Ecrubox AI Prism

Most website owners have no idea whether their site is even readable to AI crawlers. You're running attribution that doesn't cleanly tag AI referrals. Your session recordings miss the bots. Your dashboards don't show what GPTBot sees when it reaches your product page. Adobe's data shows the gap between retailers whose websites AI can actually parse and cite, versus those it can't. 62% higher AI visit share on homepages for readable sites. The aggregate 393% growth is being held up by sites that AI can understand. The rest are dragged down. If your conversion data looks flat on AI traffic, don't assume the channel isn't ready. Audit whether your website is machine-readable first. Your buyers might already be there. You just can't see them yet. That invisibility isn't a timing problem. It's a measurement problem. And measurement is what we've always been good at, and that’s why we released the Prism. A unified intelligence suite that allows businesses to monitor, audit, optimise, and report on their presence across both traditional search engines and the new generation of AI-driven discovery channels. What follows is a brief look at some of the modules the platform offers and why it represents a genuinely important step forward for anyone serious about digital visibility in 2026.

The entry point for traditional search performance. Pulling directly from Google Search Console and GA4, it surfaces organic sessions, GSC clicks, average click through rate, and average ranking position across a selectable date range. Crucially it separates branded queries from generic ones, showing how each category performs in terms of total clicks, average CTR, and average position. The SEO Audit Dashboard takes the analysis a layer deeper. Described as an AI powered Google Search Console analysis tool, it runs two distinct audit types: a Content Audit and a Technical Audit. The output is a scored assessment with a clear score trend over time, a count of total opportunities identified, quick wins available, and an estimate of monthly traffic potential that could be unlocked. The audit categorises findings into CTR opportunities, cannibalisation issues, quick wins, low click pages, and content gaps, each with a specific count of affected queries or pages. Audit history is stored so that successive runs can be compared directly, giving teams a clear view of whether their interventions are having an effect.

Two dedicated tools sit within the platform and address specific technical requirements that are easy to overlook but consequential for both search and AI visibility.The Schema Builder generates SEO and LLM friendly JSON-LD structured data markup. A user provides a page URL, an optional schema type (with autocomplete suggestions for types such as LocalBusiness, Product, or FAQPage, or the option to leave it blank and let the platform infer the correct type from the page content), and an optional logo URL. The platform then analyses the page and produces ready to implement JSON-LD output. The explicit framing of this tool as serving both SEO and LLM optimisation reflects a genuine understanding of how structured data now functions: it helps search engines understand page content, and it helps AI models parse and accurately represent a brand when generating responses.The Open Graph and Twitter Card Generator is similarly practical. A user enters a page URL, the platform analyses what social meta tags are currently present, and it produces a full analysis report identifying missing tags alongside a list of specific improvements needed. It then generates the complete recommended head tag code, ready to copy and paste. In the example visible within the platform, the analysis correctly identified missing og:image, og:type, og:site_name, twitter:card, twitter:image, and twitter:site tags, and produced the corrected code for all of them. This kind of tool removes a common source of friction for teams who know social metadata matters but find the implementation fiddly to get right.

This is where Prism breaks genuinely new ground. The LLM suite is built around the recognition that AI models are now a primary channel through which consumers discover products, services, and information, and that this channel has been almost entirely unmeasured until very recently.

The LLM Tracking module is the foundation of the entire LLM suite. Users create tracking queries by defining a target phrase (the brand name or product they want to track), a natural language query prompt that a real consumer might type into an AI assistant, and the expected URL they would hope to see cited. The platform then runs that query simultaneously against OpenAI (ChatGPT), Anthropic (Claude), and Google (Gemini), and returns results for each showing whether the target phrase was found, the prominence of the mention (high or low), and whether the expected URL was cited. Users can view the full response from each model. This is the core feedback mechanism of the platform: it tells teams not just whether AI models know about their brand, but exactly how and whether those models are recommending it in response to the kinds of questions real customers are asking.

