How to Create a Business Model of a Loyalty Program

How to create a loyalty program model that would help you increase sales? This case study demonstrates how customer analysis, proper data segmentation, and an effective loyalty program model can help a brand increase revenue and foster customer loyalty.

When you see the real numbers, it is easier to get an objective view of a business and see the dynamics in marketing and sales. Thus, a financial model of a loyalty program can help you achieve the business goals. We calculated this model using an Excel table and now want to share the results.

Get a Free Template for Building a Financial Model of a Loyalty Program

Our analysts use Excel tables to create loyalty program models. You can download a free financial model template in Google Sheets below and build a model on your own.

How to Choose a Business Model of a Loyalty Program

One can use different loyalty program models depending on the type of business one’s dealing with.

8 Must-Know Loyalty Program Models 

1. Point-based loyalty programs: a customer gets points for each purchase. Later, customers can redeem the reward points.

2. Cashback: this loyalty program model is similar to the previous one, but instead of loyalty points, customers receive cashback for their purchases.

3. Loyalty cards: in this loyalty model, after the first purchase, a client gets a special card. Each next purchase brings new stickers or stamps for the card. When the slots on the card are full, the customer gets a reward. 

4. Multilevel loyalty program model: when a client spends a certain amount of money, they achieve a new loyalty level. The higher the level is, the bigger the discount becomes. 

5. Rewards from partners: a customer gets discounts or points that can be used in partner stores. 

6. Paid points: a client pays for the loyalty program membership. In return, they get free/fast delivery, additional discounts, etc. 

7. Hybrid loyalty models: for example, a points-based program + multilevel program.

8. Non-monetary rewards: loyal clients get non-monetary rewards (for example, a meeting with famous persons). 

Each loyalty program model can be profitable, but if you want to find out which one is perfect for your business, you need to analyze several factors:

• type of your products;
• buyer behavior of your customers;
• purchase frequency;
• average purchase value, and so on.

In the case study below, we’ll share our financial model of a multilevel loyalty program. For this loyalty program model, we needed to calculate:

• what amount spent would make the discount grow;
• what discount levels there would be;
• the spending level to enter the loyalty program;
• the supposed profit brought with the help of this loyalty program.

Our first step was to gather all of the data on the transactions and customers' interactions with the brand.

Financial Analysis for a Loyalty Program Model

The purpose of a loyalty program is to motivate customers to buy more: e.g., without a loyalty program they would spend $50, but thanks to the loyalty program, they might spend $65. However, sometimes it’s not really clear why customers spend more: whether the reason is the loyalty program or the customers are just loyal to the brand.

That’s why our financial analysts should create two forecasts:

1. Sales dynamics without a loyalty program;
2. Sales dynamics with a loyalty program.

This kind of analysis requires retrospective data: we should analyze how the sales are going right now, when the loyalty program has not been launched yet.

What You Must Know Before Launching a Loyalty Program

• Offline and online sales can be different: you’ll have to create your financial model of a loyalty program considering different dynamics of offline and online sales.

• You’ll need to analyze your customers: the purpose of a loyalty program is to motivate customers, so you can’t focus on the sales volume only. You have to collect the data about your clients: how often they buy, how much they spend, etc. Most likely, you shouldn’t worry about getting this info about your online clients, but there can be issues with offline ones.

• You may need to cleanse the data and delete things like duplicate purchases, cases of bulk buying, etc.

Study Customer Behavior 

Customer analysis plays a crucial role in building a loyalty program model: it's important to understand how loyal your customers are at the initial stage.

When we collected and processed the data, we got the following useful info:

• total number of clients;
• number of loyal clients;
• purchase values.

We also created a graph to take a closer look at the clients and see how often they bought something.

The next step was to analyze the loyal customers: we analyzed the number of purchases and the time period between them.

Another method of dividing customers and analyzing their behavior & loyalty is RFM analysis. We made an article on how to conduct this analysis.

Create a Sales Forecast Without a Loyalty Program

This forecast answered the following question: “What will be the sales dynamics if we leave the situation as is?”. In our case, offline sales were growing and online sales were dropping. We needed to analyze the current situation considering the following factors:

• monthly number of clients;
• sales trends;
• conversion rate.

