We will be following up this blog with a workshop webinar that the Magento Business Intelligence team will be airing on June 22, 2017 @ 1 PM ET.
One Sunday morning, Suzie, wakes up then proceeds to open the blinds to a shining sun. She smiles knowing it is going to be a beautiful day. After making a delicious cup of coffee she asks her Alexa what the weather is going to be like for the week. To her delight, it’s going to be 75 degrees and nothing but sun for the next six days. Feeling excited to spend time outside she decides that it’s an appropriate time to buy a new sun hat.
At that moment she will go in either one of two directions:
- Remain loyal to a brand she purchased from in the past (Company A) – approx. 30% of consumers
- Venture off into the endless world of new options (some Company B). – approx. 70% of consumers
As consumers these two options come up time and time again as we need to make purchase decisions. Mentally raise your hand if that rings a bell.
Now let’s switch gears. On the business side of things, this is a critical point along all of your customer’s journeys; especially if you are Company A. You and your team already dedicated time, energy and resources to acquire that customer in the past. Now, just a few months later, that same customer might consider running to a competitor? What an outrage! What did you do wrong?
Well, perhaps you actually did nothing to upset them. However, there’s a good chance that you could have been doing more to keep them happy. This includes using data to effectively encourage and make it easier for your customers (at scale) to be more loyal. Let’s talk about how.
Product Affinity Analysis
If you’re a Magento Business Intelligence customer, you most likely already have the data you need. It’s just a matter of collecting and organizing it in a way that uncovers the historic purchase patterns of your previous repeat customers and using that information to inspire similar buyers with curated marketing content.
One of the more effective methods of encouraging repeat purchases is through a product affinity analysis, also referred to as market basket analysis. At a high level, an affinity analysis is a data mining technique which identifies relationships between customers and the attributes associated with them. In the context of generating repeat purchases, an affinity analysis considers what a customer has already purchased and uses this information to determine what they are likely to purchase next.
For more technical details on how a product affinity analysis works, take a look at our previous blog post: To Affinity Analysis and Beyond. In this post, we’re going to focus on the benefits of an affinity analysis, starting with a specific example of how you can use this information to drive repeat purchases.
Predicting time between orders
After setting up your affinity analysis, you’ll have a list of product pairings along with the Support and Confidence indicating the likelihood that the given combination of products will be purchased together, whether in a single order or over that customer’s lifetime. For example, if you know that products A and B have a high support and confidence, you can prepare a list of email addresses for all customers who have purchased A but not B, then market B heavily to those customers.
However, it’s not enough to know what a customer is going to buy next; you also need to know when they are likely to buy it. Knowing the expected time between orders and how this varies with the customer’s order number is crucial for determining that pivotal moment when a customer is most likely to come back.
As an example, consider the following sample of customers:
|Customer id||Lifetime number of orders|
For simplicity, let’s say each of them have purchased product A in their lifetime, but never B. Let’s also assume that each customer placed their most recent order today.
You already know that customers who purchase A are very likely to purchase B, so you decide to create a new marketing campaign for these customers focused on product B. When is the ideal time to send this campaign to these customers?
You might be tempted to send the campaign all at once, but this could be a mistake for a few reasons.
- Send the campaign way too early and the customer might not be ready yet, prompting a quick delete of your email and lowering the confidence that your customer has in you to deliver relevant communications.
- Send it too late and one of your competitors might have already taken advantage of this opportunity with one of their own promotions. How dare them.
With a little extra information and planning, you can find the ideal time frame to market each of your customers. Consider the sample report below, which shows the median time between orders according to the customer’s order number:
From this report, you’ll see that the time between orders for a customer who has purchased just once is much longer than the time between orders for a customer who has purchased 9 times (6 months compared to 1 month)!
Armed with this bit of information, you’ll realize that the best time to retarget your customers is highly dependent on the number of orders they have placed up to that point. Revisiting our sample customers from before, you then decide to schedule your product B campaign for each customer as follows:
|Customer id||Lifetime number of orders||Time to next marketing campaign|
|Tom Hanks||2||3 months|
You could also decide that you’d like to attempt to drive down the average time between orders. This would mean running a test that sends the marketing campaign a few weeks prior to the respective historical average time between orders. Now you can now sit back and relax, with the confidence that you’ve greatly increased the likelihood of encouraging the continued loyalty of your customers.
Affinity analysis is one of the more effective ways of determining the relationship between products in your catalog. Pairing an affinity analysis with a measurement of the expected time between orders can turn your retargeting campaigns into a formidable, repeat-purchase-generating machine. If you’re interested in exploring affinity analysis further, contact our professional services team for more information and pricing. And if you’re not yet a Magento BI customer, sign up for a demo today!