What is The Difference Between Recommendation and Personalization?
Before anatomizing the examples of personalized recommendations, it’s important to clarify 2 definitions: people often use the words ‘recommendation’ and ‘personalization’ interchangeably.
Start with the barest distinctions in the e-commerce-related definitions of these two terms! ‘Personalization’ is an extended method within e-commerce website optimization and we can see several areas where stores apply it. Even in the practice of product recommendations. Product Recommendations can be personalized (with user behavior profiling using personalization engines) or non-personalized (using the data mass of item attributes and other purchases).
Personalized or Non-Personalized Recommendations Convert Better?
Sometimes the non-personalized, sometimes the personalized recommendations entail more conversions. Professional recommendation engines usually have to make thousands of decisions in every second: ‘does customer X have enough history to get personalized offers that might imply a higher probability of conversion? Or shall we ignore the user’s profile and apply general item-to-item recommendations?’- These are special decision matrixes (fallback scenarios) that recommendation engines use very often. Sometimes they have to decide between almost equally important versions. If the engine works in a very data-rich environment (enterprise-sized e-stores with ten thousands of buyers) in a given second plenty of personalized, non-personalized and hybrid scenarios compete each other. These can be based on the given user’s history, the similar users’ histories, product sales histories etc., using the results of AB tests and the calculations of data scientists.
Types of Logics
Standard logics: the buyer has a thin profile or has no profile at all. Common recommendation logics can be authored: ‘most trending’, ‘freshly arrived’, ‘recently viewed’ etc. You can create simpler recommendation logics manually as well.
Advanced logics: Unlike in the case of the standard logics, the buyer has a meaty profile with a sound history. The product recommendation she is given is based on meticulous calculations. Or the visitor has no history at all, but the product pool is very rich and the traffic is heavy. In this case, you can set up item-to-item logics. To launch these sophisticated recommendations, you have to say farewell to the manual tinkering. You will certainly need a recommendation engine provider.
Personalized Recommendations are more likely to belong to the advanced recommendation logics. First of all, let’s have a look at these personalized recommendations with a walkthrough in the 3 prominent dimensions: On-site, Off-site, and Offline-to-Online.
Onsite Widgets serving Personalized Recommendations
1. Main page
The visitors of our main page should have already made some footprints at us if we want to display personalized recommendations for them.
First-time Visitors on the Main Page
Newcomer customers are not assumed to look for something specific. The purpose of main page recommendations for them serves the goal of informing the customers and getting them engaged. For first time visitors the recommendation engines usually show item-based offers (latest deals, most trending, freshly discounted, featured etc.). After the visitor made several movements (clicking on an item or a category like ‘teakwood tables’) and after returning to the main page the recommendation engine is able to present the first ‘recommended for you’ widgets: these can contain similar or matching products, together with item hierarchy-based accessories (I clicked on a rowboat, the engine offers a pair of oars).
Returning Visitors on the Main Page
What happens after the user gave consent that cookies can deliver information to the store and the recommendation engine provider? The cookies will store a plentiful user history that you can use to trigger personalized offers across all pages.
Displaying new products for a profiled visitor: featuring your e-store’s new products and collections in a general way for a newcomer customer is obvious. But a returning user definitely needs a fine tuning. For example, if your customer bought a road bike you can show new arrivals from that category and its accessories.
Seasonal offers, Daily deals, Rating-based recommendations for profiled visitors: you can also beef up these seasonal and impulsive categories with personalized recommendations. Build these offers on your customer’s previous year- or previous season activity.
Daily and weekly deals generally build trust in the visitor, because they represent a well-groomed and maintained e-store. Especially when algorithms filter the daily deals to display the user’s preferred items.
2. Product page
Product information page is the place where visitors find detailed info packages about the product. A significant number of visitors arrive here from non-direct traffic (PPC campaigns, price aggregators, ad displays on affiliate sites). The objective here is to keep the visitors in the sales funnel, and display the most relevant items to them. You have to stretch their purchasing willingness by recommending accessories and launching upselling efforts.
You can try many different combinations here, apply easily detectable widgets, mix their content from item-based and user-based recommendations.
Conduct experiences, Run ‘ABC multivariate’ tests!
You can test different recommendation algorithms against each other on your product pages. Try this by adding more placements or by splitting your traffic and recommending by using different logics in one widget.
You can try different approaches on your product pages:
1. ‘Customer who bought/viewed this…’ – an item-to-item collaborative filtering. This logic follows the relation of items by looking at how many times they’re present together in the histories of other users.
2. Recommendations based on the user’s preferences, without collaborative filtering.
3. Hybrid: ‘Customers who bought/viewed this…’ collaborative filtering mixed with the user’s preferred brands.
To launch these experiments, you have to possess a decent amount of user data. If you don’t have enough, set up your test with a simplified version with 2 components. But rich product catalogs and ramose category structures call for a recommendation-as-a-service provider.
3. Cart page / Checkout page
Checkout / Cart page recommendations find the buyers in a very favorable, elevated emotional state! They are about to complete the purchase, therefore they are more likely to accept a 5-10% higher priced product. Recommending similar products or accessories here can result in higher average order values.
