5 Common Mistakes Online Retailers Make When Analysing Their Ecommerce Data

Posted by Hannah Stacey 23 May 14

Getting detailed data on how your customers interact with your ecommerce store has never been easier. But the trouble is that ecommerce data on its own is kind of boring; after all it’s just a bunch of 0s and 1s in the murky depths of a server. Using data to create powerful insights that you can put into action and see tangible results? Now we’re talking!

shutterstock_169523207

Getting a handle on what to actually do with your data, however, is something that online retailers can really struggle with. After all, we’re not all data analysts, and there are a lot of factors to consider when using data to make calculations that have the potential to make a considerable impact on bottom line (for better orworse).

Here are some of the most common mistakes online retailers make when analysing their data.

1) Attaching too much significance to overall AOVs

Average order value (AOV) - or total sales revenue divided by total number of sales - is an important KPI, providing insight into customer purchasing habits.

However, where many online retailers go wrong with their analysis of AOV is that they tend to look at average order value as a single figure across the whole business, and make important marketing decisions accordingly.

The main problem with doing this is that it fails to take into account that not every customer is the same; an online seller’s customer base is likely to be made up of big spenders, bargain hunters and a wealth of people in between, so it’s a risky business making any assumptions based on an average AOV.

In its simplest form, segmenting customers by AOV (from low to high) enables online retailers to target them with products they’re more likely to buy. An even more effective way of doing this is looking at average item price per category - this helps distinguish between those who purchase fewer expensive products and those who just buy lots of cheap products.

Likewise, segmenting customers by AOV then looking at factors like source, marketing campaign, device etc. can help online retailers build a better picture of the performance of each of these channels - which of them are driving high AOV customers? How can this information be used to optimise other channels?

2) Focusing on revenue instead of profit when identifying best products and customers

Many online retailers fall into the trap of forgetting to factor in profit when trying to identify their best performing products and most valuable customers. And who can blame them when a lot of analytical tools don’t have the in-built capability to assess margins?  

Factoring in gross profit can have a considerable impact on the products online retailers choose to promote and the customers they choose to promote them to. It can often be the case that what might seem like a great product in terms of the revenue it generates isn’t the best performing once you start looking at the profit it makes.

Likewise, when trying to find the channels that acquire the best customers, online retailers need to be analysing their customers in terms of lifetime profit they give, not revenue, as it’s quite possible that a customer might return a high revenue but only be buying very low-margin products (more on accurately calculating the lifetime value of customers here).

3) Not accounting for returns or cancelled orders

When attempting any sort of data analysis it is important to get the full picture, and online retailers often overlook product returns and cancelled orders. Returns shouldn’t be taken lightly, and factoring them in can have a marked effect on how profitable an item is.

For example, I might have a dress that has a high return rate because it comes up particularly small and lots of customers are sending it back. This dress might make great margins on paper, but if half of the items are returned and I’m having to bear the costs, I don’t want to treat it like a ‘hero’ product and focus marketing resource on promoting it.

Ecommerce Product pages  

4) Not tracking site search

What gets entered into the search bar on a website gives away several important (and potentially profitable) clues as to buyer behaviour and preference. Are certain site search terms and phrases leading people to purchase a particular item in your online store? Be sure to optimise that product page for these, as people are likely to be searching for the same terms in regular Google searches - ranking highly for them may earn you more customers.

Likewise if an online retailer’s search results pages have a high exit rate, they might want to think about improving the relevancy of their site search, as visitors may be turned off by the fact a search isn’t returning what they’re looking for.

5) Calculating ROI incorrectly

There are a number of potential stumbling blocks on the way to arriving at an accurate figure for return on investment. A major problem is caused by what we talked about in point two: failing to account for profit margins and just looking at revenue.

The impact of this can be significant. For example, just looking at revenue, I may be running an AdWords campaign that seems to have an ROI of 400 per cent. I’m pretty pleased with myself. But what I’ve forgotten is to factor in profit margins - if, for example, this is 40 per cent of my sale price, it has large implications for my ROI calculation.

Likewise, calculating ROI based on a customer’s first transaction size is equally problematic, as it fails to give a bigger picture of their behaviour over time. For example, I may completely write off Facebook advertising as a means of acquiring customers because, based on first purchase data, the ROI is negative. But what if a customer acquired through Facebook is more likely to come back and make repeat purchases than one acquired through a different channel? It could feasibly emerge that, over time, Facebook delivers a far better ROI than other channels when I take into account lifetime ROI.

In order to make a more accurate calculation, customer lifetime value should be used in place of first purchase revenue in the ROI calculation:

ROI = (CLV/customer acquisition spend) - 1

Ultimately, accurate insight only comes from accurate data analysis and it’s important that online retailers do not overlook any of the issues above when using data to make key decisions.

Artboard 10 B-1.png