you figure out what it is that you want your visitors to do, you must plan every interaction around that goal. Great, we have a plan and are putting it into action. Is it working? How do we know? Uncertainty is the worst feeling you can feel as a company owner. We need to measure to know if our plan is working, or needs to change. How do we measure though?
Analytics is a collection of measurements and can show variations in those measurements over time. With this data, we can measure the impact of any change that we make. What exactly makes a good measurement? What can we even measure? How do we measure? These are common questions that come from people new to the online marketing world. Most new website owners that I meet don’t have analytics setup. Those that do, are only collecting default metrics. None are using any tag management systems and rely on their web developers for every change. This doesn’t even address the multiple types of analytics. Most are using aggregate analytics multiple Analytics or Facebook Pixel. None that I have met are using focused analytics. Now that I have you confused let’s sort through these questions one at a time.
A good measurement is something that can provide insight into a users journey to the ultimate goal. Some are easy to figure out. Tracking when someone clicks the “Add to Cart” button is an important milestone in this journey. You should track someone signing up for your newsletter because it shows them engaging with the brand. When someone clicks on a link in a Facebook post, this shows interest in what you posted, tracking this is also important. Other less obvious things come from specialized knowledge, like knowing that over 50% of people will leave if your website page hasn’t loaded in three seconds. So, page load time is an important metric to track. We have to take the right measurements though, say your important page content starts at 10% of the way down the page and goes all the way down to the 75% mark. Measuring how many people scroll to the 75% mark is good, measuring how many seconds it takes to get there would probably be a better metric for measuring engagement.
We can measure all sorts of things, besides things already mentioned. Other choices are if the user copied text and what text it was, length of time on page, where traffic came from, how many times the user unfocused the tab, how long the tab was unfocused, how the user exited the site and more. These metrics are useful in multiple situations; but, not all are useful in all situations. An experienced analytics consultant can select the metrics that matter. The most important metrics to start with though are the ones that directly lead to conversions. Examples would be an “add to cart” click, “checkout” click, submitting a “contact now” form, number of visitors that leave with items in their shopping cart, payment type, etc…
Measurements are taken by sending an event packet to an analytics program. These programs take in every event and then stores it. These events normally consist of an event type, a label, a value and a category. The values can be of many types such as a numeric value, a yes/no value or a piece of text. We can aggregate the values in multiple ways depending on the type. Analytics events are sent in reaction to user, page and site events. Events traditionally needed to be added by developers to send information to analytics packages; but, that’s no longer the case.
A developer isn’t always needed to administer analytics anymore. There’s a new breed of service quickly making this more manageable called Tag Managers. Tag Managers need to have a small piece of code installed on every page of the website. After this, there’s no more need to install analytics code on the website anymore. Tag managers allow you to add your insights and remarketing code directly to your website without opening a single HTML file. Tag managers also come with pleasing interfaces that allow non-developers to install new packages. They even allow for event and conversion tracking with zero code in most cases. Event tracking is something many smaller websites go without because the technical expertise involved is too cost prohibitive for them. Although tag managers make managing analytics code much easier, training is necessary to keep it tidy. Tag managers are powerful and highly configurable, which means they can quickly get out of hand. You need to configure them properly from the start. Training though is usually a much more palatable cost for small or young companies.
Analytics are in two main types, aggregate and focused. Aggregate analytics is simply the values you have collected and represented as a single number or range. Examples of this are averages, sums, standard deviation and ratio. These metrics, as they are called, can give us insight into our audience. They can guide our decisions and are also used to measure success. Focused analytics is individual level tracking. We track a visitors entire engagement history to understand each engagement individually.
The last point I would like to address is Vanity Metrics. A vanity metric is a measurement that’s not useful in business decisions but is still treated as important. The biggest vanity metrics are Facebook Likes and Followers, Email Open Rate and Website Traffic. These measurements sound important but without context are indicative of the shotgun effect. This means that people are seeing your message, but many of these people are simply ignoring it. Fewer than 1% of people that like a facebook page, ever return or engage with that page again. So your efforts are wasted on reaching the very small subset of people actually interested in your message. This is a nightmare for marketers as it becomes difficult to create segments and keep engagement rates up with so much noise in the data. An alternative is the engagement rate, this is a facebook insights metric that counts all interactions with the post. Interactions show the user actively doing something towards your brand. Email open rates are another vanity metric that seems useful on the surface. Once you consider that most email clients are configured to not load images by default, this skews the metric in the downward direction. A much better alternative is click-through rate (CTR). The CTR is a better metric because it involves the browser where analytics are much more accurate, and it also shows clear interest and intent by the person who opened the email. Website visitors is probably the oldest of the vanity metrics, although it is also one of the least useful. The total number of users should be a good measure of your websites popularity. If you have lots of visitors and no engagement then you won’t be in business very long. Ideally, you would have every single visitor converting into a customer on every website visit. This ratio is important and you will need to maximize it.
By ignoring conversions, you will miss critical marketing information. Another metric that augments the total page visits is the bounce rate. Generally, a bounce is when a user visits only one page and spends only three seconds or less on it. Very high page visits are usually plagued by extremely high bounce rates. This means that people are leaving your website almost as quickly as they came. Which means that conversion will suffer greatly.
Analytics is a rich field of study and full of opportunities. Companies must use and understand them, not just collect them.