Checkout page is an excellent place to launch ‘also bought’ or accessories-type recommendations. Once the users get to the Checkout Page, we already have plenty of information about them. They generally have 3-5 steps long customer journeys. On the checkout pages, the recommendation types must be formulated by their preference.
– The products they viewed. Recommend items that they previously viewed and left (and other, similar items)
– The products they (previously) purchased or added to the cart. Offer products that other shoppers also purchased or items that are connected to their previous purchases.
– Their cart value and the difference if you offer free shipping from a certain basket size. Remind them and in any event, recommend a product to help them to qualify for free shipping!
– Their preferences for product categories, price and attributes (color, brand etc.). Show them other, similar items (that are popular/on sale). These should come from their preferred categories and product attributes.
4. Category page
How to launch product recommendations on the category page? After you have some information about the users you can start experiencing with the same sets of recommendations as we suggested in the product page section.
The dilemma of “Exploration or Exploitation”
On the category page, we often face two conflicting tactics. Shall we utilize the exploration or the exploitation? Shall we help the customer to maximize the capacities of the category container, and tuck all the relevant products there? Or help the customer to explore other products? Products that do not belong to the category but have strong relations to it? (example: ‘Customers who viewed this category viewed these related products too’).
5. Onsite pop-ups
The same rule applies to the recommendation widget pop-up windows as to the pop-up windows generally. First and foremost, your pop-ups must not irritate the customer! Likewise, they must not increase the bounce rate significantly. Before implementing pop-up recommendation widgets, ask yourself: Can’t I substitute them with proper recommendation widgets? Should we block the natural customer journey here with a pop-up? Run AB tests, get the plain recommendation widgets and the pop-ups competed. And examine the buying process in its whole span: some pop-ups may increase the bounce rate on the given page, but in the end the conversion could be higher as well.
1. Recommendations in E-mail
We have recently collected enough behavior data on the course of user visits and purchases! It’s time to utilize them by launching email campaigns to boost our revenue! You can consider an email campaign as an extension of your online store. You can recommend items almost the same way as you can recommend on-site. Personalized e-mail recommendation strategies vary considerably from user to user: for example, one of your visitors might have been interested in a product very much – you can author personalized, recommendation-based email campaigns, the data is easily available about:
- – Each of your customers.
- – Each of your visitors – non-purchasing visitors should be logged in to become trackable by their email addresses. So, you have to grab them onsite first.
2. Recommendations in Facebook Messenger
For Messenger-connected e-stores, 2017 was the year of embracing a quicker response policy. Because the consumers appeared to show less patience to wait on the phone or await email responses.
Messenger-displayed product recommendations can be considered as a vital and more and more important category of personalized recommendations if the user data is collected on a passive way (recording user behavior on-site and importing the on-site-recorded user behavior into the messenger app, where, after recognizing the user, launching recommendations to him).
To process these depths of data migration and identification is certainly a ponderous work! In addition, the laws of data processing protect the e-customer more and more extensively. The results (relevant recommendations in our messenger flow) still look strange and evoke uncomfortable distrust in the user: ‘are they spying on me in the e-store? Furthermore, sell my data to the Facebook?!’ The sponsored offers occasionally popping in our messenger stream still show way lower relevance, which makes us think that this data stream is still very feebly applied in the world of e-commerce. The prominent user preference mapping still comes from the active channel: the bot usually asks the user and accordingly recommends the solution, which is undoubtedly somewhere halfway between the ‘search’ and the ‘product recommendation’ function.
Staying in the Messenger Field, or redirecting Customers to the Store?
With chatbots, 2 ways open for you:
-You can redirect the customers to the appropriate pages bearing the recommendation boxes where the possibility of sale is the highest.
-You can display the personalized recommendations right in the messenger window. However, the scarcity of space allows only one or two products plus the initiated conversations. Recommending 4+ products in one block in a chat window can be a bit too much and become counterproductive.
In short, when you get chatbots to recommend products, ‘inchat recommendation’ or product link are the options, or you can experiment with sending filtered search links too.
Connecting online and offline snippets of identities is a challenge, a jigsaw puzzle work. Last year Yusp pioneered in connecting these 2 distant areas with a well-benchmarkable solution. Using loyalty card has long been in Cora Romania’s retention marketing arsenal. In early 2017 the retail giant looked for means to utilize and leverage the 6 years of customer data accumulated in their Loyalty Program. The RFM-segmentation Yusp provided with its personalization engine generated unique and tailored promotions for each individual loyalty card holders. The blended technology Yusp implemented for Cora is generating 7500 additional store visits / week with an average 24 USD basket value.
How to Choose Logics for my Personalized Recommendations?
Although every e-Commerce store is different, a significant amount of client needs finds the ‘off-the-rack solutions’. These are recommendation algorithms that are already lab-tested by engine providers and come ‘built-in’ for you. Test the above-mentioned standards and advanced methods for yourself, and mix them with your unique approach! To sum up, the insights will help you to understand more the user behavior and ultimately increase your sales.
Ready to get started? Yuspify offers a 30-day free trial. Check it out yourself, it’s on us!