Get more out of your product recommendations — Part II: Position
This is the second part in my series on Product Recommendations. You may find the first part here.
It may seem obvious, but did you know that people spend 5x more time viewing the top 20% of a page compared to the bottom 20%? Or that only 22% of pages are ever scrolled all the way down to the bottom? — Keeping this in mind when positioning product recommendations is likely to impact the performance of the same.
Positioning recommendations high on the page will help your visitors to find it, but if you position them at the very bottom they may still get a decent amount of clicks because, well when you’re at the bottom of the page you either click somewhere or leave the page altogether.
It’s a good idea to formulate some kind of aim or goal with product recommendations in order to achieve the desired outcome. Is the main aim to:
- Guide the visitor to the product they are looking for?
- Make the visitor aware of products that could be of interest to them?
- Fill up space to make the page more crowded?
Most would probably say that the main aim is to drive an increase in revenue. But that also means that many merchants unconsciously use product recommendations only to fill up space on the page. Don’t do that.
Instead utilise product recommendations to help the visitor navigate your store in a convenient way. Either to find the correct variant of a specific product or product type, or to find new items they might be interested in.
Some guiding principles to position recommendations are:
Product page principles
On a product page of a product with multiple similar variants a recommendation block displaying similar products close to the product in question is beneficial. Simply show the visitor that you have a wide assortment of the product type.
If there are few similar variants a better approach is to display products frequently bought together or products that might be of interest. The rationale being that if the visitor don’t like the product they are looking at and you don’t have any similar product to show, the next best thing is either to present something else of interest, or to show products that goes well with the current product.
If you are using an AI powered recommendations engine you can combine the product being viewed with other information to derive the best recommendations to display on the product page. You may for example combine the product currently being viewed with the products already added to cart to figure out what products the visitor may be interest in.
Since these recommendations are derived to match an individual visitors behaviour, it makes sense to display those high on the page to make them easier to discover for the visitor.
Summarising the product page guiding principles a recommendation block should not be positioned at the bottom of the page since it is intended to help the customer navigate to the right product.
Start page principles
In general product recommendations do not perform very well on the start page. Mainly because there is no or very little information on the visitor when they first enter the store. There could however be cases with returning customers / visitors that provide information to act upon. An example would be to display previously viewed products when a visitor returns, or a set of products from a previously abandoned cart. If these kind of recommendations are used, it makes sense to display them high on the page since it is intended to help the visitor complete a purchase.
Cart and checkout page principles
Cart as well as checkout page positioned product recommendations are intended for cross-sell ans up-sell as we don’t want to confuse the visitor with products too similar to the ones already added to cart. Frequently bought together type of recommendations are well suited for these pages but should be kept fairly discrete not to interfere with the checkout process. For best customer experience the recommendations may well be displayed below the cart or in a sidebar on the checkout pages. It can also be beneficial to reduce the number of products being presented here.
Post purchase recommendations principles
The post purchase recommendations are intended to quickly add a second purchase from the same customer. these usually includes a discount and a few select products to chose from. The downside of post purchase recommendations is that the customer has to go through checkout one more time. This could however be solved using invoicing or a one click checkout solution. The upside of recommending products post purchase the richness of data available to make a really good recommendation. You suddenly know who the customer is, where they are from, what they recently bought, what the browsed but didn’t buy, what was added or removed from cart and much more.
Wrapping up this part it is important to consider why you display product recommendations. When you know the answer to that it becomes easy to position them to serve your visitors in the best way. And if you’re struggling with poor quality product recommendations or you spend a lot of time to manually select recommended products, check out zubi.ai or reach out to me for personal help.
