The pandemic might have changed a lot of things (lockdowns, masks on, lots of online shopping), but it still hasn’t stopped vendors and salespeople at department stores running after us, perfume tester in hand.
You know the drill; walk by trying to look invisible, get stopped, face a five minute ramble on a product you’re never going to buy, plot your escape.
You’d have thought you could escape it all with online shopping, but many sites are still littered with useless product recommendations that customers don’t actually want. This is why it’s so important to understand product recommendation techniques and tools—so you can avoid bombarding your on-site customers with recommendations that make them want to jump ship. And instead present them with truly useful suggestions that can boost both customer engagement and revenue.
Before we dig any deeper, let’s go over the basics.
What is a product recommendation?
A product recommendation is exactly what you think it is; a product that’s recommended to a prospective customer, usually personalised to their tastes. Sites can’t just do this willy-nilly, though, and that’s where product recommendation engines come into play!
What is a product recommendation engine?
These engines are essentially systems that collect customer data (buying habits, browsing history, etc.) to create algorithms that hit on what a customer might actually be interested in buying.
Unlike those scattergun demonstrators in the store, a product recommendation engine will put forward useful recommendations based on solid user data.
As machine learning has evolved, so too has the humble recommendation engine. There are now so many options available to ecommerce sites and online retailers when it comes to product recommendations.
Marketers and managers can even use these engines to give customers updated recommendations in real-time, as well as more familiar methods like automating emails with lists of relevant products or a prompt to revisit an abandoned cart.
How does a product recommendation engine work?
What makes product recommendation engines so effective is that they’re based on cold, hard facts. The engines collect and analyse customer data both on-site and off-site, using it to inform their algorithms so that personalised and relevant recommendations can be made to visitors.
As customers keep coming back, engines can store all this data into special user-profiles, leading to even better recommendations in the future.
Do you need to increase your online revenue?
Managers can even get behind the wheel of an engine to determine where these recommendations will be made. From landing pages and the homepage to the product page, products can be recommended at many different points throughout the shopping journey.
Want to increase your ecommerce conversion rate?
Benefits of using product recommendations
If you have an online store, you’ll want to optimise it with product recommendations to really see a lift in your revenue and user experience (UX). Customers are sick of being inundated with useless products, so instead of driving them away, show them that you care about their individual preferences with an engine that delivers relevant, personal recommendations.
Whether you’re an ecommerce site or into merchandising, your site (and customers) will greatly benefit from a product recommendation engine. Still unconvinced? Here’s the top benefits that product recommendations can deliver:
- Relevant content and recommendations
Since recommendation engines collect and analyse data in real-time, customers will always be given relevant content and product recommendations. By using customer’s browsing history and user profiles, engines can create personalised recommendations, showing customers that you actually care about their onsite experience and their preferences.
Engines are also multi-faceted. They can deliver relevant content and recommendations via email campaigns, too. All this personalisation will increase your customer’s confidence in you as a brand, leading to higher conversion rates and more clicks at checkout.
- Engage customers and boost site traffic
By making consistently relevant recommendations, you can stop your customers from having to dig deep for products they really want (leading to a better CX and higher engagement levels).
Personalisation really is the key to your customer’s hearts, so by recommending products on a range of channels you’ll be more likely to increase site traffic.
- Increase revenue
Think about it. If you’re being recommended products that you actually like, you’re going to keep coming back to that site, because it understands your wants. It values your preferences and nails great product listings.
This is the effect of personalisation, and once you start making use of product recommendation engines you’ll statistically see a much bigger increase in revenue.
- Increase average order value
Once you start making use of product recommendations, your average order value (AOV) will be boosted as a result. Recommendation engines will persuade customers to purchase additional items they wouldn’t usually buy, leading to a higher AOV.
- More opportunities for upselling and cross-selling
Product recommendation engines are a great way of subtly upselling and cross-selling to customers, since recommendations can be made in a number of ways. If you’re looking to cross-sell, simply group similar items together in your recommendations.
Product recommendations can also be used to leverage upselling, enticing customers to buy a more expensive version of a product or a more expensive alternative in a compelling way.
At PureClarity, we know the difference a strategically placed recommendation can make to your sales. With our recommendation engine, customers will be updated with personalised offers and recommendations, enabling you to effectively upsell and cross-sell your products in real-time.
How to get started with product recommendations
Though you can manually build a product recommendation engine, it’s bound to eat away at your time and requires a specialised skill set. Unless you have an in-house team of developers, it’s best to install a recommendation engine directly onto your website (it will also save you a lot of time!)
When it comes to searching for your perfect product engine, it’s important to remember that there are three main types on the market:
- Collaborative filtering engine
Simply put, a collaborative filtering engine makes product recommendations based on user data that displays similar shopping habits and purchases. It’s based on the premise that different customers who buy the same or similar things (generally) have a similar taste, and products are recommended on that basis.
