A strategy for e-commerce analytics

  • Author: Ivan
  • On the topic: E-commerce

A strategy for e-commerce analytics

An incredible marketing idea that could bring an enormous return on one’s $20 ticket took place when the Powerball lottery reached $1 billion. That day, a company offered to share, comment, and like the post with ticket numbers in exchange for a share of the winnings. The post was spread rapidly and got 50,000 likes and as many shares.

Unfortunately, the return on that investment equaled zero. Here is the explanation why: The business that attempted to promote itself in this way was a tattoo studio in Fla, West Palm Beach. But people who shared or liked the post were from all over the world. Thus, the vast majority had absolutely no direct connection to that business. There is no doubt that people didn’t suddenly want to change their casual lifestyles just to travel to that place because of the viral post.

It proves that setting up goals to measure activity on websites is significant. The basis of a good analytics program is, first of all, a measurement strategy. The end goal is to sell a product. However, all metrics are quite diagnostic toward the end goal.

Measurement strategy illustrated by a pyramid can help us to understand it better. The model represents the relative number of metrics in order to reach the end goal of increasing product sales. This pyramid model helps us to understand and not drown in data.

Another important part of any strategy which is really helpful in business is that actions can be generated based on the given metrics. Obviously, only understanding the deeper metrics gives these results; changes at its higher levels surely push you to more powerful plans.

Metrics of sales

Metrics of sales

Sales are a simple metric if they are done well. Metrics help to do one important thing – to monitor the income generated per sale. This can be a key to ecommerce success. Growing basket size is a good idea if there are lots of sales, but at the same time, income per purchase is low. When it works, you can think about reducing sale friction, for example, discounts and card recovery.

It is noticeable to remember how returns affect sales numbers. Often if the number of returns is large, presumably the problem can appear to be deeper than just a product problem. For instance, wrong users or on page copy setting up unrealistic expectations or marketing.

Soft conversions

There are some soft conversions which help you keep a customer, such as newsletter signups and account creation. Since most clients are not inclined to purchase on the first visit, soft conversions can make a miracle and one day will become a purchase, even though it can take a longer time.

A good idea is to create soft converters as a new marketing channel and track them. Afterwards, you can observe the sales behavior of soft converters that is different from others. This information can increase the number of purchases, and accordingly, result in higher cart values per sale.
In any case, soft conversions cannot become the end goal of marketing; they are only entries to new possibilities.

The example with a tattoo parlor represents likes as a soft conversion. However, a user from California who likes a business page which is physically situated in Florida will bring no income at all.

Google Analytics use leading indicators as default metrics. Online sales are generated by traffic even though they are not equivalent at all. It often appears to be second or third behind the end goal of more sales. That is why more and more e-commerce companies are becoming concerned about it. Presumably, it would be better if low-quality traffic were replaced by high-quality traffic.

Leading indicators

Some leading indicators have to be segmented into more definitive metrics. For example, for a bricks-and-clicks business it is best to create an advanced segment in a local area. Thus, the main strategy is to view the profiles of users that are likely to convert. That can become a basis for classifying users as low- or high-quality.

Diagnostic metrics

If looked at in context, diagnostic metrics can become those details needed in website analytics. Diagnostics include quality scores and bounce. More in depth, high bounce rate might be on the retail locator page only if users are able to find the physical location of a store. On the contrary, even if users cannot find the products they are looking for, a low bounce rate on a category page can still be quite high.

If you are confused about a metric and the question is if it is difficult to explain metrics correlation to sales. But the thing is, the more questions there are, the more obvious a weak correlation becomes, even if it is quite significant.

For instance, let’s take a look at a search engine quality score. How could it possibly be related to sales? Somehow, it entails a higher cost per click in search engine ads and that both makes traffic more expensive and requires a high conversion to obtain an acceptable return. Therefore, a leading indicator appears to be not quality score, but conversion rate. Definitely, it’s best to concentrate on the latter.

578 0
Leave your comment to article
Share Share Share Share
Free Web Counters