Analytics & Reports

Better Null Search Experience

Business Goal: When my customers search, they may be looking for a product line that I don’t carry. I want to suggest other searches that would bring back products I do carry, and I want those searches by driven by my customer’s behavior - not having to be curated by merchandisers.

This experience can be used on a null result landing page, and it will be powered by popular searches that returned products:


This experience can be powered by all popular searches on your site - to generate a variety of alternative searches - or by using any of the API filters to a subset of traffic that’s specific to your customer.

Sample Request

This query relies on just one factor - the fact that Recommendations will return searches that will bring back results. By querying with a matchPartial against our popular searches, we can see what popular searches contain the searched for phrase and bring back other suggested searches.

In this example, we are using a demo Apparel site and we will use the null query denim shirt:

curl -H "Authorization: --clientKey--" \
    -XPOST "" \
    -d '{"matchPartial":{"and":[{"search":{"query":"denim shirt"}}]}}'

The response will be:

  "result": [
      "count": 1,
      "query": "jeans shirt"
  "serverTimestamp": "2020-01-21T15:10:48+00:00",
  "status": {
    "additionalInfo": null,
    "code": 200,
    "message": "OK"