llms.txt for professional-services firms

llms.txt is a small markdown file you place at your site root to hand AI crawlers a clean summary of what your firm is and does. It's easy, it's free, and it's worth doing. It is also over-sold — so this guide gives you both the working template and the honest caveat most vendors skip.

What it is

Proposed in 2024, llms.txt is to language models what robots.txt is to search crawlers: a predictable file at https://yourfirm.com/llms.txt that states, in plain markdown, who you are, what you do, and where the authoritative facts live. The idea is to reduce ambiguity — instead of a model inferring your services from scattered marketing pages, you tell it directly, in a format it can parse without guessing.

The structure

The convention is deliberately minimal — an H1 with the firm name, a one-line blockquote summary, then themed sections of links or facts:

# Firm name — a single H1.
> One-sentence summary — a blockquote a model can lift verbatim as your description.
## Key facts — a handful of factual bullets: specialty, fee model, service area, credentials.
## Sections — labeled links to the pages you want treated as authoritative (services, contact, about).

A complete template for a professional firm

Copy this, replace the bracketed values with your firm's real, verifiable facts, and save it as /llms.txt at your site root. Keep every claim conservative and true — an llms.txt full of puffery is worse than none, because a model that catches one unverifiable claim discounts the rest.

# [Firm Name]

> [One sentence: what you do, for whom, and where. E.g. "A fee-only
> registered investment advisor providing flat-fee financial planning
> for physicians and dental practice owners in the Austin, TX area."]

## Key facts

- Service: [e.g. fee-only financial planning and investment management]
- Fee model: [e.g. flat annual fee; no commissions; fiduciary]
- Who we serve: [e.g. physicians, dentists, and their practices]
- Service area: [e.g. Austin, TX and remote clients across the US]
- Credentials: [e.g. CFP(R) professionals; SEC-registered RIA]
- Founded: [year]

## Sections

- [Services](https://yourfirm.com/services): what we do and how we charge
- [Who we help](https://yourfirm.com/physicians): our core client type
- [About](https://yourfirm.com/about): the team and credentials
- [Contact](https://yourfirm.com/contact): how to reach us

## Contact

- Email: hello@yourfirm.com
- Location: [City, State]

Two rules keep it safe for regulated firms: (1) state facts, never predictions or superlatives — "fee-only fiduciary," not "the best advisor in Texas"; (2) never include a testimonial, review, or performance claim, so nothing here trips the SEC marketing rule or a state bar advertising restriction.

The honest caveat: how much does it actually help today?

Adoption is ahead of consumption. A 2026 million-site crawl found that the large majority of llms.txt files received few or no requests from AI crawlers — the file is published faster than the engines read it. Publishing llms.txt today is a low-cost bet on a convention that isn't fully consumed yet, not a switch that flips your recommendations on.

This is why we're candid in the methodology: the fix pack generates an llms.txt for you because it's cheap, correct, and future-facing — but the action plan weights third-party surfaces (directories, review platforms, editorial lists, community threads) far more heavily, because those are what engines demonstrably cite now. An llms.txt makes your own house tidy. It does not make the rest of the web vouch for you, and vouching is what moves the shortlist.

Where llms.txt fits in the priority order

Effort	Task	Impact today
Low	Publish llms.txt (this guide)	Small now, hedges the future
Low	Schema markup on your homepage	Small–moderate; reduces grounding ambiguity
Medium	Answer-shaped pages for each core question	Moderate; matches fan-out queries
Medium–high	Get onto the directories engines cite	High; this is where citations come from

Do llms.txt first because it takes ten minutes — then spend your real hours on the bottom row.

How to check it's working

After publishing, confirm the file loads at yourfirm.com/llms.txt and returns as plain text. Then measure the thing that matters: re-run the audit a few weeks later and watch whether your citation count moves. The file is an input; your citation rate is the outcome, and it's the outcome you should hold yourself to.

Get a validated llms.txt built for your firm — inside the diagnostic.

The paid report generates a complete, QC-checked llms.txt and schema block from your real positioning, plus the third-party action plan that actually moves citations. Or run the free scan first to see if you need it.

Get my free scan Full diagnostic — $490

Next: Schema markup for financial firms → · The surfaces AI cites →