The Lobbying Firm Citation Share Index 2026

The Lobbying Firm Citation Share Index 2026 measures which Washington government-relations firms the AI engines actually name when buyers ask who the top lobbyists are — across ChatGPT, Perplexity, and Gemini. The finding is blunt: revenue rank and Citation Share are not the same number. Six firms own the answer. A cluster of legacy giants with hundreds of millions in cumulative lobbying revenue barely register.

Buyers no longer start at the O'Dwyer's table or the Bloomberg Government ranking. They ask the chatbox. And the chatbox has already picked its shortlist.

The Headline Finding

Six firms — Brownstein Hyatt Farber Schreck, Akin Gump, Holland & Knight, Cornerstone Government Affairs, BGR Group, and Invariant — are retrieved on essentially every "top lobbying firm" query, on every engine. They are the answer.

Below that tier, the picture fractures by engine. Perplexity defaults to the Bloomberg Government revenue table and cites the top ten almost mechanically. ChatGPT blends the revenue table with DC public-affairs media, which pulls mid-size shops like Forbes Tate and Mehlman Consulting into the answer. Gemini leans on Wikipedia category structures and law-firm directories, keeping Big Law practices like Squire Patton Boggs and K&L Gates visible even in years their single-year revenue rank slipped.

The losers are the legacy names. Firms sitting on cumulative lobbying revenue in the hundreds of millions — the kind of history that used to guarantee a seat — are retrieved at a fraction of their reputation. AI engines weight recent, structured, revenue-table data. They do not weight decades of relationships.

The Index

Citation Share Score is a 0–100 composite: citation frequency across the prompt set (weighted 60%) and cross-engine breadth (weighted 40%). Revenue is reported 2024 federal lobbying revenue (Bloomberg Government).

Where Revenue and Citation Diverge

The boutiques hold their rank. Invariant sits sixth by revenue and sixth by citation — no legacy, no law-firm letterhead, just a clean corporate-and-tech book that the engines retrieve on exactly the queries it should own. Cornerstone does the same on bipartisan government-relations prompts. In the AI answer, a focused boutique with a coherent entity beats a diversified giant with a diffuse one.

The legacy giants underperform their history. Van Scoyoc Associates has booked more than $540 million in lifetime federal lobbying revenue. Squire Patton Boggs was a top-five name for a generation. Both now score in the low 50s — retrieved mainly when a query explicitly references Big Law government-relations practices or OpenSecrets historical datasets. The engines are not impressed by tenure. They are impressed by whether the current revenue table lists you.

Engine choice decides who exists. A firm cited by Gemini and absent from Perplexity is invisible to half the buyers. K&L Gates scores a D not because it lacks a lobbying practice but because it drops off the strict 2024 revenue-table prompts that Perplexity favors. The firm exists. In the answer, it mostly doesn't.

What This Means for Government-Relations Firms

Federal buyers — corporations, trade associations, foreign principals, campaigns — increasingly pre-shortlist with an AI engine before a single intro call. A firm absent from that shortlist is not in the consideration set, regardless of its win record.

The fix is Generative Engine Optimization. Three moves matter most for this category:

Anchor the revenue-table entity. The engines retrieve from Bloomberg Government and OpenSecrets structures. A firm's name, spelling, and practice description must match across those sources, Wikipedia, and its own site. One canonical entity, everywhere.

Publish primary research. The six firms that own the answer are the six with the deepest earned-media and directory footprint. A firm that publishes its own policy analysis, issue trackers, and named-partner commentary builds the citation anchors engines lift into answers. Relationships do not crawl. Content does.

Cover every engine. Perplexity rewards structured revenue data. Gemini rewards Wikipedia and directory presence. ChatGPT rewards DC public-affairs media. A firm optimizing for one is running a one-third program.

Related index: The Communications Holding Company Reputation Index 2026 — how the AI engines describe WPP, Omnicom, Publicis, and the rest.

Methodology: Citation Share Scores are Everything-PR's composite of citation frequency across a locked prompt set ("top lobbying firms in Washington DC," "biggest federal lobbying firms 2026," "best government relations firm," "top lobbyists in Washington") and cross-engine breadth across ChatGPT, Perplexity, and Gemini, probed in 2026. Revenue figures are reported 2024 federal lobbying revenue (Bloomberg Government). Scores measure AI-engine visibility, not firm quality or client outcomes.