# The short version

When a question is broad, AI tends to cite a local page. When the question names a specific service or product, AI tends to cite the page built for that intent. In this brief, an intent is a service or product a customer asks for by name. An intent page is the brand-owned page built to answer that named need.

**What to do:** Keep the local page strong, then give each high-intent service or product a clear intent page that says what the offer is, where it is available, and how it connects back to the local page.

# How to read the numbers

Read each bar as the share of brand-owned AI citations that went to an intent page. Gray marks a broad category query, and blue marks a query that names a service or product intent.

A local page answers where the business is and how to reach it. An intent page answers whether the business offers the specific service or product in the query. The two page types solve different citation problems, and the category labels in the figures are representative and generalized from the original customer category wording.

## Finding 01

## Why the Page Changes

Broad category queries usually ask whether a business exists nearby and can serve the customer at a location, which is why 64% of citations for primary-category queries landed on local pages in this analysis.

Intent queries ask a narrower question about whether the business can handle a named service or product request, which requires a more exact page with intent-level evidence.

That is why the page type matters operationally. The local page can answer who and where, while the intent page can answer what is available and what the customer can do next.

### Primary and Secondary Queries Point to Different Pages

| Query type               | Location page | Intent page |
|--------------------------|---------------|-------------|
| Parcel services          | Core category | shipping     | 60.9% | 39.1% |
| Named intents            | service examples | 13.5% | 86.5% |
| Vehicle parts            | Core category | parts       | 83.3% | 16.7% |
| Named intents            | product examples | 48.8% | 51.2% |
| Quick-service            | Core category | menu        | 47.1% | 52.9% |
| Named intents            | item examples  | 23.3% | 76.7% |

Read each sector as a pair. The top row is the broad category question. The second row groups named service or product intents, where the intent page takes a larger share.

70%  
Intent queries landed on intent pages more often than on local pages in the page-citation analysis.

## Finding 02

## Intent Queries Use Intent Pages

In the parcel-services cohort, the broad shipping query usually cited the local page, reaching a 60.9% local-page citation share.

When the query named a specific intent, the intent page became the main answer, and representative intent queries cited intent pages 84.1%, 86.2%, and 89.1% of the time.

The reason is practical. A broad query asks for a nearby place, while an intent query asks whether that place can do a specific job or provide a specific product, so a page about the intent gives the AI answer a clearer source.

### Intent Queries Cite Intent Pages

| Location page                  | Intent page                  |
|--------------------------------|------------------------------|
| Primary core shipping query     | 60.9% | 39.1% |
| Service document disposal       | 15.9% | 84.1% |
| Service document verification    | 13.8% | 86.2% |
| Service ID photo service       | 10.9% | 89.1% |

The primary query usually cites the local page. The three intent queries usually cite the intent page built for the named need.

## Finding 03

## Page Fit Explains the Pattern

Page fit means how closely the cited page matches the intent named in the query. A local page can be useful for the broad category, but it may only mention a specific intent briefly.

The scatter below compares representative intent queries across three sectors. Farther right means the cited page is a closer topical fit for the query, and higher means AI cited the intent page more often.

The pattern is not perfectly even, but the upper-right cluster shows why intent pages matter. When the page is built around the named need, AI has a more direct brand-owned source to cite.

### Higher Page Fit Usually Means More Intent-Page Citations

| Intent-page citation share | Higher dots mean more intent-page citations | Lower page fit | Higher page fit |
|----------------------------|-------------------------------------------|----------------|----------------|
| Document disposal           | 84.1% intent-page share                  |                 |                |
| Document verification       | 86.2% intent-page share                  |                 |                |
| ID photo service           | 89.1% intent-page share                  |                 |                |
| Visibility accessory        | 64.0% intent-page share                  |                 |                |
| Power replacement           | 48.6% intent-page share                  |                 |                |
| Brake parts                | 41.0% intent-page share                  |                 |                |
| Side menu item             | 76.7% intent-page share                  |                 |                |

Each dot is a representative intent query. Farther right means the cited page is a closer fit for the query, and higher means the intent page received a larger share of brand-owned AI citations.

## Finding 04

## The Pattern Appears Across Models

The same page-choice pattern appeared across multiple AI models on one representative intent query, which makes the finding less dependent on a single answer system.

- Anthropic cited the intent page 96.9% of the time.  
- Perplexity reached 93.0%.  
- Gemini reached 84.2%.  
- OpenAI reached 68.4%.

The model shares are not identical, and that variation matters, but each model still cited the intent page more often than the local page for that specific intent query.

### Each Model Cited the Intent Page Most Often

| Intent-page citation share |  |  |  |  |  |  |
|----------------------------|----------|----------|----------|----------|----------|----------|----------|
| 0%                         | 25%      | 50%      | 75%      | 100%     | Intent-page citation share |  |  |
|                           | Anthropic | Perplexity | Gemini | OpenAI |
| 96.9%                       | 93%      | 84.2%   | 68.4%   |  |

The exact share changes by model. In each case, the intent page received a majority of brand-owned citations for the representative intent query.

## Important distinction

## Local Pages and Intent Pages

This report separates two jobs that can look similar in a citation table but serve different reader needs. Local pages answer where the business is and whether a nearby location can help, while intent pages answer whether the business handles a specific service or product request.

The claim is limited to that distinction. A strong local page still matters, but a specific query often needs a page that matches the intent more directly.

## What you can do

## Three Practical Steps

### 01

#### Map the query set

List the services and products customers ask for by name. Keep those separate from broad category terms.

### 02

#### Match the page to the task

Give each high-intent service or product a page that names the offer, explains availability, and links back to the local page.

### 03

#### Keep local context close

An intent page should still say where the service or product is available and which location can fulfill it.

A full scan can come later. The first pass is a simple page inventory against the services and products customers already search for.

## Common questions

## Questions Readers Ask

**Does this replace the local page?**  
No, because the local page still carries the broad category and location answer, while the intent page carries the narrower service answer.

**Should every intent get a page?**  
Start with services and products that customers ask for by name and that the business can fulfill across many locations, then expand once the highest-intent pages are covered.

**Why generalize category labels?**  
The businesses are anonymous, and service labels are generalized so the pattern is visible without exposing exact customer category wording.

## Methodology

## How This Was Measured

**The data**  
The analysis covers Q1 2026 brand-owned page citations for 10 businesses across 1,800 U.S. locations.

**The sample**  
Brand names are withheld and replaced by sector descriptors. Service labels are representative rather than exact customer category wording.

**The comparison**  
Primary-category queries are broad terms tied to what a brand is known for. Secondary queries are narrower service or product intents.

**The measure**  
Page share counts brand-owned citations that land on a local page versus an intent page for a specific service or product.

**The labels**  
Visible category names are representative labels. Original customer category wording has been generalized.

## Future Research

Next studies can expand the sector set, compare more intent types, and track whether newly strengthened intent pages gain citation share over time.