When search algorithms began relying more heavily on generative models, many large platforms experienced a sharp drop in organic traffic. Traditional SEO stopped being reliable, and snippets in the search results were increasingly replaced by fully generated AI answers. One major platform felt the impact particularly strongly: its usual stream of users shrank, engagement fell, and content that had previously ranked consistently no longer appeared in leading positions. It was at this point that the AIO approach – optimisation for AI-driven algorithms – was chosen.

Reframing content for new answer formats

The first step was understanding how AI generates responses. Generative models do not display websites directly – they reconstruct information, prioritising structure and semantic completeness. As a result, the platform’s content was rewritten to become as clear, precise and machine-friendly as possible.

All materials were organised into logical blocks: clear definitions, concise explanations, and self-contained semantic segments. Anything that failed to add meaning – empty introductions, promotional inserts, overly decorative language – was removed. This kind of content became a suitable source for generative systems, which prefer structured knowledge they can easily extract and reuse.

Restoring thematic coherence

After years of publishing, the platform had accumulated thousands of loosely connected articles. This meant AI models did not perceive it as a unified expert resource. As part of AIO, a restructuring audit was carried out: materials were grouped by theme, core areas of expertise were defined, and interconnected clusters were established.

As a result, the site began to resemble a coherent knowledge system rather than a collection of isolated posts. Generative models increasingly pulled information from these clusters, as they offered completeness and predictable semantic logic.

Adapting to AI-first search results

A decisive stage was transforming the content formats themselves. Short explanatory blocks were introduced – easily integrated into AI-generated answers – alongside expanded article versions suitable for deeper processing, and precise phrasing aligned with user intent. The content now addressed not only what users search for, but also how AI turns those queries into final responses.

Additional materials were created specifically for the types of questions users often pose to generative assistants. This allowed the platform to fit naturally into the new query “language” – part conversational, part analytical.

Strengthening authority and modality

One of the key criteria for AI is source trustworthiness. The platform therefore reinforced its authority: it added author profiles, clear expert attributions, and specialist-verified data. The articles became less anonymous and more structured around factual credibility.

Algorithms began classifying the site as a reliable source, which increased the likelihood of being cited in AI responses and boosted visibility in generative snippets.

Traffic recovery and renewed growth

Several months after implementing AIO, the platform recorded a gradual return of traffic. Mentions in AI answers increased first, followed by the return of users clicking through from extended result blocks. Legacy search rankings began recovering as well – but more importantly, a new channel emerged: a steady flow of traffic from generative recommendations.

Over time, the platform began receiving even more organic visibility than before the decline, as AI surfaced its content in a wider set of contexts than traditional SEO ever could.

This case demonstrated that optimisation for AI is not a passing trend, but a new standard for the information landscape. AIO not only restored lost traffic, but created a foundation for sustainable growth in a world where generative answers are becoming a central component of search.