How effective are the personalized product recommendations for eCommerce?
What and how to affect the ability to convince a visitor to the purchase, how they work and how much bring to increase the sales volume?
Not widespread and little-used in the market, they are very present in other realities to online sales with more realism and less capacity and blinkers.
Several online vendors that follow use them with success and increases the total performance of absolutely effective online sales, both in relation to sales that bring the cost of using these are Software as a Service. The greater the use in time, the greater their efficacy, as a function of the ability to learn and respond to the browsing habits of a visitor.
In practice this is to install on your eCommerce software that recognize the visitor who returns on the online shop, attracting their attention by presenting, in boxes dedicated for the purpose, the products we saw earlier and which best corresponds to their intentions navigation, to their interest and ultimately its purchasing intentions.
Personalized product recommendations, and revenues
For August 2015, Marketing Sherpa has published a research based on 1.5 billion of shopping sessions over a period of three months and of products ranging from clothing, to sports, food, wine, books, music, electronics, and furniture.
The dynamics recommendations are generally set on the home page, on the category page, product page, on the shopping cart page and are personalized on the visitor’s data, on its comportment and the history of its navigation.
The chart above shows the 10 kinds of recommendations that have the best results for receipts brought eCommerce.
Where to put the recommendation box
Do yourself a question and give yourself an answer.
If Amazon puts the recommendations highlighted in first place in view of the customer, there may be a valid reason?
Customized recommendations are highlighted, in the foreground, the view of the visitor on the home page and for a very obvious reason to the category page is personalized recommendations.
No general recommendations of the seller, often based on nothing when not on sale completely wrong assumptions.
The basic rule that I should mention to every seller is that if a product does not sell, do not sell.
And an indisputable reality and it is not highlighting a product that nobody wants you increase your sales. The effect will always be a general detriment of sales.
One does not enter a shop, if exhibited in the showcase unsaleable goods. When a product does not sell will adopt other tactics: it is removed from the sale or drastically cut the price, or to do more, but not stands out as such.
Instead, there are many online merchants’ lenses, which do not move their inventory and they build with “best seller rigid box”, the “featured products”, the “best discounts”, often hoping to cheat the visitor with static presentations needlessly relegated to the bottom of the page.
The titles of the recommendation box have a different conversion capacity
Do not put titles at random, do not put the random recommendation box, you do not make evidence which is not supported by A / B testing, they do not bring non-verified and supported by unverified assumptions recommendation box into place.
“Visitors who viewed this product also viewed” is the product recommendation that generates the most revenue.
“Visitors who viewed this product also viewed” was the kind of recommendation that generated the most revenue – more than two-thirds (68.4%) of all revenue.
Another significant revenue generator is “Visitors who viewed this item ultimately purchased” This is followed by “You might also like”, which is a highly personalized product recommendation based on current and past behavior of visitors on the site.
These results are not surprising since social proof, or our tendency to assume that an action is more correct if the others are doing; it has often been tested, has proven to work and as a result has become a buzzword in marketing.
You may also like to read another article on Tradenligne: Customer Reviews and E-Commerce: 5 growth strategies for your online shop
“Customers also bought” is abused by eCommerce sites
Tactics that most eCommerce sites use are those that generate the most revenue?
The research data show some discrepancies between what sellers think positively influence their customers (percentage of sites that use a type of recommendation) and what the customers really do (percentage of revenue from a recommendation type).
The increased revenues are always brought by customers purchasing multiple items or customers who buy items of greater value.
And this too is a fundamental rule of the sale, wherever they take place.
Two categories have shown the most other discrepancies between the use and effectiveness.
- 30.9% of sites use “Customers also purchased” on their shopping cart page, while only 7.9% of revenues came from this type of recommendation.
- Although 42.5% of websites use “may also like” on the entire site, only 16.1% of revenues came from this type of recommendation.
Another notable discrepancy is “Top Sellers on the site”: 22.3% of the sites sites involved in the research is using, while only 8.5% of revenues came from this type of recommendation.
“Visitors who viewed this product have also purchased” it is underused by eCommerce.
The tables have turned – a higher percentage of revenues and a smaller percentage of sites that use this type of recommendation – with “Visitors who viewed this product have also purchased” on the product page. Although 25.1% of revenues came from this recommendation, only 7.3% of sites use it.
What does this mean for marketers and for the store manager?
Overall, online merchants may overestimate the value of the suggestions as “Customers also bought” on the cart page and “may also be interested / like” site wide. They may underestimate the value of presenting “Visitors who viewed this product have also purchased” on the product page.
If you’re deciding what types of user recommendations, to decide when and where to present them should be the first very important question. The goal is to make navigation of the customer and the process of convincing the simple and rewarding as possible, adding value without taking the intention to purchase the origin of the research.
5 tips to create effective product recommendations
#1- Relevant and actionable suggestions
Easier said than done. To understand what is relevant to a shop on your site, you should have a system that measures the behavior and preferences of buyers and react quickly to signs. A good recommendation system does exactly that. A consumer, who selects the products in size XL, should not then see recommendations of products that are not in the XL size during a shopping session.
#2- Build engaging recommendations
Such as using dynamic images as a good way to make attractive product recommendations and easy to operate. If a customer is over the image of a shirt, the picture may change to a different view without having to click on the product.
Dynamic content increases the commitment and requires little effort on the part of the visitor.
#3- Understanding the customer’s needs in its path to purchase
There are two blunders they commit many online merchants.
- Do cross-selling on the product detail page.
The consumer who navigates on the product detail page is still in a research phase and is looking for the right product. And ‘correct to present similar options, then upsell on the product detail page to help visitors quickly find the right product.
- Show similar products on the shopping cart page.
At this point of purchase, the customer is in close vicinity to close the purchase and should not be distracted with similar options. The shopping cart page is the place to try out cross selling.
# 4 – Customers must have a sense of control and security.
Privacy is a constant concern of surfers and online customers. It should be explained that you are using a recommendation engine, it should be explained that the customer can erase their history, and you could ask for comments and use surveys to obtain feedback on personal recommendations.
# 5 – The recommendation for the product are a customer service, not a technological mystery.
We are talking about automatic systems for recommending products to customers. Good. Do not let technology be the only method to use to get to a final goal: to help customers find what for them is important and valuable.
The automated technology, marketing automation, is an effective way to accomplish this, but it is necessary that the seller should focus on all the ways that help customers discover the products, for more information, to manage a personalized relationship with the seller.
The sale does not stop to personalized recommendations, but it lives and thrives with all the actions that facilitate customer purchase decisions.