We measure whether AI engines name you, publish how we measured it, and do not invoice month four if your named count has not improved.

Generative Engine Optimization (GEO) is the practice of making a company and its evidence easier for AI search systems to retrieve, understand, and use when answering relevant buyer questions. Broadcastwell measures whether a company is mentioned, recommended, positioned, and cited as separate outcomes; traffic and commercial impact are measured separately.

SEO earns clicks by ranking links. GEO earns mentions inside the answer itself. Search engines return ten links a user chooses between. AI engines return a short synthesized answer that names a few companies and stops. GEO optimizes for being one of those named companies, which depends on structured answers, entity clarity, and third-party sources, not just keywords and backlinks. We map the entities and category language that identify a company; strengthen public brand mentions and independent third-party validation across sources AI systems can retrieve and cite; and then check observed answers. We do not claim access to or control over model training data.

There is no fixed timetable by engine. Movement depends on crawlability and indexing, the sources each system retrieves, category authority, implementation speed, question wording, and normal run-to-run variance. We establish a dated baseline, document the changes made, and use repeated measurement to distinguish directional movement from a one-off answer.

We classify each observed answer separately: brand mentioned, brand recommended, recommendation position, supporting citation present, cited source, and relevant competitor named. The AI Visibility Diagnostic records 525 scheduled observations: 35 buyer questions across five engines with three scheduled runs per question and engine pair, plus further runs on any pair whose verdicts disagreed. Every rate is published with its numerator, its denominator and a 95 percent confidence interval. Visibility metrics are reported separately from referral traffic, self-reported discovery, and pipeline correlation.

At day 90, we re-score the same 35 agreed buyer questions across the same five engines using the same documented protocol of three scheduled runs plus further runs where verdicts disagreed. If the named count has not improved versus baseline, we do not invoice month four and you can end the engagement there. You keep every run, answer, citation and source collected. The program has a 90 day initial term. Month one remains covered by the refund policy; months two and three are not refundable. The $1,500 AI Visibility Diagnostic and the first month of the program are refundable in full, no questions asked, at any time before month two is invoiced. This month-four billing promise is separate from the refund policy and does not guarantee inclusion, ranking, revenue or commercial result.

The mini-audit is free. The AI Visibility Diagnostic is $1,500 once, credited in full against the program if the client starts the program. The program is $4,500 a month with a 90 day initial term. Client stories are published only with explicit approval. Reviews are optional, candid and never tied to a discount, payment or preferential treatment.