For example, if customer A purchases products from a line that customer B regularly purchases from, a collaborative filtering engine would assume that both customers like similar things, and might start recommending items that both have bought to each other.
- Content filtering engine
Unlike collaborative filtering, a content filtering engine uses customer data (such as browsing history) to make recommendations based on products that the customer has already bought or added to their wishlist. It’s based on the premise that if a customer likes a certain item, they’ll be more keen to purchase a similar one.
Netflix is a great example of a business that uses content filtering to appeal to its users. The algorithm looks at a user’s watch history and recommends movies based on what films users prefer to watch. If you’re a fan of period dramas, for example, then Netflix might recommend films and series from that same genre such as Downton Abbey or Jane Eyre.
- A hybrid mix of both
Combining the best of both worlds, you can also create an engine that utilises content and collaborative filtering to make for even more personalised product recommendations.
This hybrid mix is usually specific to each site visitor, using data collected from both methods in combination with the user’s attributes. Spotify is one of many sites taking this hybrid approach, curating personalised playlists based on the songs you listen to most.
What all three engines have in common, however, is their use of machine learning. At the core of each approach is an intricate system, which is why it’s generally recommended to utilise an existing product recommendation engine instead of making one from scratch.
Do you need to increase your online revenue?
7 Techniques in using product recommendations
Product recommendation engines are multi-faceted and can take on a number of different approaches. The most common techniques for recommending items is usually based on the site’s purchase history or the customer’s shopping habits.
Below, we’ve gone through seven of the most common product recommendation strategies, including:
- Most popular or trends
Similar to your best sellers, you won’t need to dig into user data to make the most of this technique. All you have to do is find your highest-rated products and popular items (and it’s even better if you can spot what’s trending) and place them in your product recommendations.
Like best sellers, this technique is especially useful if you don’t know much about a customer or are targeting a first-time buyer. By recommending your hottest products first you’re more likely to get a response. If users can see that a product is in high demand, its value increases and the likelihood of them making a quickfire purchase soars as a result.
It’s all about building a sense of brand trust; once users can see that people are clamouring for a particular item, it affirms that that product is worth it.
- Ratings-based recommendation
Another technique that’s great for hooking new buyers, ratings-based recommendations are another form of social proof and can help build a visitor’s trust in your product.
It’s a no-brainer; if they’re being recommended products with consistently high reviews, they’re going to be wowed.
- Based on recently or previously purchased
Even more powerful than browsing history, basing your recommendations on purchase history is a great way of putting relevant products in front of customers that they likely already have an interest in.
You may have seen it on sites like Amazon and Ebay, where product recommendations are made in the form of, “customers who bought X, also viewed Y.”
Once again, it’s a technique that utilises collaborative filtering to get you interested in what other shoppers are buying. It also develops a sense of trust in the product, since an item that’s been bought widely by other customers is surely reliable—social proof at play, once again.
- Based on browsing history
If you want to get to the bottom of what your customers actually want, analysing their behavioural data when shopping on your site is a great place to start. By looking at their browsing history, you can make use of collaborative filtering and recommend products to large groups of people based on what products they’re viewing.
You might find that customers A, B, and C all have a similar browsing history and love the colour pink. Your product recommendations would then come in the form of, “you might also like”, and display similar items that all three customers would like and might have added to their wishlist before.
Featuring your best sellers in product recommendations gives customers an extra vote of confidence in what you’re selling, increasing the likelihood of a purchase by utilising social proof. That’s the idea that if other people are buying something, it must be good, so you don’t want to miss out.
This is especially effective when visitors land on your site for the first time, as it hooks them in and gets them interested in the products that everybody seems to be raving about.
Unlike most of the other techniques, featuring your best selling items doesn’t require any in-depth analysis of individual user data. You simply have to dig up your most popular products and voila; you’ve got yourself a winning tactic.
- Frequently bought together
Sites that make use of the “frequently bought together” technique usually do so with widgets. Data is collected by the product recommendation engine on all items and the number of times they have been purchased together. It essentially draws on user data and the kinds of items customers purchase together.
- Recommended product pairings
Similar to the “frequently bought together” technique, product pairings are usually also done with widgets, but instead of recommending items based on solid, user data and purchase history, these recommendations are made based on items that generally go well together.
These recommendations don’t just come in the form of similar products, they’re usually complementary.
You’ve probably seen it in action on ecommerce sites, where customers are presented with other products that complement the item they’re viewing. If you were browsing on a clothes store, for example, you might find that the engine recommends other products to complete your look based on items you’ve added to your basket. It’s an easy way to cross-sell and get your customer’s cart loaded with a few more items before checkout.
Are you struggling to increase your AOV?
9 Product recommendation best practices
- Integrate product recommendations in your email marketing strategy
Don’t just limit your product recommendations to onsite visits. Entice your customers to buy more frequently by incorporating it into your email marketing. Your best bet is to analyse user data based on a customer’s purchase history, so you can make recommendations related to products they already have an interest in.
