FAQs

AI platforms like ChatGPT, Perplexity, and Claude answer product questions by fetching web pages in real time, extracting text, and ranking passages by semantic relevance to the user's query — no search index, no click required. For outdoor and cycling brands, where high-consideration purchases often begin with a question like "what's the best trail brake for heavy riders" or "most packable sleeping bag for backpacking", the buying decision is increasingly being shaped before a customer ever visits your site. If your content isn't retrieved and ranked in the top results for those queries, your brand doesn't exist in that answer — regardless of how strong your SEO or ad spend is.

Google Analytics and Shopify analytics measure what happens after someone arrives on your site — sessions, conversion rate, and revenue by channel. SEMrush measures your position in Google's traditional index. REACH measures what happens before any of that: whether your brand's content is being retrieved and ranked by AI discovery systems when a potential customer asks a relevant question. You can have flawless conversion data and strong Google rankings and still be completely invisible to the AI layer that is increasingly driving the first moment of brand consideration. REACH is the only diagnostic that shows you your performance inside the actual retrieval pipeline, stage by stage.

Traditional SEO optimizes for Google's index-based ranking signals — backlinks, domain authority, keyword placement. AI retrieval systems don't use an index. They perform direct HTTP fetches of web pages and rank content using dense vector embeddings and cross-encoder models that measure semantic similarity to a query. A page can rank number one in Google and score poorly in AI retrieval if its content isn't semantically rich, structurally clean, or fully extractable by an AI crawler. The skills and signals that win in traditional SEO are necessary but not sufficient for AI discovery — and the gap between the two is where most brands are currently losing ground.

AI platforms that perform direct web retrieval fetch pages via plain HTTP requests, extract visible text content, convert it into vector embeddings, and rank passages by semantic similarity to the user's query. The brand whose content is most completely extracted, most semantically aligned with the query, and best structured for relevance scoring wins the recommendation. Crucially, this process has no awareness of brand awareness, ad spend, or historical search authority. A smaller brand with semantically rich, well-structured product content can outrank a category leader whose pages extract poorly or whose descriptions are thin on specific, query-relevant detail.

REACH runs a multi-stage diagnostic pipeline against queries specific to your brand and category. It measures whether your pages are successfully fetched and extracted by AI crawlers, how your content scores in dense vector retrieval, how you rank after fusion of multiple retrieval signals, and your final position after cross-encoder reranking — the most precise relevance signal in the pipeline. Outputs include a stage-by-stage Excel report showing your brand's performance and competitor performance at each layer, a competitive attribution summary tracking where your content enters and exits the pipeline, and an interactive 3D visualization plotting every result across all three retrieval dimensions simultaneously.

Yes — competitive visibility is a core output of every REACH diagnostic. Because the pipeline evaluates your brand and competitor content against the same queries simultaneously, you can see exactly where competitors are outperforming you: whether it's at the extraction stage (their pages are more AI-crawlable), the vector similarity stage (their content is semantically richer for your target queries), or the reranking stage (their passages are better structured for relevance scoring). This turns a vague competitive gap — "they keep showing up and we don't" — into a specific, layer-identified finding with a clear optimization path.

REACH surfaces findings at three levels of effort. Configuration fixes — adding FAQ Page or Product JSON-LD schema, correcting content extraction failures, restructuring page sections for AI crawlability — are typically developer tasks measurable in hours. Content improvements — adding semantic depth to product descriptions, building query-specific landing pages, restructuring FAQ content — take days but often produce the highest retrieval impact. Strategic gaps — thin category coverage, missing content formats, structural site issues — require longer roadmap work. REACH prioritizes findings by retrieval impact so effort is directed where it actually moves your position in the pipeline, not where it's easiest to act.

AI platforms are handling a rapidly growing share of high-intent product research queries, particularly for considered purchases in outdoor, cycling, fitness, and specialty retail — categories where buyers ask detailed questions before committing. Brands that establish strong AI retrieval presence now are building a structural advantage: the content quality, schema implementation, and semantic depth that drives AI retrieval performance also reinforces traditional SEO. The compounding effect runs in both directions — brands that act early accumulate advantages, and brands that wait cede ground to competitors who are already optimizing. The query universe is being shaped now, and early positioning is difficult to displace once it is established.

Yes. REACH evaluates the AI retrievability of your owned brand presence — your DTC site, product pages, and content assets. For brands with significant wholesale or marketplace revenue, strong brand content determines whether AI platforms recommend your specific products or simply recommend the category, leaving the customer to discover whoever ranks highest on Amazon or at the retailer. Thin or poorly structured brand content means AI systems default to generic category answers, eroding brand preference before the purchase decision is made. REACH identifies whether your owned content is doing the brand-building work that drives demand through every channel you sell in.

