How have D2C e‑commerce businesses used data to optimise throughout the pandemic

The perennial question in commerce, both physical and digital. Should you focus more money, time and effort into acquiring new customers or building stronger relationships with those you already have?

Whilst it remains true that there’s no right or wrong answer, since the beginning of the pandemic the goalposts have shifted for D2C e-commerce businesses. Today, customer acquisition is more difficult than ever. So should customer retention take priority?

For the last 18+ months there have been more people shopping online than ever before. Subsequently, there have been more businesses operating online than ever before, resulting in a digital marketplace that’s crowded and competitive to levels we’ve never seen. For a more detailed explanation on all of this, check our article on the state of the D2C e-commerce landscape today.

Customers shopping online have more choice. There are more stores and more offers. They’re bombarded with more ads, making yours not only less effective but more expensive to run as the space becomes increasingly contested.

As a result of the pandemic, customers are also more price sensitive. With huge numbers either losing income or losing their jobs entirely, it makes sense that consumers today are looking for the best deals. And in that race to the bottom there’s simply no room to compete with the big players like Amazon who can afford to lower prices beyond anything that more modest e-commerce businesses can afford.

For first time buyers, price and shipping convenience are more often than not the single most important driving factors when making a purchase. In the current climate it’s more likely than ever that they’ll go somewhere where they can get the best deal. We all know how well Bezos did out of the pandemic, this is a big reason why.

So what does all this mean? Quite simply: attracting customers into your store is more difficult than it’s ever been. Customer acquisition costs had already been steadily rising for the five years leading up to 2018, and the worldwide rise in e-commerce driven by the coronavirus pandemic has only accelerated that trend.

If you choose to prioritise customer retention all of the hard work described above has already been done. The customer has successfully navigated the marketplace to find your store and make a purchase. They see something in your brand and/or product that they like and now the job is to encourage them to return, to make them feel valued and to give them the best possible user experience.

Existing customers are easier to reach (you likely already have contact details of theirs stored somewhere) and are more likely to buy again so long as they have a positive experience in your store and with your brand.

But which customers specifically do you want to come back? Preferably all of them, of course, but let’s be realistic here. You’ll only retain a portion of your customers, it’s up to you to identify which portion.

Are all of your customers of equal worth to your business? Some will buy big once in a blue moon, others will buy little but frequently and others will be somewhere in between the two. All of these are bringing in revenue but the trick is to work out which group brings in the most. Once you know this, you know which group to focus your efforts on when optimising your store for customer retention.

So how do you identify which customers fall into which category, and which category is the most profitable to your business?

Data science has entered the chat.

Data science can separate your customers into segments based on specific characteristics. With your customers uniformly split into different groups you can identify which group is most valuable and then identify how to keep the members of that group coming back time and again.

So how does it work?

By tracking events throughout your entire website you can build a comprehensive record of your customers’ journeys. Conversions, cart abandonment, clicks, drop-off points, average number of sessions before purchase, all of this and more can be tracked through your website.

To the human eye this results in a jumbled, unstructured spreadsheet of numbers. But by building an AI algorithm, data scientists can extract clear patterns from the chaos in a matter of seconds.

Let’s take a super simple example:

• Say 100% of your customers each buy 4 products. There’s no distinction to be made on the number of products purchased. So any AI here would not use the variable "amount of products purchased".
• However, if 50% of those customers purchased during the weekend and the other 50% during the workweek, there’s a clear distinction here. So the AI will use "day of purchase" to create 2 distinct groups.

The real world scenarios are far more complicated than this, but hopefully you get the idea of how it works. The AI looks for customer patterns and characteristics that can separate them into uniform groups.

With the groups established it’s then possible to identify which is most valuable to your business. Without getting too technical here, data scientists will enrich customer segments with both internal and external data sets. From here they can calculate how much money is lost each time a customer from each group abandons a purchase and subsequently identify which group has the biggest growth leverage.

The customers with the biggest growth leverage are the ones you want to retain as often as possible. This is what the more successful businesses have been doing throughout the coronavirus pandemic, as customer acquisition costs have gone through the roof.

It’s far more cost efficient to introduce data science into your e-commerce strategy, identify which existing customers are bringing in the most money and then optimise to retain those customers, than do battle with the big businesses to continually acquire fresh customers.

Once you’ve identified your most valuable segment of customers, how do you optimise your store to improve their UX?

Luckily, if you’re using data science you have all the answers you need right in front of you. Every action a customer makes in your store represents a preference for something. Whether they make a purchase, leave your store without browsing or abandon their cart, all of the actions they make leading up to that point are a trail of preferences that you can read and act upon.

Which banners they click, how many drop-down menus they use, if they read blog posts, FAQs, or the returns policy - all of this tells you what encourages a customer to buy and what puts them off. Quite simply, with some technical data analysis you can identify all of the features that your most valuable customers expect in a store.

Then it’s up to you to make sure those features are in place. By building a store that matches your customers’ preferences as closely as possible, you’re far more likely to see them return again and again.