Defining KPIs and Measuring with Online Analytics

So far, we have taken a high-level view of the website property’s analytics. As we’ve seen, there is plenty of work to be done. However, let’s pause and parse out a framework that will help us move the website forward.

Drowning in Data

The temptation with such powerful analytics tools is separating the signal from the noise. With just a few clicks, you can quickly overwhelm yourself with data. As you familiarize yourself with the interface, suddenly an infinite number of options start to present themselves for ways to slice the numbers. This gives rise to the phenomenon known as “paralysis by analysis”.

Digital marketing gives us tremendous feedback in real-time on how our marketing efforts are working. While traditional marketing like print ads or television had a longer feedback loop, the data stream from idea to revenue was much more obscured. In a sense, identifying attribution for a sale was near impossible. We can now start to close this loop.

However, the tremendous power with digital marketing analytics comes at a price. The trick is no longer capturing a potential or current customer’s engagement – as we’ve seen that is readily available.

Rather, it falls back to an age-old marketing problem: defining what is success.

It cannot be overstated enough that the upfront work of defining a Key Performance Indicator, as driven by the strategy, is the most important part of web analytics. In my years as a digital marketer, I have been constantly amazed at how little effort goes into this exercise when working on a campaign. Assumptions are baked into these efforts, media is purchased, content is created, and technology is built – all without little consideration into what success looks like.

It is a conversation well worth having with the decision makers on your project.

The fascinating thing about KPI’s is they cut right to the heart of the mission of the brand and the business. In legacy companies with consistent sales, KPI’s can be especially murky. After all, “we’ve always done things this way” is the common response. In startups, it’s even worse given the absence of a viable business model and historic data.

An old Chinese proverb states: If you do not know where you are going, any road will get you there. In the case of web analytics, all these roads have signs pointing different directions and can take you anywhere.

A Framework for Analytics: Define, Measure, Question, Experiment, Compare

To make sure you do not wander in your online efforts on infinite roads, you must do the hard work of defining KPIs. (The Strategy section covers this in detail). In Avinash Kaushik’s excellent book Web Analytics 2.0, he suggests a similar framework (Kaushik, 63).

A quick tip: Go the extra mile to make these KPIs as simple as possible.


A poor KPI: “We want to grow Visitors on the website”. For reasons stated, we know this is way too fuzzy. Unique Visitors? Return Visitors? What does engagement look like?

A good KPI: “Want to grow Unique Visitors by 20% through organic SEO. We want this traffic to be targeted and engaged and we’ll measure this by shooting for an average Bounce Rate of 35%. We hope to achieve this goals over the next 6 months. We would like to see 2% of our site visitors opt-in for our monthly e-newsletter for our nurturing efforts.” This shows what kind of visitors, how much growth, how they’ll arrive, how they interact, and how long we’ll run this experiment and what a site conversion looks like – all easily trackable with online analytics.

Key Takeaway: Online analytics are only as useful as the thought that goes into defining the goals and objectives upfront. Without this preparation, the data becomes overwhelming and it’s easy to lose your way in all the noise.