Remember to keep it brief and follow a template that’s not overtly salesly. A newsletter is a nice way of dropping in a few recommendations, just be sure to steer clear of including any prices.
Your job is to compel and pull readers in, leading them to buy later. If you introduce a price straightaway, they’re going to really feel like you’re selling to them rather than giving them any value. It’s better to create an alluring experience that entices readers to look on your website, eventually leading to a sale.
It isn’t always easy to draft up dazzling emails and newsletters that hook AND convert. So, PureClarity’s product recommendation engine can also be integrated with your ESP to give you access to templates and other assistance that helps maximise revenue and customer loyalty.
- Provide access to user’s shopping history
Give customers access to their browsing history to make looking up items easier. They might have viewed an item they really liked, but if they can’t trace it they may end up looking elsewhere (leaving you with one less sale).
- Leverage side-door traffic
To get as many visitors onto your site as possible, make sure you utilise all major channels like social media and email, as well as search engines. Develop your brand across all these channels to reach a wider audience.
- Automate product recommendations
Save time and install an engine to automate product recommendations (more on this later).
- Show best-sellers across the product categories
To really entice customers, display best-sellers on the homepage for different product categories. You can also directly include these on the individual category pages, but either way, it’s social proof that an item is desirable.
- Make sure that recommendations are relevant and timely
You wouldn’t change a product page without doing some A/B testing, so don’t just make recommendations out of the blue, either. Keep everything relevant and timely, always using data. This is why installing an engine is beneficial, as it does all the hard work for you and is based solely on user data and purchase history.
- Know your visitors
Really get to know your customers and regularly analyse user data so you can make recommendations that are personalised.
- Optimise with widgets
Widgets are a great way of padding a customer’s shopping cart with a few more items before checkout, just make sure you place them appropriately (before checkout and when viewing an item).
- Places where you can put product recommendations
Some ideal places to display product recommendations include:
- Cart page
- Home page
- Widgets across site
- Landing page
- Product page
- Category listing page
Introducing PureClarity’s intelligent product recommenders
If you’re looking for an easy, cost-effective method for implementing product recommendations, our intelligent engine might be just what you’re looking for.
AI driven recommender
Provide product recommendations to customers in real-time with our AI driven recommender. Adapting each time a customer visits your site and makes a purchase, user data will be stored securely for future use.
Personalised and crafted to suit each customer, we utilise data to provide recommendations created for the individual. Avoid product duplications and recommend what your customers are truly looking for.
We know that no two businesses are the same, so our product recommendation engines can be tailored to suit your company’s specific needs and goals. Choose from a wide range of recommender types, and have the option of adding manual recommenders into the mix to supercharge your strategy.
- Cold start feature
Experience results from day one with our “Cold Start Feature”. You don’t need any trading history to get started; our recommenders will sift through products with similar functions and properties, providing recommendations to first-time customers with ease.
Our engine simply finds products with similar properties and functions, grouping them together in recommendations to help optimise your campaigns.
- Upsell and cross sell
Encourage customers to add more to their cart before checkout. Incentivising extra product purchases, our engine increases your chances of upselling and cross-selling by recommending high-value products to customers and those that are frequently brought together. Customers will be encouraged to add extra items to their cart with complementary cross-sell recommendations, leading to an increase in revenue.
3 Key benefits
- Real time updates
Our AI-driven recommenders aren’t static; your customers will always be provided with updated product recommendations through our real-time data analysis. Each time a new user visits the site, data is collected, enabling the AI to adapt to current behaviours, avoiding product duplication.
- Website and email integrations
Unlike standard engines, PureClarity’s recommenders can be placed across many channels, including email and on site. Help customers find what they are looking for with our “Personalisation within Search” feature that enables you to place recommendations on your site’s search bar and results page.
- AI and data driven
Our recommendations are based on data and facts; we only display relevant and personalised products to customers.
Product recommendations are a great way of boosting your site’s traffic and revenue, showing users that you value their preferences and can accommodate their tastes. A product recommendation engine, like we offer at PureClarity, is an efficient solution to providing customers with these relevant and personalised recommendations, all without the hassle of a manual set-up.
Create engaging and personalised shopping experiences for your customers today, and try our 30 day free trial to experience the benefits first-hand.
Frequently asked questions (FAQs)
What is a product recommendation?
Part of a marketing strategy, product recommendations are pointers to items that are strategically displayed to users on a webpage or email to create a personalised shopping experience.
How do you recommend a product?
Product recommendations are made by analysing user data such as browsing and purchase history, as well as items frequently bought together.
What are the different types of recommendations?
There’s a wide range of techniques that can be used, but they all generally fit into either collaborative or content filtering methods.
There are a number of metrics you can look at to measure the success of product recommendations, including:
- The locations that have the highest purchase rate (places where you have displayed recommendations).
- Interaction metrics (analysing the CTR on the recommended products).
- Conversion (analysing the AOV increase and increase in revenue).