The LLM Dashboard translates the tracking data into commercial metrics. It shows LLM sourced sessions, conversions, revenue, and average session duration over a selected date range, broken down by individual AI platform. In the live data visible in the platform, traffic is broken down across ChatGPT, Gemini, Claude, and Perplexity, showing sessions, items viewed, add to basket events, checkouts, purchases, revenue, engagement rate, bounce rate, and average session duration for each source. An ecommerce funnel visualisation shows how LLM sourced traffic progresses from item view through to purchase, with conversion rates and drop-off points at each stage. This is a genuinely novel capability: the ability to see the commercial value of AI platform traffic with the same granularity that teams have long applied to paid and organic search.

LLM Visibility provides an aggregated view of brand presence across all LLM tracking results, automatically detecting which competitors appear alongside the tracked brand in AI responses and calculating a visibility percentage for each. The output is a ranked table of brands by LLM visibility share, with trend data showing how each brand's share is moving over time. This competitive benchmarking dimension is important because share of voice in AI responses is a zero sum game: a brand that is underrepresented in AI recommendations is, by definition, losing ground to the brands that are overrepresented.

The LLM Audit Dashboard provides the most operationally detailed view in the entire suite. It scores the brand's overall LLM readiness on a rolling basis, tracks that score over time, and maintains an issue tracker categorising open problems by severity (Critical, High, Medium, and Low) and by type (Structured Data, Content Quality, and others). Each issue comes with a specific description, the affected URLs, and a direct link to the details and fix guidance. In live usage, the audit identified issues ranging from insufficient brand authority information and incomplete schema on collection pages through to thin content on key landing pages and missing material definitions in structured data. The audit history allows previous runs to be compared, so teams can see directly whether resolved issues have improved the score.

Public Perception completes the LLM section but addresses a distinct question: how is the brand being discussed across online forums, news sources, and reviews, not just within AI model responses. Users enter a brand name and a geographic scope, and the platform generates a sentiment score alongside a comparison against competitor brands. The output shows recent analyses with scores and positive or negative sentiment labels, allowing teams to track whether public perception is improving or deteriorating over time and how they compare to the competitive set. This is broader than LLM performance: it is monitoring the raw material from which AI models form their understanding of a brand.

The analysis identifies which channels are dragging overall funnel efficiency, where the critical bottlenecks sit, and what the likely causes are. Previous analyses are stored so that improvement over time can be tracked.This module is particularly valuable because it contextualises AI driven traffic within the broader acquisition picture. A team might discover that LLM sourced sessions convert significantly better than paid social traffic, or that a particular AI platform sends users who abandon at checkout at a disproportionate rate. Those insights shape where budget and effort should be directed in ways that neither SEO data nor LLM tracking data alone could reveal.

The platform is relevant to a broad range of organisations, but it will be of particular value to digital marketing agencies managing multiple clients across competitive categories, in house SEO and content teams at brands of meaningful scale, and any business operating in a sector where AI search has already begun to influence consumer discovery significantly.

For agencies, the combination of multi project management, comprehensive Google Search Console integration, LLM tracking across multiple AI platforms, and structured reporting makes it possible to deliver a genuinely differentiated service. Rather than reporting on the same traditional SEO metrics that every other agency is measuring, Prism enables agencies to demonstrate value across the new and strategically important dimension of AI visibility, with the commercial data to back it up.

For in house teams, the platform addresses a gap that has been widening for the past two years. Most teams have tools for measuring traditional search performance and most have some awareness that LLM visibility matters. Very few have had any systematic way to measure it, track it over time, understand how their specific content and technical choices are affecting it, or connect it to commercial outcomes. Prism fills that gap comprehensively.

For any business or agency that takes digital visibility seriously, Prism deserves close attention. The question of where brands appear when AI systems respond to consumer queries is not going away. If anything it will become more consequential over the next few years as AI assistants become further embedded in how people research and make purchases. Starting to measure it, understand it, and actively optimise for it now, with the right tools in place, is not a nice to have. It is rapidly becoming a core competency.