RFM analysis increases eCommerce sales.  Today, we explore how we can use recency, frequency, and monetary data to unlock personalized marketing, and maximize cohort analysis to increase retention.

To discover all the powerful RFM segmentations you can create through example, click here. Otherwise, we start with the fundamentals: what is RFM analysis, and how can we use it?

RFM analysis is a data driven customer behavior segmentation technique.

RFM stands for recency, frequency, and monetary value.

The  idea is to segment customers based on when their last purchase was, how often they’ve purchased in the past, and how much they’ve spent overall. All three of these measures have proven to be effective predictors of a customer's willingness to engage in marketing messages and offers.

While RFM analysis was born in direct-mail, it is a powerful tool for eCommerce stores to use today. 

Above is a great illustration of how customer segmentation empowers businesses to speak to customers specific values. Image source: Intercom 

The first known applications of RFM analysis was in the catalogue industry. Pioneers include Land's End, JC Penny's and others. Since it's inception, many varients of RFM have come been developed, including

Conducting an RFM analysis on your customer base and sending personalized campaigns to high value targets has massive benefits for your eCommerce store.

How should you calculate recency? Or score frequency? What is a good monetization threshold? 

Defining thresholds is the first step in segmentation. Below we discuss common metrics eCommerce companies can use for recency, frequency, and monetization.

Recency measures the time since last purchase.

There are two challenges for eCommerce stores when calculating recency.

First, in an omnichannel world, it can be difficult to tie together purchase data from each channel.

Second, each business will have different interpretations of what a good recency score is. For example, consumables have an inherent need for frequent orders, making the required time since last purchase shorter to qualify for a higher score.

Above, Starbucks' products are typically consumed within a day.  Their product mix requires a different interpterion of recency data compared to slower moving products with much higher product lifecycles. 

The same concerns in recency present themselves in frequency analysis as well. 

Again, product lifecycle 

Calculating recency, frequency, and monetization scores present similar challenges. The reality is, each business is unique. It is incredibly difficult to create accurate benchmarks. 

Luckily, there are a few common approaches to properly assign RFM metrics.

Perhaps the easiest way to create RFM scores, quartile analysis allows you to quickly and fairly assign scores based on relative performance.

Each quartile will give a score, 1 through 4. The final RFM segmentation will use these scores together.

Here is a great, simple overview for using quartiles to define your RFM segments using Python. 

Image Credit

RFM segmentation doesn't have to be complicated.

Below we will show how you can create a RFM model in Excel. We go step by step, and include screenshots so you can easily replicate the model. 

Keep in mind that your eCommerce brand is unique.

This isn't going to be the best way for your particular business, but it will be a great template for you to iterate on.

Let's begin.

Before you can begin, you need to define the one KPI that matters most to your business for each segmentation vector: recency, frequency, and monetization.

To do this, you will have to connect your purchase history to each customer, and select a timeframe that you want to work with.

For our example, we are going to use the following KPI's and timeframe.

Recency: Date of Last of Purchase
Frequency: Total Number of Orders
Monetization: Average Order Value
Timeframe: 2 years 

As you will see, RFM analysis is a straight forward process. The goal is to systematically score each customer according to recency, frequency, and monetization. We do that by first rank ordering all customers across our chosen metric, and then scoring them based on how well they perform relative to the other customers in your database.

There are a number of KPI's you can use for Recency. Example KPIs include 

For this walk through, we are going to use days from last purchase as the primary metric. 

You should intuitively know which metric makes the most sense for your business. Chances are if you are a traditional eCommerce store, date of last purchase will be your metric of choice also.

First, we want to download your client information with your specified KPI's. Here, we simply downloaded this information directly from Barilliance, and uploaded it into a Google Sheet.

Next, we want to clean up the sheet a bit. 

I delete some unwanted columns I forgot to close out: Sessions, Last Seen, First Seen, First Ordered, and AOV.

Next, you are going to add three columns for your RFM scores. Title them across the top "Recency", "Frequency", and "Monetization". 

Finally, I apply a filter on the data to make it very easy to sort across. If you don't know how to apply a filter the process is easy. Select all of your data including your headings (you can do this quickly by holding down shift+command+arrows). 

Once all of your data is selected, click data->filter. 

After you are done, your sheet should look something like this. 

Navigate over to your Recency KPI, and sort the list in the appropriate order. In our case, we go to "Days from last purchase" and sort in descending.

Finally, score each customer according to their position. 

While there a few methodologies for doing this, I've found the easiest and most helpful to be with quartiles. 

Take your total number of customers divided by four. Next, give each quartile a score reflective of the position. 

