Bringing you Tips on Running....

A|B Testing | Optimization|Experimentation

  • Experimentation Expert

D3 Series | Isolate COVID-19 Data in Post Analyses


Your Data is the foundation and bone structure of your customer. You can wrap functions and layers of logic around it like muscles but it won't change what your Frankenstein looks like. We'll call this step your vitamin D3. You need this data to tell you what to do next. So make sure it is clean and untampered.

Once your test wraps up, hopefully in the near future for all our sake, dive deep into these insights to identify changes in your mean and understand the new type of customer surfing your channels.


Take observations from the small businesses that thrive on the "X Business to your doorstep": Lyft, Uber, Hello Fresh, Blue Apron, Door Dash, Postmates, Amazon Prime delivery, Netflix, Disney +, Hulu etc.

Ensure that you keep these frightened customers at bay and meet the many demands. What can you learn from the small business structures that are prevailing and failing?

The public is in search for answers on how to hold onto a piece of normalcy when it comes to necessities. At a time like this, people actually need to spend more on business, at least to their best ability, to soften the blow of our ultimate comeback. Buy your items online, pick a few neighborhood supply runners (kids) to reduce the foot-traffic in stores. And buy that Jimmy Johns sandwich because somehow they are still freaky fast.


We are still running tests. So how can we isolate the variables of customer response to COVID-19 in our metrics?

  • Include session replays if applicable

  • Add supporting metrics for your primary KPI to identify the subtle changes in behaviors according to the purpose of the page

  • Separate your customer types by groups of identifiable attributes. Keep it clean and straight forward, you can always narrow it down into subcategories

  • Explore isolating geographical conversions and non conversions by: regions, hospital counts, employed vs unemployed rates by state, etc.

  • Keep a list of questions from the business to follow up on once you have unpacked this data

  • Identify whether your business has a positive or negative correlation to health crises by product or category. Next time, you will have an action plan to deliver to the customers in a timely fashion

  • Gather most searched terms during your tests and outlying navigation patterns from your typical observations YoY, MoM etc.

  • Keep in mind you most likely WILL NOT reach confidence using <50% of your site so these statistics will be part of a library of possible trends to explore later

Want customized web analytics dashboards and help on relevant data? Subscribe now to get access to exclusive documents.


Signed,

Experimentation Expert

2 views0 comments