From 6 Cited Pages to 31, in 60 Days
A client that engineers silicon-carbide solar inverters and EV chargers for Indian grid and climate conditions came to us with a familiar problem: strong products, established distribution, and almost no presence in the answers AI engines were giving buyers who asked about the category.
In 60 days, a technical foundation rebuild and a disciplined weekly content program took the brand from 6 to 31 pages cited by AI engines, a 5.2x increase, without naming a single competitor.
Unique pages cited
Up from 6 at baseline, a 5.2× increase in AI source coverage.
Total citation instances
Up from 16 at baseline, a 4.2× increase across every tracked engine.
Combined share of voice
Up from 15.8% at baseline, holding steady while citation footprint grew fast.
Why Wasn't the Brand Showing Up in AI Answers?
Because its own pages were rarely what an engine actually pointed to. At baseline, only 6 pages across every tracked engine had ever been cited as a source, even though the content already existed to answer the questions buyers were asking.
The client engineers silicon-carbide solar inverters and EV chargers built for Indian grid and climate conditions, a category where AI answer engines already had well-worn default answers built around a handful of larger, longer-established brands.
Engines simply weren't finding a reason to cite the brand's own pages over those defaults. Closing that gap, not writing more content in general, was the actual brief.
Midway through the sprint, the client set one non-negotiable rule: no content, anywhere, would name a rival by name. Not in a blog post, not in a LinkedIn carousel, not in a directory listing. Every workstream below operated inside that rule.
What Changed, and on What Cadence?
Five workstreams ran in parallel for sixty straight days, each on its own fixed cadence rather than a one-time push: content, discovery, reviews, community, and the technical layer underneath all of it.
| Workstream | Cadence | What it covered |
|---|---|---|
| Technical | Ongoing rebuild | Structured data per page type, the machine-readable brand file, authorship and trust signals |
| Content | ~8 posts / month | Hyper-contextual articles built around the exact question a buyer was typing in, not a keyword |
| Discovery | 2 to 3 placements / month | Directories, reference sources, Quora Q&A, paid and earned press |
| Reviews | Ongoing | Verified installer and buyer reviews, public and checkable |
| Community | ~4 posts / month | LinkedIn articles, carousels, and brand updates |
None of these three, regularity, precision, restraint, are individually remarkable. Repeated without exception, week after week, for sixty straight days, they're what actually moved the numbers.
What Actually Moved in Sixty Days?
Citation footprint grew fastest: 5.2x more pages cited and 4.2x more total citations. Combined mention share held roughly flat, which is normal early in an AEO program: citation footprint moves first, and mention rate follows as engines re-crawl and re-index.
| Metric | Before | After | Change |
|---|---|---|---|
| Unique pages cited as an AI source | 6 | 31 | 5.2× |
| Total citation instances | 16 | 67 | 4.2× |
| Combined share of voice | 15.8% | 16.2% | +0.4pp |
What Actually Made This Work?
Two habits, repeated without exception for sixty days: a cadence that never broke, and writing precise enough to be the correct answer to a question, not just an available one.
The cadence never broke. Blogs and LinkedIn both went out weekly, on schedule. Directory and press outreach ran in parallel the whole way through.
Every piece was built around one real question, and was checked against the product itself before it shipped.
Just as much judgment went into what didn't happen. An outdated founder bio was caught and corrected before it reached the site's machine-readable footer.
None of these three are individually remarkable. Repeated without exception, week after week, for sixty straight days, they're what actually moved the numbers above.
Frequently Asked Questions
Case study anonymized at client request. Metrics drawn from weekly AI-visibility tracking across ChatGPT, Gemini, Perplexity, and Google AI Mode, comparing the engagement's baseline week to its week-60 checkpoint.