What goes in a content strategy for AI search visibility?
Broadcastwell measures and fixes whether AI search engines name B2B software companies.
Run the free checkA content strategy for AI search visibility starts from a frozen set of buyer questions and a dated baseline, ranks the questions where you are absent, assigns each fix an owner and a status, and states how a later run will be compared. Our free template has 18 fields, in Markdown and CSV, and below we fill it on the fictional Kalvenor sample (Sample data, fictional company).
Updated 11 November 2026.
A strategy is a plan for one category over a stated period. It sits above the briefs: each priority it names becomes one content brief. It names no target rank, because no one can make ChatGPT, Claude, Perplexity, Google AI Overviews or Google AI Mode name a company.
Download the template
| File | Format |
|---|---|
| Content strategy template | Markdown |
| Content strategy template | CSV |
| Filled fictional Kalvenor example | Markdown |
| Filled fictional Kalvenor example | CSV |
Four parts of the template
The frame. Period, category in the buyer's words, the question set and the engines. Our method uses 35 brand-free questions for a full baseline: 10 shortlist, 10 buyer-role, 10 use-case and 5 evaluation or pricing. Freeze the wording before the first capture, because changing it starts a new baseline. Our guide to choosing buyer questions helps draft them.
The baseline. Capture dates, runs, valid answers and exclusions, then your named count, the questions where you appear at all, the rival named most, how often your own pages are cited, and the outside domains cited most. Each number carries its base, and from 30 observations up, its 95 percent interval.
The priorities. Up to three at a time, each naming the questions it serves, the action, one owner and a dated status. An action can be a page on your site or a change to an outside page the engines cite.
The comparison. Which runs are compared with the baseline, the rule for calling a change, the latest separated change, any provider releases between captures from our model release tracker, the limits of the record and who approves publication.
A filled example: fictional Kalvenor sample
Fictional sample. Every figure below comes from our published fictional Kalvenor sample: 35 buyer questions, observed 8 to 12 July 2026, with a monthly run on 12 August 2026. Kalvenor Systems is not a client, and no figure describes a real company.
| Field | Value (fictional sample) |
|---|---|
| Category | field service management software |
| Baseline | Observed 8 to 12 July 2026; 3 scheduled runs; 518 scored answers; 7 excluded |
| Named today | 125 of 518 answers, 24.1%, 95% interval 20.6 to 28.0 |
| Questions with an appearance | 21 of 35 |
| Most named rival | Northvale FSM, 329 of 518 answers, 63.5%, 95% interval 59.3 to 67.5 |
| Own pages cited | 66 of 518 answers, 12.7%, 95% interval 10.1 to 15.9 |
| Sources cited most | hvacoperations.example 105 of 518; servicetechforum.example 99 of 518; veranofsm.example 97 of 518 |
| Priority 1 | Q18, the procurement checklist question: a public page with contract terms and data export detail. Content lead (sample). Approved, since 16 July 2026 |
| Priority 2 | Q2, Q6 and Q7: recurring maintenance contract detail on the dispatch page, submitted to the buyers guide the engines cited. Product marketing (sample). Measured again, since 12 August 2026 |
| Priority 3 | Q12, job level margin reporting: a case note pitched to the operations publication the engines cited. PR lead (sample). Sent, since 22 July 2026 |
| Latest change | ChatGPT: 0 of 10 answers in the monthly run of 12 August 2026, against 15 of 29 in the baseline run of 12 July 2026; the two 95% intervals do not overlap |
The latest change in this fictional sample is a fall, and the template records it the same way it would record a rise. A strategy that only reports good news cannot be checked.
Where the strategy fits
A strategy like this underlies the monthly work in Care, $490 a month, and Program, $5,500 a month. The workstreams that carry it out are listed in our service model, and the finished record of a strategy's re-check follows our completed-case format.
Sources
- Broadcastwell, fictional Kalvenor sample, 35 questions, observed 8 to 12 July 2026, monthly run 12 August 2026 (read 9 November 2026).
- Broadcastwell, Methodology, method v1.1, on question groups, frozen wording and intervals (read 9 November 2026).
Related: the workshops page walks through filling this template in a self-paced session.