Product recommendations

Personalizing product offer increases conversion rate,
revenue and improves user experience

Thanks to personalization, the visitor does not have to search for products that interest him.
Persoo displays them automatically.

  • Maximizes the offer attractiveness
  • Increases average order value, conversion rate and revenue
  • Improves users’ experience, who will more readily return
  • Saves time and money on manual insertion of related products. Persoo creates the relationships automatically.
  • Personalizes the offer according to the visitor’s preferences (color, brand, price range, gender, destination etc.)
  • Reacts to current trends automatically – if a product or a category reach a temporary high
  • Uses cross-sell and upsell effectively
  • Reacts to visitor’s behavior in real-time

In average we improve e-shops’s performance:

+14%
Revenue
+12%
Average Order Value
+16%
Number of Transactions

Recommendation types

Choose from ready-made templates or create your own

Cross-sell

Offer products, that are related or complement the one, that is in the basket. The visitor will be happy to have inspiration and you’ll increase the average order value.

Steve Jobs: “People don’t know what they want until you show it to them.”

+12,6%
Revenue
+7,4%
Average Order Value
+4,3%
Conversion Rate

More information

Logic: The reason to buy a related product is because it complements the main product.

When to display: When the first/main product is added to the basket. Because when the visitor is in the process of choosing the main product, related products don’t interest him.

Algorithm: Cross-sell calculation uses:

  • Visitor’s behavior – what he prefers, bought, visited or what is in the basket
  • Other visitor’s behavior – what is most commonly sold with this product, this category or similar products
  • Strict rules – e.g. for ensuring compatibility
  • Similarity of product’s parameters – for offering matching colors, brands, gender etc.
  • Calculated metrics – e.g. higher importance for products that have all sizes available, are bestsellers etc.
  • Manually selected related products – if you want to highlight specific products
prebasket

Alternatives and upsell

When the visitor is searching for his product of interest, it’s good practice to show him alternatives to keep him browsing.

You can also use these alternatives to direct the visitor to products with slightly higher price or higher margin.

More information

When to show: On the product detail page.

If the displayed product is sold out, this box with alternatives can be placed higher in the page.

If the visitor comes from remarketing campaigns, the alternatives should not be displayed as not to distract the visitor from finishing the order.

Logic: The selection of alternatives is based on:

  • Parameters similarity (category, product type, brand, color, gender etc.)
  • Other visitor’s behavior (what others viewed or bought after viewing this product)
  • Visitor’s behavior (depending on what he prefers – viewed, bought or put in the basket)
  • Other limits (higher priority to upsell products, or strict upsell)
alternatives

Bestsellers/top products

In certain situations, the visitor will believe the crowd, and rely on the fact that the product is sold often. This removes the burden of choosing.

Top products can also be new items that have high potential.

More information

When to show: Popular products are typically shown on the category pages.

Logic:

  • Bestsellers: Usually show most sold products in a given period.
  • Trends: Our algorithms quickly catch current trends by watching ratios of views/transactions.
top products

Homepage recommendations

Homepage is and important landing page, where we can use the information gathered on the (returning) visitor.

Since we know the preferences of those returning visitors, we can show what is relevant, and help start the session.

More information

Logic:

  • New visitors will get a variety of bestsellers from multiple categories.
  • Shopping in progress: Persoo knows, what the visitors looks at, and offers similar products.
  • Return after longer time: Persoo remembers preferences of each visitor. It can recommend repeatedly bought products or relevant cross-sell.
  • Persoo varies the product offer, so we don’t waste any single opportunity to make the visitor interested in a product.

When to show: Always on homepage, but vary depending on the shopping cycle and segment.

homepage

Last viewed products

Onsite remarketing of last viewed products makes the navigation easier and improves user experience.

Display even those products viewed on other devices by the same visitor and simplify the path to transaction.

More information

When to show: Always when the visitor viewed at least one product. It’s ideal to have one place, where the visitor will remember he can find the last viewed items. It’s usually above the footer or in a sidebar.

Logic: Last viewed are not ordered only by time, but also by the interest shown. If the visitor saw some product six times, it will stay among the last viewed longer, before other products push it away.

last viewed

Nonexistent/error page (404)

In the unlikely event of the visitor finding a 404 page, he should have a way to navigate away easily.

Display products that might interest the visitor to prevent him leaving the site.

More information

When to show: Tn the 404 error page

Logic: The recommendation is a combination of:

  • Of personalized offer according to the visitors behavior (preferences, related products to bought products, etc.) 
  • General bestsellers and trending products
  • Last viewed products
Img_404

Simple deployment

1  Web

Set up the Javascript snippet
to all pages.

2  Products

Insert the link to an XML feed
(by Google specifications or custom)

3  Logic

Select the algorithms for product selection,
set-up conditions and location.

Deployment options

Quickly and without the need for IT. Choose the template in Persoo administration, CSS selector of the location, and let Persoo display the content in the same way as ads are inserted into the page.

The HTML inserted can use your CSS styles, so the personalized content is indistinguishable from the the other content.

Custom box rendering. Persoo can only supply data, if you want to handle content from your server. Persoo will only send the IDs of products to be shown. This can allow for displaying real-time product availability and custom prices.

Persoo also has REST API for delivering personalizations.

Fine-tune the recommendations

Persoo can use all the parameters you have available. Author for books, color for clothing, destinations for vacations etc.

You can change the rules, that Persoo uses:

  • To ensure compatibility
  • To offer upsell
  • Specific algorithms for unique situations

Detailed statistics for recommendations

To be in control of all the recommendations, Persoo offers the following statistics for each box:

Real impressions and clicks

Real impression happens, when the visitor really sees the offer. Rendering the offer outside screen area does not count as an impression.

Box CTR

CTR calculated as the number of clicks / number of impressions.

Impact on revenue

If the visitor clicked any recommendation and then purchased something, we attribute the value of the purchase to the recommendation box.

The attribution model distributes the transaction’s value evenly among all boxes clicked before transaction (in the last 30 days).

Successfully recommended and purchased products

If the visitor viewed, clicked and purchased a product from a recommendation box, this purchase is counted as directly affected by the recommendation box.

If the visitor clicked the same product in multiple boxes, the revenue is divided evenly among all participating boxes.

These KPI allow you to evaluate AB tesets of various algorithms and conditions.

These metrics can also be exported into CSV files.

A/B tests and experiments

Anything in Persoo can be A/B tested. Whether you want to test different content, compare algorithms or targeting conditions, it’s easy.

ab test screenshot

Start With Personalized Recommendations Today

I am interested in free demo

Fill in your contact info and we will send you
demo account access and more details.



* = required item