This article outlines how to troubleshoot issues with the search backend related to location searches. Location search is achieved by setting an NLP filter on builtin.location and returns results based on the distance from the user or location in the search query.

Be sure to follow the How to Start Debugging Backend Search Issues help article for general troubleshooting steps first. Once you resolve your issue, you may skip the remaining steps.

 

General Troubleshooting Steps

If location search doesn’t seem to be working, start with these troubleshooting steps below. If a step resolves your issue, you may not need to run through the remaining steps.

1. Open the search log for the query you’re seeing issues for. Either:
   1. Navigate to Search > [[Experience Name]] > Home or a Configuration sub-tab. Run a test search and then select Debug Query from the drop-down menu.
   2. If you’ve already run the search in a staging or production link, navigate to Search > [[Experience Name]] > Search Logs and click on the relevant search log. Use the search bar or filters to help you find the correct search log or run the search again to have it appear at the top.
2. Check location detection: In the search log metadata, check what is detected as the user location. If this is incorrect, make sure to allow location detection in your browser, then run the search again and review the new search log. If this is still incorrect, submit a support ticket.
3. Check NLP filters: In the search log, check if an NLP filter was applied on  builtin.location for the relevant vertical.
   1. If not, check that the NLP filter is turned on for the field by navigating to Search > [[Experience Name]] > Verticals. Add the searchable field, click Save, and then run the search again in Test Search on the right-hand side.
   2. If you already have an NLP filter on for builtin.location, check to see if there is an NLP filter conflict where only one NLP filter will be applied, for example between location search and a location subfield like address.region. You can set an NLP filter order to break ties between the fields. If the conflict ranks above location search, you’ll have to remove it if you’d like location search to be applied.
4. Check country restrictions: Check that your configuration includes the relevant countries to search in, which defaults to only the United States if not set.
   1. Navigate to Search > [[Experience Name]] > Additional Settings.
   2. Check the Country Restrictions field includes all countries that should be searchable.
5. Check bounding box: Check whether your configuration specifies a bounding box and whether the desired searchable region is contained within that bounding box. Bounding boxes will scope searches to a particular geographic area.
   1. Navigate to Search > [[Experience Name]] > Edit as JSON.
   2. Use a tool like bboxfinder to map out any existing bounding box. Confirm this covers the geographic region needed.
   3. If no bounding box currently exists, consider adding one if the business is local to a region to improve the accuracy of search results.
6. Check that results exist: Check that the place being searched has entities nearby that are within the search radius.
   1. Navigate to Search > [[Experience Name]] > Edit as JSON. Check whether a minimum location radius (minLocationRadius) is set (this number would be in meters). If not, the default is 25 miles.
   2. Run a generic vertical search to return all results (for example, run a search for the entity type) and check whether there are entities returned nearby, specifically within the radius found above.
   3. You can also check for entities in the platform. Navigate to Content > Entities and filter for the state/region or country name to find nearby locations.
7. Check for experience training overrides: The desired NLP filter will not show up if it was rejected or another NLP filter was approved for the same search query.
   1. Navigate to Search > [[Experience Name]] > NLP Filters.
   2. Use the search bar to filter to the search term you are troubleshooting.
   3. Click to turn on the toggle for “Show completed”.
   4. Check to see if there are any NLP filter entries for the relevant vertical.
      1. If a location NLP filter was rejected for the desired location, this NLP filter would not be applied for this search query. If you would like it to apply, undo the completed experience training override.
      2. If a different NLP filter was approved, a location NLP filter for the desired location will not be applied.
         1. Undo the incorrect filter.
         2. Toggle “Show completed” off.
         3. Approve the desired result by clicking the green checkmark next to it.
      3. If there are multiple active overrides, undo undesired ones so that there is only one active override.

 

I’m unable to find places outside of the United States

Place Search only searches the countries listed in countryRestrictions. If this field is not populated, it defaults to search only in the United States.

1. Navigate to Search > [[Experience Name]] > Additional Settings.
2. Check that the countryRestrictions field is populated with the countries that you want included in the search and add any missing countries.

For more information, check out countryRestrictions in the Answers Configuration API doc and the Additional Vertical Settings unit.

 

Location Search picked the wrong option between two identically-named places

Read about how the location search algorithm uses proximity bias and place prominence to choose between two identically-named places in the Search Algorithms reference doc.

Setting a bounding box and country restrictions will help to improve the accuracy of results by limiting the scope of the search to a particular region. Check the General Troubleshooting Steps above for how to do so.

 

I searched for a specific city, but locations from other cities are appearing or appearing higher

Location search works by finding the central latitude and longitude of the search token, for example, a city, and then returning the locations that are closest to that latitude and longitude. If locations from other cities are within the search radius, they’ll appear as well. There may be cases when locations in a different city are closer to the city center than locations in that city itself, and so they’ll appear higher in the results. For more details on the location search algorithm, check out the Search Algorithms reference article.

If you would like to filter to only show locations in the token searched, set an NLP filter for that address subfield, such as address.city, which NLP filters treat as regular fields:

1. Navigate to Search > [[Experience Name]] > Verticals.
2. Click + Add / Update Fields.
3. In the search bar, type address. to view the possible address subfields and select the one you want to use. Then click Update Fields.
4. In the Searchable Fields table, turn on NLP Filter for the newly added field and turn it off for builtin.location if desired.