Show me your website. Show me your app. What do you want the user to do? How is your product being used? I believe that successful digital analytics is impossible without a solid understanding of the strategic business objectives, the website goals, and the target-audience. It is also crucial to run requirements gathering sessions with all stakeholder involved to understand business requirements, reporting needs, as well as the data architecture and pipeline.
Let's say you are a web-publisher. Your revenue model is purely based on selling your ad inventory (regardless of whether that is smart or not...). This business objective is directly tied to your website goal: Engage your users & make them consume content, so you can create & sell many high-quality ad impressions. Given your business model, you are certainly interested in how many of your visitors are running an adblocker when browsing your site.
Adblock-usage might not be a built-in feature of your analytics tool. However, if this piece of information has been found to be crucial during the needs analysis, there are always ways to adjust the tagging and collect the data necessary.
Analytics-tagging constantly has to be debugged and adjusted during and after implementation. This involves, hit inspection, tag crawling, and constantly reviewing the user-identification mechanisms, such as the cookie-configuration. Handy helpers include in-browser development tools, source-code access, the console, and debugging proxy applications like Fiddler & Charles.
Does the code work properly? Is the metadata correctly attached to the tracking-pixel? Is the hit being sent to the analytics server for processing?
Tag inspection is vital for everyone involved in analytics and can easily be done with in-browser DevTools:
Although processing is something that is usually done on the side of your analytics vendor, consulting is important here as well. How should the data be processed? Does data collection, processing, and storage comply with your local data privacy regulations? Sampling / no sampling? Server and database structures? All these issues can play a crucial role in your analytics set-up.
On top of that, a lot of nifty things can be done via pre-processing. Correcting data, manipulating data or mirroring/excluding particular traffic can enhance your data for the analysis later on. Finding creative ways to optimize and clean up your data collection before it is being processed leads to a consistent data output.
How do you mine your data? How do you query large quantities of data in order to answer your specific question?
There are endless possibilities of organizing & exploring the collected data across all the available dimensions, segments, and metrics.
Now that the adblock-data is collected, you need to explore it. Ask questions. Does the usage change over time/across browsers/devices/localities? Are certain user groups more drawn towards using adblockers than others? Does adblock usage affect consumption-, engagement-, or conversion-metrics? The right usage of dimensions & metrics as well as the application of segments is crucial when it comes to data exploration.
Some people work with raw data, others need top-level management dashboards. Data Democratization means making the data accessible in the right granularity & the right format for anyone in an organization.
Business users, data scientists, content-editors, data analysts, top-level management - they all want data! But they all have different needs, skills, and expectations. That is what makes a good visualization- and distribution-strategy so crucial. It requires hands-on skills with reporting- & analysis tools as well as an eye for design and an understanding of the business-context.
Using either built-in tools (vendor specific dashboarding-applications) or working with APIs/connectors in order to dynamically connect to external data analysis- or visualization tools (such as Power BI, Tableau, RStudio, Python, Google Data Studio, or MS Excel) enables me to deliver the right data in the right format to the right stakeholder. Sometimes this even involves writing custom JS scripts or web applications to cater a specific need.
The last step of the cycle is all about using the insights and turning them into actions that drive business value. Most of the time, a bulletproof recommendation requires a more in-depth look at the data. Applying statistical analyses or machine learning algorithms to web analytics data (using programming languages like R or Python) is often part of the value creation process.
Analytics data can then be the basis for a wide variety of actions that can drive your strategic business objectives. Possible application areas include content-strategy, product design, UI/UX-optimization, personalization or targeted advertising-strategies.
In our little exmaple, there are plenty of possible implications. Should you integrate some sort of adblock-notice or warning? Or block the content for adblock-users altogether? Should you target certain user-groups? Or certain sections of the site? Leverage your analytics data to turn the insights into action. Note that you will want to monitor the numbers over time and run some further statistical analysis in order to investigate whether there is real potential business value to be expected from these implications.