Data without context is dangerous.
Because business owners are willing to pay for these vanity metrics.
Vanity metrics are data that makes you look good but does not actually provide any valuable insights. The last thing a business owner wants to do is to spend their marketing budget on vanity metrics and the last thing a marketer wants to do is to just report numbers. Unfortunately, many fresh digital marketers or DIY Marketing business owners get overwhelmed by the amount of data available in marketing platforms and get stuck in a sort of data paralysis. The data is there but is never synthesized into something actionable.
The sheer volume of data provided can be intimidating but knowing how to read, interpret, and pull actionable insights from data is a skill that anyone managing marketing campaigns should be familiar with.
I’m going to show you how to read, understand and create actionable insights from data in Google Analytics, AdWords, and any other platform that requires data analysis. Using a popular information science framework, DIKW.
Although the origin of the framework isn’t exactly known, it’s attributed to T.S. Eliot’s 1934 play The Rock. Since then it was adopted by the information science community to help describe the relationship between Data, Information, Knowledge, and Wisdom.
Each of these levels of understanding provides an additional level of context that allows more and more valuable insights to be extrapolated from them. Although the actual model has been under scrutiny from data scientists it still serves as a good starting point to understanding how to navigate data.
Data is the fundamental building blocks that allow us to synthesize insights and information from. Data itself does not tell you anything rather it’s a signal of fact or symbol of something. Many times data in digital marketing is a metric (Users, Bounce rate, Session Duration, Conversions, etc).
For example 722 users is a metric and is data we know.
Information is organized data with meaning to it. This phase is understanding what the data is. Information itself has a bit more clarity to it but is not valuable by itself. Continuing with our previous example, knowing that the 722 users are from the Display channel answers the “what” but does not provide us any insight or actionable items.
This is typically where vanity metrics are reported. Campaign managers may simply look at the amount of Users and assume that 43,000~ users visiting their website is a fantastic number but how did we get those users? Is that actually a good number compared to previous date ranges? These are the questions that the next step of the model helps us understand.
The knowledge phase allows us to start pulling insights from the data. Knowledge in this framework is defined as answering the “how”. By synthesizing multiple pieces of information we can come to a conclusion and create frameworks/theories that allow us to run our data through them to test. It can be interpreted as understanding and being able to explain something.
To illustrate this in our example we can extract this information from the analytics view:
- The display channel makes up 722 users or 1.63% of traffic.
- The display channel has the highest bounce rate compared to other channels.
- The display channel has a 0% conversion rate which is low based on the average conversion rate.
These pieces of information start to illustrate the insight that the display channel may not be a good channel for us to invest our marketing budget into.
The final stage of the DIKW model is Wisdom. This is the most esoteric phase and focuses on answering the question of “why”. It uses knowledge and integrates it with experience and takes into consideration the other parts of the data ecosystem. This phase requires a holistic view of the data you interpreting and having the context to answer why is the way it is. To anchor this into a more concrete concept, we have the knowledge that the display channel isn’t performing too well for us relative to the other channels, but why is that happening?
We cannot answer this question by just looking just at the data, we need to identify the different variables of the display campaign, holistically, to see if there are any discrepancies that could answer why this campaign is performing relatively poorly. This could be anywhere from the ad copy of the display ad, the imagery, or the targeting settings. After we identify why we believe this is happening, we take action to modify the variable then wait for the data to come in to repeat the analysis.
Data analysis is an essential job function for most modern day jobs. Being able to read and understand data is especially important for anyone managing campaigns who use these insights to inform strategy decisions. Although this may seem tedious to go through every time you analyze a report, it is a trainable skill. After analyzing enough data and experience in your field, the different phases become second-nature.
Tired of flashy numbers and useless reports? Make sure your AdWords and SEO campaigns aren’t wasting you money. Take this step-by-step checklist to make sure you’re getting the most out of your marketing budget.