First Quartile: 1

Second Quartile: 2

Third Quartile: 3

Fourth Quartile: 4

Don't worry too much about getting everything "perfect".  For example, here we have two customers who have both ordered in the last three days. I simply assigned them both a recency score of 1, even though that puts 5 customers with a score of 1 and three with the score of 2. 

At this point, your sheet should look similar to the one below.

The process is very similar for both Frequency and Monetization. 

While we are using the total number of orders over the past two years as our Frequency KPI, there are a number of competing metrics you can select. Some of them include:

As mentioned before, we are going to be using # of orders as our Frequency KPI. Repeat steps 2.b and 2.c, expect use your Frequency KPI as your guiding metric. After scoring, your sheet should look similar to the one below.

Finally, you are ready to knock out the Monetization scores. 

Monetization KPI's include

As before, repeat steps 2.b and 2.c, expect use your Monetization KPI as your guiding metric. After scoring, your sheet should look similar to the one below. 

Awesome! 

At this point you have the analysis done. Next comes the fun part - actually using these newfound insights to identify profitable segments. 

There are many ways marketers have used this segmentation to help guide their marketing. Here are a few ideas to cherry pick from.

RFM Score: 111

Who They Are: Highly engaged customers who have bought the most recent, the most often, and generated the most revenue.

Marketing Strategies: Focus on loyalty programs and new product introductions. These customers have proven to have a higher willingness to pay, so don't use discount pricing to generate incremental sales. Instead, focus on value added offers through product recommendations based on previous purchases.

Above, Uber targets their "core" RFM customer segment, introducing their new Uber Eats offering.

RFM Score: X1X

Who They Are: Customers who buy the most often from your store.

Marketing Strategies: Loyalty programs are effective for these repeat visitors. Advocacy programs and reviews are also common X1X strategies. Lastly, consider rewarding these customers with Free Shipping or other like benefits.

Here, Costco compliments their membership business model with a custom credit card to further increase repeat purchase rates and expand wallet share. 

RFM Score: XX1

Who They Are: Customers who have generated the most revenue for your store.

Marketing Strategies: These customers have demonstrated a high willingness to pay. Consider premium offers, subscription tiers, luxury products, or value add cross/up-sells to increase AOV. Don't waste margin on discounts. 

RFM Score: X13, X14

Who They Are: Customers who return often, but do not spend a lot.

Marketing Strategies: You've already succeeded in creating loyalty. Focus on increasing monetization through product recommendations based on past purchases and incentives tied to spending thresholds (pegged to your store AOV). 

Above is an example of Target using lifecycle marketing on specific RFM segments. Notice how they pair financial discounts with spending thresholds to drive repeat purchase and increase customer profitability.  

Another example comes from AirBnb. Here, they send triggered messages based on customer's viewing activity to prompt a booking.

RFM Score: 14X

Who They Are: First time buyers on your site.

Marketing Strategies: Most customers never graduate to loyal. Having clear strategies in place for first time buyers such as triggered welcome emails will pay dividends.

Starbucks is excellent at moving customers across RFM segments. Above, they use email to bring customers into their loyalty rewards program. You can see our full case study on Starbucks here. 

RFM Score: 44X

Who They Are: Great past customers who haven't bought in awhile.

Marketing Strategies: Customers leave for a variety of reasons. Depending on your situation price deals, new product launches, or other retention strategies.

Barilliance empowers you to not only create personalized experiences for each RFM segment. It also gives you the ability to create multivariate ab tests. 

You can quickly see which offers resonate most with a given segment, discover what content leads to sales, and more.  

One of the significant challenges with RFM Analysis (and segmentation in general) is creating a full, 360 degree view of a customer.  Today's omnichannel shopper isn't much different than the direct mail world RFM was born from. RFM is still an incredibly powerful model.

But it's effectiveness is determined by the quality of the data you have.

This is where Barilliance shines. It creates a holistic customer view, across devices, shopping sessions, and channels. You can see which pages customers interact with, time since last purchase, order volume, brand engagement and more.

In other words, you have the ability to create highly detailed segments, including all the ones we talked about above, through one simple interface. 

Take action.

RFM is an established, clear way to get more out of your client list. 

But...

If you want to take RFM to the next level you need to consider Barilliance's Retention. It improves traditional RFM analysis in two fundamental ways.

First, it connects your online and offline purchases, web behavior, and demographic data in one place - allowing you to make more accurate segmentations. 

Second, it lets you automate campaigns to defined segments. It connects to Web Personalization and individual email product recommendations to customize each interaction not at the segment level, but at the individual level.