Updated: Sep 22
I want to talk about digital analytics data quality that confuses lots of people who are new to this world.
Analysts who are used to the structured data which gives them perfect data will have frustration at the beginning if they don’t understand digital analytics dynamics. In structured data world, you are based on transaction data which is the exact number since data is created by application during transaction , there is no third party involved. This means when transaction happens, your data is also being created. If you have transaction, you have data.
In this new digital world where we track user clicks, data is not passed from application but it is passed from web browser. Websites are not connected to the visitors directly, there is a medium between them: Web Browsers like Firefox or Safari. Visitors are accessing to company’s digital assets through this medium. So if you want to collect data from customer who is visiting your site, you need to ask to the browser to send data to you. You usually need to use another third party analytics vendor to get this data.
This is the essential thing to know when you look at this data.
Customer clicks something on the page, browser sends this click as a web request to the website and site gives response from its servers, browser picks up this response and shows it to the visitor. During this traffic you will be collecting data through your web analytics vendor. As you can see there are many things that can go wrong during this transaction that consists of many steps, parties and hardwares. Even if you have the perfect web analytics implementation, you really can’t control things like visitor’s computer, browser performance, or visitor’s internet connection. You even can’t control your site 100%. Servers can fail, you end up with bugs and pages might have errors.
This is what you got and you need to look at data keeping this in mind. There is no such thing like perfect digital data and you just need to use your business and analytical skills to make sense out of this data. That’s why good digital analytics professionals look at trends to create insights rather than looking at raw numbers.
This is also one of the reasons why you should not let everybody to make analysis. When an analyst does not understand the product and infrastructure of digital assets well, they are going to end up with lots of wrong insights. At the end they are going to blame web analytics vendor or technical implementation. Implementation might be wrong but you need to have people who can understand data quality in a blink of eye
Second type of people who are not comfortable with differences in data are those who are not employing analytical skills that is needed. When you are highly analytical you can easily make connections between different reports, KPIs and business objectives. You would recognize wrong data but you will also recognize that in some cases this wrong data will not deviate your insight, you will still get something out of that one.
Hire people who can create insights with data that is available to them. That’s the skill we need in this fast-paced digital world.