Currently Available: Need a skilled Software Developer for your next project?
Categories
SEO, GEO and AEO

Tracking Image and Video Search After Console Interface Changes

When Google changes Search Console reports or search-type filters, a chart can still look familiar even though it shows a different kind of search activity. To keep image and video reports accurate, developers and site owners need to check which Search Console report they are using and which search type is selected. They should keep that search type attached to every export and dataset.

Search Console’s Search results Performance report measures search traffic such as clicks, impressions, click-through rate, and position. A filter separates the data into Web, Image, and Video search types. Google also added reporting for web multimodal searches, where a person starts a web search with an image. That category is separate from searches in the Google Images tab. This distinction keeps image and video traffic out of ordinary web-search totals.

Start with the report surface

For traffic from Google Search, open Performance → Search results and check both the report name and the active search-type filter before reading a chart or exporting data. Google’s Performance report documentation describes the available search types and notes that the default is Web. Seeing a new filter option does not mean the property has data for that search type. What appears depends on the selected property, date range, and traffic.

Image and Video are search types within this Performance report. The Video indexing report answers a different question: whether Google detected videos on pages it has added to its search index. Use Video in Search results to measure video-search traffic, and use Video indexing to investigate indexing. Google explains that the reports aggregate impressions differently, so their totals do not have to match. The Video indexing report documentation also says the report counts indexed pages where Google detected a video, rather than every video on a site.

Image rows need a similar distinction. In the Pages tab, a URL identifies the page associated with an image result, not the specific image file. Search Console treats an image as a link to its host page and does not separate multiple images on that page in its clicks, impressions, or position metrics. Check the metric definitions for image search before treating page rows as measurements of individual image files. For a quick interface check, see how to identify the report’s search type.

Keep each search type separate

Google’s September 2026 update added web multimodal reporting for image-initiated web searches. It covers activity from Google Lens and Circle to Search on Android, as well as image uploads to Google Search and Chrome’s “Search this image” feature. Those searches belong to Web multimodal reporting; they are distinct from results shown in the Google Images tab. Google’s announcement of web multimodal reporting describes the covered experiences.

Treat Image, Video, Web, and Web multimodal as separate values in dashboards and stored data. Similar metrics do not make the categories interchangeable. Before you present a combined total, check that the selected categories do not overlap. Confirm that they use comparable filters and that Google summarized the data at the same level. Otherwise, keep the series separate and label each one clearly.

For every export or API result, record the Search Console property and date range, then note the search type and dimensions (the fields used to break down results, such as date and page). Also record the summary level and extraction time. Include the type in the row key or table partition. For example, a daily page-level image row and a daily page-level web row should remain distinct even when the date and page match. This is the same principle used when maintaining dashboard accuracy across search-type filters.

Update API and BigQuery reporting

The Search Analytics API uses the request field type; its documentation marks the older searchType field as deprecated. The API supports distinct values for web, image, video, and webMultimodal. Send a separate query for each type instead of requesting one broad dataset and treating its rows as a single organic-search series. For example:

{
  "startDate": "START_DATE",
  "endDate": "END_DATE",
  "type": "image",
  "dimensions": ["date", "page"],
  "rowLimit": 25000
}

Run a separate request with type set to video or webMultimodal when those categories are part of the report. The Search Analytics API reference says the API returns rows ordered by clicks. Google also documents a limit of 50,000 rows per day per search type. The API can omit lower-ranked data from detailed page and query groupings. Compare a detailed extract with an unsegmented total only when both use the same summary level and the API’s row limits do not exclude data. Google recommends daily, one-day queries for managing quota; its performance data guide says data is typically available after two to three days.

There is a documented scope difference to account for when reporting web multimodal data. Google’s Search Central announcement includes Chrome’s “Search this image” feature, but the API reference describes webMultimodal using Lens, Circle to Search, and image uploads without naming that Chrome feature. The documentation does not establish whether the API excludes Chrome activity or simply omits it from the description. Avoid making a feature-specific coverage claim until Google clarifies it.

Before adding a multimodal filter to a BigQuery query, check the current schema and exported rows. Google’s November 2026 API update says multimodal data will be available in BigQuery for properties with active bulk exports, while the current BigQuery table reference does not list webMultimodal among its documented search_type values. Check the exported schema and actual rows before changing a query; do not assume the BigQuery field or value matches the API enum. For background on the export, see querying image and video data in BigQuery.

Validate like-for-like data

When a number changes after an interface update, first compare the same property, date range, search type, dimensions, and aggregation level. A Search Console export uses the active filters, but Google limits direct report exports to 1,000 representative rows. The displayed totals can include rows beyond that table limit, so the export is useful for checking a filtered view but may not contain every row. See Google’s report export details.

If a new type returns no rows, check the property and date range. Then confirm that the property received traffic in that category and that the API or export method supports that search type. An empty result by itself does not establish a site or pipeline failure. For BigQuery, inspect the current table schema and exported values; for the API, confirm the request’s type and dimensions. Keep those checks separate from changes to the site’s markup or indexing settings.

What I'm building

Delegate tasks. Get software.

Give Vroni a GitHub issue, bug report, spec, or rough idea. It reads the repo, plans the change, writes code, runs checks, and works toward a review-ready pull request.

Take a look at vroni.com

Email updates

Usually a new article and a few links I found interesting.

No spam. Unsubscribe with one click.

Leave a Reply

Your email address will not be published. Required fields are marked *