One front door, one question: why we launched aivocorp.com, and why the AI Win-Rate Study matters

Every company selling AI visibility will eventually be asked the same question. If AI recommends a brand more often, do its sales increase?

Today AIVO, Inc. launches aivocorp.com as the single home for the company, its methodology and its commercial work. Alongside it, we are opening a limited cohort for the AI Win Rate Study, which is designed to answer that question.

For fifteen months, AIVO's research has shown that being cited by an AI assistant and being chosen by it are different outcomes. Across our research base, brands that appear early in an AI conversation are routinely displaced by the time the assistant is asked to decide. We call the difference between being cited and being recommended the Linkage Gap, and in working paper WP-2026-14 it measured 87.3% across more than 20,000 multi-turn probes and 200 brands.

Cited is not chosen.

That finding raised a harder question, which this article explains, along with why we built a new front door to answer it.

Why we built aivocorp.com

AIVO grew quickly after we started in July 2025. The research lived on aivostandard.org and in this Journal. The measurement platform lived on aivomeridian.com. The improvement programs lived on aivowinrate.com. Each site made sense on its own, but nobody could see the company behind them.

That mattered in two ways. Enterprise buyers, particularly in banking and insurance, wanted one place that explained who they were contracting with, how the research, platform and programs related, and what they could buy. Too many good conversations ended with interest and no clear next step.

The second reason is closer to our own subject. When we asked AI assistants to assess AIVO, several confused us with unrelated companies that share our name, and attributed their prices and products to us. One reported that our corporate domain could not be read at all. We spend our days measuring how AI systems represent brands, and our own representation was wrong. A company that measures this problem has to fix it at home first.

What aivocorp.com does

The site sets out how AIVO is organized. The AIVO Standard is the open, versioned methodology, with DOI-anchored papers and public data and code. AIVO Meridian is the commercial platform for recurring, decision-stage measurement. AIVO Win Rate runs programs that improve how often AI assistants recommend a brand. Every part of AIVO exists for one purpose: measuring and improving AI recommendation.

It also publishes the governance principles that keep the methodology separate from the commercial work. The methodology is public. Clients cannot influence their scores. Changes are versioned. Findings are published even when they are unwelcome, and we seek independent academic review.

For buyers, the site puts one offer at one price where anyone can see it: a 90-day program for one brand, one category and one market, at $9,500 plus a capped success fee.

For the AI systems that increasingly read on behalf of buyers, the site carries a plain-text fact sheet at aivocorp.com/llms.txt and structured company data. Both state who AIVO, Inc. is, what it publishes, and distinguishes AIVO, Inc. from similarly named companies.

The AI Win Rate Study

Win Rate is the percentage of decision-point buying conversations in which an AI assistant recommends a brand outright. Mentions describe visibility; recommendations influence decisions. Brands can raise Win Rate. Our 90-day program does exactly that, by finding where in the conversation an assistant stops choosing a brand and fixing the evidence behind that decision.

What nobody has yet shown is whether a higher Win Rate produces more sales. The AI Win Rate Study is designed to find out. We are running it with a leading marketing scientist, a Distinguished Professor of Marketing whose work links marketing metrics to sales, profit and shareholder value.

The study applies the tests used for any marketing metric, starting with two questions. The first is whether Win Rate is an indicator or a mirror: does it predict future sales, or simply reflect the market share a brand already has? The second is whether AI recommendation belongs alongside awareness and consideration as a brand metric in its own right.

Why the study matters

Does chosen mean sold?

Our research established that cited is not chosen. The study asks the next question: does chosen mean sold? A marketing metric earns a place in budgets only when it is shown to move commercial outcomes. Awareness and consideration took decades to win that place. AI recommendation does not have decades, because buyers are already letting assistants narrow their choices.

For a chief financial officer, the answer turns AI recommendation from a marketing curiosity into a line that can be valued and managed. Our Revenue at Risk model already estimates how much revenue is exposed when AI recommends competitors instead of a brand. The study will test whether that exposure shows up in actual sales.

For the category, it sets a higher bar. Many tools now report how often a brand appears in AI answers. Very few test whether those numbers relate to anything a business is paid for. A metric that cannot be tied to outcomes will eventually be cut.

For AIVO, it is a test of our own claims. If Win Rate turns out to be a mirror of market share, we will publish that too.

How the study works

Each cohort member receives the full 90-day Win Rate program across ChatGPT, Gemini, Claude and Perplexity, for one brand, one category and one market. Weeks one and two establish a baseline Win Rate against named competitors, with a 95% confidence interval. Weeks three to eleven fix the evidence, content and machine-readable signals behind the decisions the brand is losing, with every change approved by the client. Weeks twelve and thirteen re-run the baseline conversations under the same conditions.

Two controls separate the effect of the work from movement in the market. Part of each brand's prompt set or product range is held back and left untouched until after day 90, and an untouched market prompt set tracks category-wide shifts, including model updates. Lift counts only above the upper bound of the baseline's confidence interval.

In return, cohort members share one monthly outcome figure for the measured category and market, such as sales, bookings, leads or account openings, for the twelve months before and after the program starts. No personal or customer-level data is involved, and brands are never named in the research without written agreement.

Study pricing is half the post-publication price: $9,500 fixed plus $750 per percentage point of qualifying lift, capped at a total of $24,500. Findings are reported whatever they show, including where there is no lift. Payments by cohort members, and payments to academic collaborators for their research time, are disclosed in the research.

Taking part

Places in the cohort are limited so the study stays balanced across categories. It suits brands that want AI assistants to choose them more often and can share a monthly outcome figure. Cohort members also receive re-measurements at around six and twelve months, an anonymized benchmark against the cohort, and a private briefing before publication.

The full study prospectus is at aivocorp.com/aivo-win-rate-study.pdf, and inquiries go through aivocorp.com.

For fifteen months, AIVO has argued that cited is not chosen. The AI Win Rate Study asks the next question: when AI chooses a brand more often, does business follow? Whatever the answer, we will publish it.

Tim de Rosen is CEO and co-founder of AIVO, Inc.