Using data segmentation, we forecasted what sales we would get without the loyalty program.

You can’t create a financial model of a loyalty program being 100% sure that your sales will grow; instead, look at the real numbers and don’t forget about previous trends and seasonal activities. 

Design Business Models of a Loyalty Program That Would Increase Sales

As we’ve mentioned before, we picked a multilevel loyalty program model: the discount would grow depending on the purchase value. We created four variants of this loyalty program model. 

However, you can’t just assume that sales will grow and your stores will be crowded. That’s why we suggested three forecasts for each loyalty program model: pessimistic, conservative, and optimistic. 

Thus, we created four loyalty program models in three versions (optimistic, conservative, and pessimistic) for offline and online.

Still, we needed to consider the number of new clients to come. We didn't know the exact growth, but we could create a forecast. To make our predictions more accurate, we suggested three of them: pessimistic, conservative, and optimistic.

However, the growth might depend on the discounts too: a bigger discount could attract more clients. Thus, different business models suggested different growth:

We’ll showcase the examples for the online loyalty program model only. 

Forecast the Number of Clients in the Loyalty Program

At this point, we have the data on clients and purchases and a loyalty program model. The next step is to bring together the data and the loyalty program. 

In our case, we gathered the existing data on the clients and their purchase values and divided them by their status in the loyalty program. Thus, we understood how many people would get this or that loyalty status.

Indicate Customers Who Can Be Easily Motivated to Get the Next Loyalty Level

Some customers were close to getting a discount. For example, if the loyalty status began with $150 spent, customers who’d spent $130 were easy to convert into a new loyalty level. 

This way, we segmented the customers according to their loyalty status. The more the client bought, the more loyal they were considered.

Forecast the Sales Uplift From the Loyalty Program

The final part is to calculate the sales uplift brought by the loyalty program. At this step, you need to consider the following points:

• don’t forget to take into consideration your expenses on the loyalty program and match them against the revenue. 
• instead of thinking about the total purchase value, focus on the additional sum that a customer spent to enter the next loyalty level. Of course, this additional sum will be lesser than the usual purchase value. In our case study, the average cross-sale was considered 80% from the average purchase value.

How We Chose the Most Profitable Loyalty Program Model

To choose the model that will increase your sales, you need to create monthly forecasts for the loyalty program models using the factors we’ve mentioned in the previous section. Each case is unique, but you may use our example and see the way we organized the data in the Excel table below. 

How We Calculated the Loyalty Program Models

First, let’s sum up how we calculated our loyalty program model. We created:

• four loyalty program models;
• three forecasts of customer growth (pessimistic, conservative, optimistic);
• calculations for both offline and online sales;
• 12-month forecast for every loyalty program model;
• calculations for new and loyal customers separately;
• a forecast for the clients who spent less than $50. 

As a result, we used 312 Excel tables for all the calculations.

The next step was to turn the monthly data into a yearly forecast. Thus, we put together the figures and created a forecast for a year in three versions: pessimistic, conservative, and optimistic. After that, we compared this forecast with the previous forecast without the loyalty program.

In this table, we put together online and offline sales.

Select the Most Profitable Loyalty Program Model

The final step is to choose the most efficient loyalty program model: the one that gives you the bigger uplift with lesser expenses on rewards. 

In our case, we got the following winner: a loyalty program model that offered 0%, 10%, 15%, and 20% discounts for $100, $200, 370, and > $370 purchases. 

Final words 

In this case study, we described only one way you can calculate a certain loyalty program model. Although you may face other issues with your unique loyalty program, the principles are pretty much universal:

• the calculations must be 100% accurate;
• you need to create a loyalty program model using the real data and considering the previous sales trends;
• offline and online sales are different, so you need to calculate loyalty program models separately;
• you need to consider reward expenses for motivating customers to enter a new loyalty status. Also, don’t forget about the delivery costs.
• the financial model is only the beginning of your work: every forecast is just a hypothesis. You need to implement the loyalty program and check whether it’s working. Perhaps, something will change and you’ll have to modify your financial model. 

The Final Results: Our Business Model of a Loyalty Program