How can a “recommendation engine” that enables proposals tailored to customers be used in business?
In this article, we will explain how to use the recommendation engine in business, use examples other than EC sites, and the advantages and points of introduction.
An overview of recommendation engines and how they are used in business
Even if you have a vague image of a recommendation engine, there may be many who are not clear about how it can be used specifically.
In order to make it easier to imagine how to use and utilize the recommendation engine in business, we will first explain the basic knowledge of the recommendation engine.
What is a “recommendation engine” that enables suggestions tailored to users?
A recommendation engine is, as the name suggests, a device for making recommendations. Refers to functions and tools that recommend products and content that are likely to be of high interest to users who have visited an EC site or company’s own site.
For details on the mechanism of the recommendation engine, please refer to “What is the mechanism of the recommendation engine and how to make it? Explanation of the 7 main types of AI (machine learning) algorithms”.
Business Use Cases for Recommendation Engines
Recommendation engines became famous after being introduced by Amazon. Nowadays, it is installed in many online shopping malls such as Rakuten Ichiba and ZOZOTOWN.
In fact, when you open the smartphone case page on Amazon, you can also suggest the purchase of a protective film as a product that is purchased together, or display related products, giving users many attractive choices.
These suggestions make it easy to find similar products.
It is possible to discover products that were not originally planned to be purchased and encourage simultaneous purchases.
In recent years, in addition to EC sites, recommendation engines are also being used on content sites and news sites.
By displaying content that is related to the content that the user has viewed and content that is considered to be of interest, it is possible to encourage more page viewing.
5 benefits of introducing a recommendation engine in your business
Here are five benefits of introducing a recommendation engine into your business:
- Up-selling and cross-selling of existing customers will be possible, and the average purchase price and purchase frequency will improve.
- Improve LTV by optimizing customer experience
- Increased dwell time on site
- Can be used to consider marketing measures
- Reduce new customer acquisition costs
Let’s take a closer look.
[Merit 1] Up-selling and cross-selling of existing customers are possible, and the average purchase price and purchase frequency are improved.
Recommendation engines allow users to suggest products and content that they would not have found on their own. You can upsell to a more expensive product or cross-sell related products at the same time. You can expect an increase in the average purchase price and purchase frequency.
[Advantage 2] Improve LTV by optimizing customer experience
When the number of products and contents is enormous, there are many cases where it is difficult for customers to find all the products by themselves.
If you introduce a recommendation engine, it will make the best suggestions for users based on the following data.
- Keywords users are searching for
- Products and content you are viewing
- Purchase history
- Items purchased by people who actually purchased the same product
Improve the customer experience with compelling offers they might not have discovered on their own.
As a result, it leads to repeat use, and you can expect an improvement in LTV.
[Advantage 3] Longer stays on the site
The recommendation engine enables us to present attractive content to users.
Users spend more time on your site as they check out new content.
A joint study by Fanplayr CEO Simon Yencken and the University of Bocconi in Italy found that the longer a visitor spent on a site, the higher their CVR.
On an EC site, a long stay time can be an indicator that the customer has sufficiently compared and considered within the site.
[Advantage 4] Can be used to consider marketing measures
When using a recommendation engine, data will be accumulated in the system, making it possible to more specifically understand user behavior and preferences.
Personas are clarified and can be used for marketing measures.
For example, if it turns out that many people who have purchased A also purchase B, consider measures that better meet customer needs, such as conducting a campaign to sell products A and B as a set. I can do it.
[Merit 5] You can reduce the cost of acquiring new customers
The recommendation engine can present attractive options to customers, making it easier to connect to CVs from the first time. It is easier to acquire new customers, and it also reduces the cost of acquiring new customers.
Points to note when using a recommendation engine for business
Although it is a recommendation engine that has great benefits for business,
- If the number of products that can be proposed is small, the effect will be small.
- Lack of customer or site data leads to lower accuracy
Two points must be noted. Details are provided below.
If the number of products that can be proposed is small, the effect will be small.
If the number of products that can be recommended by the recommendation engine is small, the products that can be proposed will be standardized, so the proposal will not be attractive to customers.
If the number of products is limited to the extent that users can check all products by themselves, it is difficult to demonstrate the effect.
Lack of customer or site data leads to lower accuracy
The recommendation engine can recommend related products (content-based) or according to certain rules (rule-based). By using the proposed mechanism (item-based, user-based), more effects can be expected.
Therefore, if the site has just been opened and there is little data on customers, behavior and purchases, the accuracy will be low. Remember, if you want more accuracy, you need to accumulate data.
3 things to consider when choosing a recommendation engine
When using a recommendation engine, it is common to install the tool on your own site. There are three things to consider when choosing a recommendation engine:
- Check if it can be introduced on your site in the first place
- Check if there are functions such as AB testing and reports
- Check billing type and cost
Details are provided below.
[Point 1] Check if it can be introduced on your site in the first place
In some EC sites, the recommendation engine cannot be used if the shopping cart ASP is used. Start by checking whether it can be introduced systematically.
If the number of accesses or conversions is too small, introducing a recommendation engine may not be effective. It is also a good idea to check the introduction criteria of the tool you want to introduce.
[Point 2] Check if there are functions such as AB tests and reports
The recommendation engine is not completed after installation, but is more effective when operated continuously. Also make sure that the tools you introduce have functions that make it easy to measure effectiveness, such as AB testing and report creation functions.
[Point 3] Check the billing method and cost
As the functionality increases, so does the cost. It is a good idea to determine the functions you need for your company and start with those that have the minimum necessary functions.
Considering that data will be accumulated in the recommendation engine, choose one that allows you to add functions later.
Most recommendation engines charge based on views. If the number of views is large, the amount will be higher than the amount displayed on the site, so it is safe to know the amount in advance if you introduce it to your company or get an estimate.
Learn how to use the recommendation engine and apply it to your business
The recommendation engine is an attractive mechanism that can be expected to improve customer experience, LTV, and CVR by presenting attractive options to customers.
Since it can be installed not only on e-commerce sites, but also on sites that provide content, why not use this article as a reference to consider whether it is possible to introduce a recommendation engine to your company’s website and what effect it will have when it is introduced. Is not it.
