How To Avoid Failing at E-commerce Analytics?

Luna ShirleyGeneral

How To Avoid Failing at E-commerce Analytics?

If you have your data management in a top condition, nothing is stopping you from running your automated marketing campaigns to boost sales on your online store.

Yet, you can’t set up these campaigns without insights from your database. What could hold you back during such analyses and how could you solve these challenges?

Basing Decisions on Dirty Data

Severity: High

Make sure that you are working with clean datasets, otherwise the outcome of your analyses could be polluted by nonsensical data. If you are running an online store with low traffic volumes, even failing to exclude your internal traffic may skew your findings.

What to do about it?

Apply common sense and add filters to your analyses to exclude values which may pollute your dataset (such as test purchases done by your colleagues) and before concluding any outcome out of your analyses, review a sample of the data if it doesn’t contain some faulty values which you’ve initially missed.

The old saying, “if it looks too good to be true” frequently works twice in e-commerce analytics.

Lack of Transparency

Severity: Medium

One of the most commonly occurring issues in e-commerce analytics is a plain misuse of metrics.

With marketers dealing with a high number of metrics which often tends to slightly differ in the way they are calculated, it’s an easy mistake to pick the wrong one while analyzing your data. Such a mistake can lead to a tiny deviance without any significant impact or could result in making a completely incorrect business decision.

What to do about it?

The most foolproof way is to make sure that you add descriptions to each metric, event, customer attribute and to any other data type that you work with in your database.

It may sound like a no-brainer, although you need to build a habit of it, which could be difficult to uphold if you tend to create new metrics on a daily basis.

Focusing on Superficial or Secondary Metrics

Severity: Low

Vanity or secondary metrics are often set as targets to optimize in e-commerce analytics even though they themselves may not be as impactful as other, more telling metrics and as such marketing teams aren’t using their time effectively.

As an example, when analyzing newsletter performance, marketers tend to focus on Open Rate, yet this metric by itself isn’t necessarily indicative of newsletter performance and its improvement may not result in increased revenue from newsletter traffic.

What to do about it?

Before deciding which metric or set of metrics you will be analyzing, make sure to ask yourself which of them you can optimize in such a way that improving it will have a direct impact on your company level KPIs.

Desynchronized Time Zones

Severity: Low to High

If you are analyzing the performance of an online store which sells to multiple countries and has customers all around the world, or at least in different time zones, one of your challenges that you are facing is to make sure that your time-stamped data is synchronized to either your head office or to local time zones.

Mixing time stamps could lead to improper business decisions, misfiring of your automated marketing campaigns and other serious issues.

What to do about it?

The core solution is to make sure that each time specific datapoint contains two timestamps, one local based on your customers’ local time zone and the other one synchronized to your head office.

AdBlock Skewing Front-End Tracking Results

Severity: Low to Medium

Adblocks could be very aggressive and skew the results of your front-end tracking. If your customers are tech-savvy and are often using adblocks, you may find out that your on-site banners and on-demand web layers aren’t being shown to your customers even though you are tracking that they were.

What to do about it?

The most difficult part is to identify that something like that is actually happening and testing your on-site campaigns with adblock turned on will help you identify potential issues and seeing a severe underperformance of your newest on-site campaign may be a sign of that as well.

To mitigate the probability, try not to name your on-site campaigns in such a way that they won’t contain typical keywords connected to ads such as advertisement or banner.

Unfortunately, in this case, it’s nearly impossible to give you a fool-proof method to make sure that your on-site campaigns won’t be blocked for sure.