EEAT Mechanics

Search plan (fan-out): the query plan travels into the second pass

One schema object, PlannedRetrievalAnnotations, suggests that the plan of searches behind a query is kept as an annotation after ranking. Google confirms the fan-out technique for AI Mode; that this object holds its plan is RESONEO's reading.

By Thomas Wawra· Published · Version 1.0· Systems referenced: Mustang / SuperRoot etc.

01What the schema shows

This entry has a single schema object, PlannedRetrievalAnnotations, in the quality_twiddler namespace. An annotation is information left on the query or on results for later steps; it does not move anything by itself. The name says that some retrieval was planned. Any twiddler that reads it only recommends; the framework in Superroot, the layer that fans a query out to Google's indexes and merges what comes back, decides.

RESONEO connects it to query fan-out: a question is first turned into a plan, a list of searches to carry out, and that plan remains attached to the query during the second pass, where twiddlers can read it as one more signal. Which twiddlers read it, and what they do with it, is not known.

Source per claim (2)
C
The search plan entry consists of one schema object, PlannedRetrievalAnnotations, in the quality_twiddler namespace: an annotation, not a boost or a filter.[1]
C
RESONEO reads the object as the fan-out plan, the list of searches carried out for a question, which remains attached to the query in the second pass as one of its ranking signals.[2]

02What Google says about query fan-out

At Google I/O on 20 May 2025, Google described query fan-out as the technique under AI Mode, its conversational search mode. A question is split into subtopics, and many searches run at the same time on the user's behalf. Deep Search, a more thorough option inside AI Mode, uses the same technique on a larger scale and can run hundreds of searches.

The post does not use the word plan. It describes splitting a question and running searches, which needs some list of what to run, but how that list is stored is not public. The step from fan-out to this schema object is RESONEO's.

Source per claim (3)
O
AI Mode uses query fan-out, "breaking down your question into subtopics and issuing a multitude of queries simultaneously on your behalf."[3]
O
Deep Search in AI Mode uses the same query fan-out technique at a larger scale and can issue hundreds of searches before writing a cited report.[3]
C
Google's I/O post does not speak of a plan; linking PlannedRetrievalAnnotations to the fan-out technique is RESONEO's inference.[4]

03E-E-A-T and the system map

An annotation recommends nothing, so this entry belongs to the twiddler layer itself on our map. It has no E-E-A-T bearing: a plan of searches says what was looked for, not how good the results are.

For site owners the practical side of fan-out is known from AI Mode: a page can be found through one of the subqueries rather than the original question. What this annotation changes in ranking after the score is not known.

Source per claim (1)
C
The search plan annotation has no E-E-A-T bearing; it records what was searched for, and its effect on ranking after scoring is unknown.[4]

04References

This Deep Dive is Layer 2 content. It interprets the material and always points back to Layer 1, the reference layer. Every claim is mapped to a source, labeled by type: [A] Sworn testimony · [B] Verbatim from the leak · [O] Documented by Google · [W] Peer-reviewed study · [S] Checked on the page · [C] Our assessment · [P] Patent.

© Thomas Wawra · Senior SEO Manager · wetter.com — a Funke Digital company