REACH onboarding starts with a query universe session — defining the specific questions your target customers are asking AI platforms at each stage of the purchase journey. The diagnostic pipeline then runs those queries, fetches competitor and brand pages, extracts content, scores dense vector retrieval, fuses ranking signals, and delivers cross-encoder reranked results, typically within 48 hours. The initial diagnostic establishes your baseline position across all pipeline stages and identifies the highest-impact optimization opportunities. From there, REACH operates as an ongoing measurement layer — tracking how your retrieval performance shifts as content and structural changes are implemented, and monitoring competitive movement in your query universe over time.

No — and this is one of the most important structural differences between AI-mediated discovery and traditional search. Google's PageRank and paid auction systems rewarded brands with budget and domain authority, creating a visibility landscape where spending determined presence. AI retrieval systems score content using embedding models and cross-encoders that evaluate semantic relevance directly — there is no ad auction, no authority signal for sale, and no black box algorithm that favors incumbents. A brand with a $500K Google Ads budget and a brand with zero paid spend are evaluated identically by the retrieval pipeline. What wins is content quality, semantic depth, and structural extractability — all of which are fully within a brand's control and require investment in craft, not in media spend.

Yes — and the transparency is one of the most significant advantages AI retrieval offers brands willing to invest in understanding it. Traditional PageRank was a proprietary algorithm with hundreds of undisclosed signals, making optimization a combination of best practices and guesswork. The core mechanisms of AI retrieval — dense vector embeddings, reciprocal rank fusion, cross-encoder reranking — are published, peer-reviewed techniques with well-understood behavior. REACH exposes every stage of the pipeline with actual scores: your dense vector similarity score, your fused rank, your cross-encoder relevance score. You can see exactly why a passage ranks where it ranks, identify the specific content or structural issue causing underperformance, and measure the direct impact of changes. This is optimization with the lights on.

AI retrieval results move faster than traditional SEO — and the reason is structural. There is no index crawl delay, no domain authority accumulation period, no waiting for Google to reassess your page. When an AI platform fetches your page, it evaluates the content it finds at that moment. Configuration fixes like adding FAQPage JSON-LD or correcting extraction failures can produce measurable retrieval improvements within days of deployment, as soon as AI crawlers re-fetch the updated page. Content improvements — richer product descriptions, query-aligned FAQ content, improved semantic structure — typically show movement within one to two weeks. The compounding dynamic works in your favor here: unlike paid search, where visibility stops the moment spend stops, retrieval improvements based on content quality are durable. You are building an asset, not renting a position.

REACH is a measurement platform first — which means every recommendation comes with a defined retrieval metric to move, and every subsequent diagnostic run produces a direct before-and-after comparison. When REACH identifies that your product page is failing extraction, the fix has a measurable outcome: extraction success rate and chunk count for that URL. When it identifies that your content scores poorly in dense vector retrieval for a specific query, the improvement has a measurable outcome: your dense score and rank position for that query. You are never in the position of implementing changes and hoping the algorithm noticed. Each REACH diagnostic run produces the same pipeline output — stage-by-stage scores, competitive attribution, and ranking positions — so progress is quantified, not estimated, and competitive movement is visible in real time alongside your own.

Yes — and this is one of the most underappreciated benefits of retrieval diagnostics. The process of identifying which queries your content scores highest for in AI retrieval reveals the specific language, concepts, and product attributes your target customers use when they are actively researching a purchase. That query intelligence translates directly into Google Ads keyword strategy and ad copy that matches demonstrated buyer intent rather than assumed intent. On Meta, the semantic themes that drive high cross-encoder relevance scores — the specific product attributes and use-case framings that resonate with high-intent queries — translate into audience messaging and creative angles that speak to customers at the consideration stage. For Klaviyo, understanding which content clusters your brand owns in AI retrieval helps define the educational and product content that belongs in nurture sequences for different customer segments. REACH gives you a map of how your customers think about your category — and that map is useful everywhere they encounter your brand, not just in AI search.

REACH diagnostics are designed to be precise — and precision cuts both ways. The platform identifies exactly what is underperforming and why, but acting on those findings correctly requires understanding how changes interact across the full pipeline. A product description rewrite that improves dense vector scores can inadvertently introduce extraction noise that reduces chunk quality at the next stage. A schema implementation done partially is sometimes worse than no schema at all. A content expansion that adds word count without adding semantic depth moves the wrong metric. The brands that see the strongest retrieval improvements from REACH diagnostics are those working alongside an experienced team that understands not just what the scores mean, but how the stages interact, how to sequence change for compounding impact, and how to avoid the common implementation mistakes that neutralize good diagnostic findings before they reach the ranking layer. REACH tells you exactly what the pipeline sees. Getting the pipeline to see what you want it to see is where expertise earns its return.

