Brandon Lazovic

Pulse · August 25, 2026

Observed

Google now shows an AI Overview above the stock-price widget for ticker searches

Google search results for ticker queries such as GOOG stock or MSFT stock now show an AI Overview above the stock-price widget, reversing a layout search marketers documented in June 2026 when the AI Overview appeared below the chart. Len Raleigh spotted the change on X on August 24, 2026, and Search Engine Roundtable's Barry Schwartz independently replicated it the next morning. Google has not commented on the change.

In plain termsThe widget here is the boxed stock-price chart Google already shows for a ticker search, and Google is now stacking its AI-generated summary on top of that box instead of underneath it, the reverse of the layout from a few months ago.

Why it mattersA placement change like this can shift click share away from the stock-price widget and toward whatever sources the AI Overview cites, changing who gets the click on high-volume ticker queries.

Our takeThis kind of placement flip is easy to miss because it ships with no changelog entry, and a single screenshot can be a short-lived experiment rather than a permanent layout change. The safer read is to recheck a handful of tickers over the next few days before treating an AI Overview above the chart as Google's new default, since this same query type already flipped once before, from below the chart in June to above it now.

Confirmed

Google's crawlers don't parse JSON, and Merchant Center likely shares that pipeline

Google's Gary Illyes said on Bluesky on August 25, 2026 that Googlebot and Google's other crawlers never parse JSON. They only download the file, and a separate system, such as the Search indexer, does the actual parsing. Asked whether the same applies to Merchant Center, Illyes said MC is almost certainly built on the same shared parser, though he has not worked directly with that team.

In plain termsIt works like a mail carrier and the person who opens the envelope, where the crawler only delivers the file and a completely separate program does the actual reading of what's inside, so a clean delivery says nothing about whether the contents made sense.

Why it mattersFor anyone debugging a Merchant Center feed error, this confirms the fetch step and the parsing step run in separate systems, so a successful download of your feed file says nothing about whether its field values were read correctly.

Our takeI argued in Google's crawler split that the Shopping crawler appears to be an HTML-only reader for structured data, based on a Google engineer's spoken comments rather than a published spec. Illyes' statement is a second, independent data point for that same fetch-versus-parse split, generalized past structured data to JSON broadly, and it makes the shared-infrastructure read on Merchant Center a reasonable bet rather than a confirmed architecture.