When it comes to product recommendations, there are two main schools of thought: persona-based and product-based. The former takes into account the specific needs and wants of the individual, while the latter focuses on the features of the product itself. So, which approach is better?
There's no simple answer, as both have their advantages and disadvantages. Persona-based recommendations can be more tailored to the user, but they require a lot more data about the individual in question. Product-based recommendations, on the other hand, are easier to generate but may not be as relevant to the user. Let's take a closer look at them both to understand how we can utilise the strengths of each to maximise sales and Average Order Value.
In the world of e-commerce, product-based recommendations are more common. This is because when a new customer lands on your website you know pretty much nothing about them. But you do know everything about your products, their features, and their sales performance, and as soon as that customer views a product you can use the data you have and look for similar or related products to recommend.
Let's take sun cream - if a customer views a bottle of sun cream you know the factor of the sun cream, and can show recommendations of other bottles of sun cream with the same factor, or you know the brand of the sun cream and can therefore show different factors of the same brand. This is a really simple example that illustrates how product based recommendation algorithms use data about the product being viewed to make the recommendation. The algorithms can be really complex and can take a lot more product features into account - size, make, model, material, brand, fit......the list of features for a product can be far more in depth than the example I've used, but you can start to see already how using those features to make product based recommendations can have a really powerful impact on the product discovery
and customer experience.
Personas are stories that composite together characteristics and interests of different users into one ideal customer, and if done correctly, a persona can be eerily accurate in predicting the needs of a customer. Using personas to generate recommendations is a newer form of marketing that is growing in popularity because it allows for a much higher degree of customization and engagement with the customer. Download your FREE copy of 'Managing AI - The 4 key areas of AI to manage in your online store'
Let's continue with our example of sun cream and explore the difference in how persona based recommendations would work. If I was selling sun cream and I considered my ideal customer I may decide that they are the parent in a family with children preparing for a summer holiday. I can then start to understand the customers needs and wants through their eyes - e.g when planning a summer holiday I may also want after sun lotion. The recommendations are no longer based on the features of the product, they are based on what the customer is thinking and feeling. This opens up so many more products to recommend - when I'm planning a holiday not only do I want sun tan lotion, I also want after sun lotion, bug and mosquito repellent, upset stomach medication, luggage tags, sunglasses, a hat and more. By understanding this I can now recommend all these as products if any of my customers look at sun cream.
Product vs Persona Based Recommendations - which is best?
So, which type of recommendation is better for your website? It really depends on your business and your goals. If you’re focused on selling specific products with detailed technical specifications, then product-based recommendations may be the way to go.
Persona based recommendations are estimated to be up to 30% more effective than product based recommendations. And if you think about it using our sun cream example you can start to see why. Using product based recommendations the outcomes can only be that the customer views sun cream and decides to buy or not. With persona based recommendations not only is the customer viewing sun cream, we are also introducing other products that they are likely to be interested in. The potential outcomes are now customer buys sun cream or not, buys sun cream and additional product(s), buys additional product(s) but not sun cream. We have given ourselves a far greater chance of a conversion by introducing more relevant products, whilst also introduced the chance of greater order value through purchasing more than one product. Ultimately, if you want to increase your eCommerce store's sales
, any recommendations are good, but persona based recommendations will give you the best chance.
Whichever route you choose, make sure you’re using the best software to power your recommendations. GrapheneHC is a great option for both product based and persona-based recommendations. It’s easy to use and provides a deep level of customisation and control of the recommendation outputs. So, if you want to give product and persona-based recommendations a try, GrapheneHC is a great place to start. Book a demo today: Book Now