Problem
Viggle was not a small site. By September 2025 it ranked for more than 36,000 keywords and pulled over 400,000 organic sessions a month. On paper it looked like a site that had already won its category.
Underneath, the picture was different. Ninety-five percent of those ranking keywords sat beyond page two, most of them on URLs that existed for no commercial reason at all. Crawl budget was being spent on pages nobody would ever land on, while the pages that actually mattered, product and use-case pages, were thin, inconsistent and in several cases missing entirely.
The larger gap was newer. Viggle's audience, creators comparing AI video tools before they commit, had already moved a chunk of that research into ChatGPT, Perplexity and Google's AI Overviews. Across all seven major answer engines in September 2025, Viggle was cited a total of six times. Four pages on the entire domain were quotable. The category conversation was happening in a place the brand did not exist.
Baseline
We fixed the measurement before touching anything. Index composition, ranking distribution, backlink profile and citations across seven AI answer engines were captured as a single September 2025 snapshot, so every later claim could be checked against it. This is what we were working against.
Engines tracked: ChatGPT, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Copilot and Grok. Only three of the seven cited the domain at all, and none of them more than three times.
Strategy
The obvious move was to publish more. We did the opposite. Viggle already produced plenty of content; the constraint was never volume, it was whether anything on the domain was structured well enough to be found, trusted and lifted into an answer. Three principles carried the engagement.
No content sprint, no new blog category, no rush to publish against competitors. Viggle's problem was never a shortage of pages. Adding more before clearing the index would have deepened the exact issue we were brought in to fix, and it would have made the first three months look busy while moving nothing that mattered.
Execution
SearchAxe ran the engagement end to end, from the audit through to the pages going live. Every change was scored and sequenced so the work shipped in priority order rather than landing as one unmanageable list.
-
Ran a full technical audit and turned it into a scored, sequenced remediation roadmap with traceable keep, consolidate and redirect decisions.
-
Removed non-commercial URLs from crawl and index, then rebuilt the sitemap structure around priority templates.
-
Reallocated crawl budget toward product, use-case and comparison pages.
-
Built and shipped more than ten landing pages across product and use-case intent.
-
Restructured on-page hierarchy for extraction: question-led headings, direct answers above the fold, scannable specifications.
-
Rewrote the blog for dual optimisation, built to rank in Google and to be lifted by LLMs.
-
Identified and disavowed spammy backlinks before they eroded domain trust.
-
Built brand mentions and contextual links through Reddit and topically relevant placements.
-
Added and validated schema markup on product and use-case templates so specifications could be read without interpretation.
Results
The index got smaller and the visibility got larger. Viggle now ranks with more than ninety percent fewer URLs than in September 2025, and earns more sessions, more top-three positions and vastly more AI citations from that smaller footprint.
Lessons for other AI products
Scale is not visibility
Thirty-six thousand ranking URLs and four quotable pages. The size of an index tells you nothing about whether an engine can find something on it worth citing. Audit what is citable, not what is indexed.
Crawl budget is an AI problem now
Those 21,427 dead URLs were not just SEO clutter. They were consuming the crawls that should have been spent on the pages built to be quoted. Clearing them was the single highest-leverage action in the engagement.
Format decides whether you get quoted
The same claims, restructured into question-led headings and direct answers, went from unciteable to quoted across all seven engines. The content did not get better. It got machine-readable, and that turned out to be the difference.
Community mentions carry unusual weight
Links built for rankings and citations earned in research spaces are not the same asset. Reddit and topical placements moved AI visibility noticeably faster than they moved classical rankings.
Removing pages can be the growth lever
Cutting more than ninety percent of ranking URLs raised sessions rather than costing them. When an index is mostly non-commercial pages, subtraction compounds faster than publishing does.
Sequence the work, not just the list
Every change was scored and ordered before anything shipped, so the highest-impact work went live first. An unsequenced audit is a document; a sequenced one is a roadmap that actually gets executed.
Ready to turn search into a growth channel?
Let's build a stronger organic growth engine.