Yes — and the downstream effects are broader than most brands expect. The same content quality, structural clarity, and semantic density that drives AI retrieval performance also determines how well every other AI system in your stack performs against your brand's content. Your site's AI chat widget answers customer questions by retrieving and synthesizing content from your own pages — the same extraction and relevance mechanics that REACH diagnoses. If your product pages extract poorly or carry thin semantic content, your chatbot gives vague, incomplete answers regardless of how sophisticated the underlying model is. The same applies to AI-powered email personalization engines that pull product context from your catalog, shopping feed optimization tools that rely on structured product data, and any AI platform that ingests your content to generate recommendations, descriptions, or customer-facing responses. Content that is clean, well-structured, and semantically rich for retrieval is, by definition, content that every AI system in your stack can work with effectively. REACH optimizations are not a narrow AI search fix — they are a foundational content infrastructure investment whose benefits propagate through every AI touchpoint in your customer experience.

Significantly more important — and the reason is one of the most consequential dynamics in the coming era of AI-mediated commerce. Google's Universal Commerce Protocol, co-developed with Shopify, Target, Walmart, and Etsy and endorsed by Visa, Mastercard, and Stripe, establishes a standardized layer through which AI agents handle product discovery, cart building, and checkout autonomously across any retailer. When that infrastructure matures, every brand competing for AI-driven purchase recommendations will be feeding the same protocol the same structured fields — product title, description, price, availability. The structured data layer becomes a commodity. Every brand has it. Every brand looks, at the protocol level, roughly the same. What doesn't become a commodity is semantic content quality. When the protocol fields are equal, the AI agent's recommendation decision falls back to the same retrieval and relevance mechanics REACH already measures — dense vector similarity, cross-encoder relevance scoring, and semantic depth of the content behind the product listing. The brands that win in a UCP world are the ones whose underlying content is semantically richer, more precisely aligned to buyer queries, and more completely extractable by AI systems than their competitors' content. REACH measures exactly those dimensions — today, before the protocol matures, when building that content advantage is still a differentiator rather than a survival requirement. The window to establish a structural retrieval advantage before AI commerce protocols standardize the playing field is open now. It will not stay open indefinitely.

A single diagnostic is a snapshot — and in AI retrieval, the snapshot expires faster than most brands expect. The query universe your customers are using to discover products in your category is not static. New competitors enter, existing competitors update their content, AI platforms shift their retrieval weighting as their models are updated, and seasonal intent patterns change which queries are high-volume and high-stakes at any given time. A brand that ran a REACH diagnostic six months ago and implemented the findings is not optimized today — it is operating on six-month-old competitive intelligence in a landscape that has moved around it. Beyond competitive drift, your own content changes — new products launch, descriptions get updated, pages get restructured — and each change has retrieval consequences that are invisible without measurement. REACH is most valuable as a continuous monitoring layer: establishing your baseline, tracking the impact of content and structural changes as they are deployed, and surfacing competitive movement in your query universe before it translates into lost consideration. The brands building durable AI retrieval advantage are not the ones who ran one diagnostic. They are the ones who treat retrieval performance as an ongoing operational metric, the same way they treat conversion rate or return on ad spend.

Different page types carry fundamentally different semantic responsibilities in AI retrieval, and treating them the same is one of the most common and costly optimization mistakes we see. A homepage needs to establish broad brand authority and category ownership — its retrieval job is to answer "who is this brand and what do they make" with enough semantic richness that AI systems can accurately represent the brand in overview and comparison queries. A collection page needs to own a category concept — "best trail running packs," "hydraulic disc brakes for enduro" — and its retrieval performance depends on how well it frames the category, names the relevant use cases, and positions the brand's range within that context. A product display page has the most precise semantic duty of all: it needs to answer the specific, high-intent questions a buyer asks immediately before purchase — compatibility, performance under specific conditions, comparison to named alternatives, suitability for specific rider profiles or use cases. Each of these page types requires a different content strategy, different schema implementation, and different query targeting to perform well in AI retrieval. REACH evaluates each page type against the queries appropriate to its semantic role, identifies where content is failing that role, and prioritizes fixes by retrieval impact at each layer of the pipeline. Optimizing a product page like a homepage, or a collection page like a product page, produces work that scores well on the wrong queries — and invisible on the ones that actually drive purchase decisions.

Generally REACH has been designed to elevate products over services or retail partners. We might in the future adapt the process for other segments but that is not the current focus of REACH.