How much does political opposition research cost?

The short answer

Political opposition research in 2026 runs from about $1,000 for an AI-assisted report on a single subject to $15,500 to $37,000+ for a traditional research book from a human research firm, with AI-and-human hybrid books published in between at $7,500 and up. Doing it yourself costs staff time instead of dollars, which for most campaigns is the most expensive currency they have. The price differences are not arbitrary: each tier buys a different depth, speed, and level of human verification.

And the price gap points at a strategy conclusion this article will defend: at every level of race, the right first move is the AI-assisted report. For about $1,000 it may surface everything the race needs. When it does not, it becomes the blueprint that makes every expensive human research hour that follows more effective. For local and downballot races it is simply the best answer available; for a statewide or national race it is the map you hand the research firm before the meter starts at $225 an hour.

What actually drives the price

• The size of the paper trail. A first-time candidate with no voting record is a smaller job than a ten-year officeholder with thousands of votes, filings, and public statements. Traditional firms price this directly: Nesbitt’s published packages step up from $22,450 for a judge or prosecutor to $29,200 for officials with ten-plus years of record.
• Turnaround. Speed is a paid feature everywhere. Nesbitt publishes a 25% surcharge to compress a book from 12 to 9 business days ($19,375) and 50% for 6 days ($23,250). AI-assisted products compete on exactly this axis: hours or days as the default rather than the rush option.
• Human verification depth. The core of a traditional book’s price is researcher hours: Nesbitt publishes $225 per hour with a ten-hour minimum for consulting work. AI products organize what is publicly available fast and leave verification to your team, which is the honest reason they cost a tenth as much.
• Scope: one subject or the field. Bulk pricing is standard. FullPAC publishes bulk discounts for more than one report, and consultants typically run research across every race they staff rather than one at a time.

Tier 1: Do it yourself ($0 plus time)

Everything opposition research is built on is public: court records, business registrations, property filings, campaign finance reports, voting records, news archives, and what candidates have said on the public web. A staffer with time and discipline can assemble a serviceable file on a local opponent.

The cost is that “time and discipline” line. Weeks of a senior staffer’s cycle is real money on any campaign budget, the work arrives late or shallow when the news cycle will not wait, and an amateur file misses the archived, dormant, and cross-referenced material that professional processes exist to find. DIY makes the most sense for very small races and for a first pass on your own candidate.

Tier 2: AI-assisted reports (from about $1,000)

This is the newest tier and the reason the price floor dropped an order of magnitude. FullPAC’s GOTV Intel publishes it plainly: a single report typically starts around $1,000, with bulk discounts for more than one, nothing charged at the request step, and scope confirmed before any work begins. AI agents research structured public records and lawful open sources across six report types, from background and campaign finance to digital footprint and reverse username search, and most preliminary reports are available within hours. Consultants deliver the reports white-label under their own brand.

This is the tier every campaign should start in, and the arithmetic is why. At roughly $1,000 against a $15,500 book, the AI report is a rounding error that answers the most expensive question in research: what is actually out there? Often the answer is everything the race needs, organized and sourced, the same week you asked. When the report surfaces a thread that deserves a professional pull, you now brief the human researchers with a map instead of a blank page, and every one of their hours lands on something. Paying for weeks of human hours to discover what an AI pass would have surfaced in an afternoon is the expensive way to learn what is public.

What this tier is not: a substitute for human verification. Every finding should be checked against the cited record before it shapes a mail piece or an ad, and our own product page says exactly that.

Tier 3: AI-plus-human research books (from $7,500)

Civly, an AI-forward research shop, publishes research books “starting at $7,500” with delivery “in days instead of weeks” (civly.ai/politics, read September 14, 2026), and a $999 down-ballot bundle combining lighter research tools for small races. That price point buys more human structure and narrative work than a raw AI report while staying well under traditional-firm books, and their own marketing frames the comparison the same way this article does: the work “costs $30,000+ at legacy firms.”

Tier 4: Traditional research firms ($15,500 to $37,000+)

Nesbitt Research is one of the few traditional firms that publishes a rate card (nesbittresearch.com/candidate-campaigns, read September 14, 2026): a core research book at $15,500 on a roughly 12-business-day standard turnaround, pre-built packages from $22,450 to $29,200 depending on the subject’s office and record, a full workup at $37,150, rush surcharges of 25% to 50%, and hourly consulting at $225 with a ten-hour minimum.

That money buys the thing the cheaper tiers do not claim to sell: experienced human researchers building a vetted, sourced, litigation-conscious book, the standard when an attack has to survive a lawyer’s review and a hostile fact-check. Statewide and competitive federal races usually budget for this tier, often alongside a self-research book on their own candidate, and it is the right call when the stakes justify it. If your race needs that bar, pay for that bar. Just run the $1,000 AI pass first, so the firm starts from a map instead of a blank page.

The AI-first playbook, by race level

Whatever the race, the sequence is the same: run the roughly $1,000 AI-assisted report first, then decide whether anything it found justifies deeper spend. What changes by level is only how often the second step happens.

• School board, city council, small local: the AI report usually is the research program, start to finish. It costs less than a week of a staffer’s DIY time, covers more sources, and spending five figures on research for a $30,000 race is a budget error. This is the level where AI-assisted research is simply the best option, full stop.
• Competitive local and legislative races, consultant books: the AI report as the default line item on every race in the book, escalating to a deeper book only on the races that tighten. Consultants attach it precisely because it is cheap enough to run everywhere and decisive about where the deeper spend belongs.
• Statewide, federal, high-spend independent expenditure: these races will buy the traditional book, and should. The AI report is the blueprint that comes first: it scopes what exists, briefs the firm on day one instead of week two, and points the $225-an-hour work at threads instead of at a cold start. Skipping the $1,000 pass before a $37,000 workup is leaving the cheapest leverage in the budget on the table.

The questions people actually ask

How much does opposition research cost for a local race? With published 2026 pricing, as little as about $1,000 for an AI-assisted report on a single subject, or $999 for Civly’s published down-ballot bundle. The five-figure books exist for races whose stakes justify them.

How long does opposition research take? Published turnarounds range from hours (AI-assisted preliminary reports) to days (AI-plus-human books) to about 12 business days standard at a traditional firm, where faster delivery carries published surcharges of 25% to 50%.

Is AI opposition research as good as a human research firm? Different product, same starting move. AI-assisted reports organize publicly available records fast and cheaply; a traditional book adds human verification, narrative judgment, and a standard built to survive legal review. But the order is not a coin flip: run the AI report first, because it may cover everything the race needs, and when it does not, it makes the human work faster and better aimed. AI for speed and coverage, humans where a claim will carry an attack.

Should we research our own candidate? Yes, and every tier offers it. Knowing what is publicly findable about your own candidate before an opponent finds it is the cheapest insurance in politics.

The playbook in one line: start every race with the AI report, escalate only where it earns it. GOTV Intel is where that first step lives, with the report types, the request flow, and the published starting price on the page, and fifteen minutes will scope a specific subject. Consultants comparing the wider platform should start with the consultants’ platform guide.