Understand the what, why, and how of each search result

The Explain tool from the Empathy Platform Playboard helps you understand what’s happening in your search experience. Get insights into Empathy Platform's ranking algorithms and unveil how product scoring determines product positioning on the search engine results page (SERP).

Why are search results ordered in the way they are on the SERP? Why is a product in a certain position or why is it positioned higher than other similar products? Understanding product scores and the factors that contribute to product ranking, such as equalize, query ranking, wisdom of the crowd (your shoppers’ collective behavior), customization (custom business rules affecting the products’ relevance), or the term position in the query helps you to influence product results positioning at all times.

interact

For more information on how products are ranked on the results page and the criteria used to calculate product scores, see Product scoring.

Explaining the what, why, and how

You can use Explain in different ways to enhance product relevance in your commerce store:

Understand how products are scored

Empathy Platform supports multiple complex search configurations that can affect product ranking. For example, you may want to apply different business criteria affecting a particular query in addition to the global weighting set for the product catalog using Equalize. The more configurations you make, the harder it's to guess what happens when these changes are combined, and how they impact the order of search results.

With Explain, you can launch a specific query and preview almost real-time search results with detailed information on each individual product score and the scoring criteria that are applied for the query. Top search results are listed in descending order of relevance according to the overall score, so boosted products always appear in the first positions and buried products in the last. Play with the bar chart and the results grid to see product scores for each product, broken down by scoring criteria in the list shown when the view selection is expanded.

For example, for the query "T-shirt", you have multiple results for Brands A, B, and C. Using Explain, you see that all T-shirts have the same product score. However, one Brand B T-shirt appears first in the bar chart and the results grid as it has been boosted for the period. At the same time, Brand C has 10 T-shirts that appear at the end of the first page of results, since they were the last to be indexed into the product catalog.

Check Product scoring for more details on scoring criteria and product ranking.

Compare products

If you can’t understand why a product is listed in a higher position compared to another for a specific query, you can compare multiple products at once to get a better idea of what’s happening in terms of product ranking.

Take the same example as before, where you have results for Brand A, B, and C. You want to understand why the Brand C T-shirts appear in a specific order. You use Explain to compare the position of one product against another. You click the Brand C red T-shirt and Brand C green T-shirt to see why they have been ordered as first red and then green. Brand C red T-shirt has a higher product score due to the wisdom of the crowd points, so it's placed first; Brand C green T-shirt has a lower score compared to the red T-shirt, so it appears next.

Check Comparing multiple search results to learn how to compare products with Explain.

Visualize Semantics suggestions for no results

It's rare that your shoppers don't obtain results for the query they're looking for, but it happens sometimes. For this no results scenarios, Empathy Platform provides several fallback mechanisms to avoid shoppers' frustration, for example, Semantics recommendations. The Semantics recommendations are relevant product suggestions based on query semantic affinities normally displayed as product carousels.

The products in these suggested carousels are also ranked, so you can break down the scoring criteria involved in the product ranking of each carousel using Explain.

Check Visualizing Semantics results to find out how no results are managed in Explain.

Take action on what you see

Once you understand how product scores determine product ranking, you can make better decisions to optimize search results. Take advantage of the information you see in Explain to determine whether any improvements need to be made to meet your business needs and intention. For example, check whether a specific product needs to be boosted or buried, the weighting of product attributes needs to be adjusted according to your business goals, or a group of products needs to be promoted on the results page based on their product attributes.

To start optimizing search results, take into account the scoring criteria that impact products' organic ranking. Explain gives you direct access to the Empathy Platform Playboard tooling via the shortcuts placed on the top left side of the Product score panel section, so you can instantly take action to improve product relevance.

For example, if you want to promote all Brand C T-shirts, click the Query Ranking shortcut in Explain and boost the Brand C attribute value for the query "T-shirt". Now use Explain to view the new product scores for Brand C products and see how the Brand C red and green T-shirts have changed position in the bar chart and the results grid, and have a new overall score thanks to the business rules points.

Spot the difference

Product ranking is associated with the order in which products appear on the SERP according to the product's search relevance resulting from the product scoring criteria: equalize, query ranking, wisdom of the crowd, customization, and term position. On the other hand, shoppers can choose to change the order in which products appear by using the Sorting option to arrange search results as they like (for example, by popularity, price, newest first, or top products). Product sorting overrides any product ranking criteria defined. All catalog rules, boosts, custom configurations, etc. are ignored in favor of the shopper's sorting option.

Try Explain to…

• Gain insight into product ranking and understand what’s happening behind each search result in almost real time.
• Learn how product scores are calculated.
• Compare product scores for different products.
• Deeply understand your product catalog and how its structure impacts product ranking.
• Increase conversion rates by making informed improvements on product relevance. Reduce the distance between your product catalog and shopper expectations.

note

Ready to start using Explain? Read Understand search results using Explain.

interact

See what Raquel Bernard has to say about Explain in Explain Tool: Understanding your search results (opens new window) and check the Explain Feature Whitepaper (opens new window) (both open new window).