Component 08
Onelo does not exist in AI-generated search results. Not weakly — absent. Across 12 queries tested on ChatGPT, Perplexity, and Google AI Overviews, Onelo was not mentioned once. Every primary competitor was mentioned in multiple responses. This is the most urgent secondary constraint in the engine. The window to establish AI presence before competitor positions become entrenched is narrowing, and the remediation programme has a 6–12 month compounding timeline that must start now.
This document covers all 12 signals in the AI Visibility component. For each signal, you will find: what was assessed and why it matters, the specific findings for Onelo, evidence supporting those findings, and the recommended intervention.
A signal is a subcomponent of any of the ten layers that make up an organic growth engine. Each signal is assessed thoroughly following our methodology and assigned a status: Healthy, Fragile, Blocking, or Missing. For each signal, there is supporting evidence and recommendations for how to turn each signal healthy.
AI Visibility is Missing because no infrastructure for it exists. The remediation programme is not about fixing something broken — it is about building something that was never built. That requires a different mindset than the remediation work in other components: it is longer, it is less immediately measurable, and it compounds slowly before it compounds fast.
The good news is that the work required for AI visibility is largely the same work required for Category Presence. Category landing pages built with FAQ schema, structured data, and citeable content serve both components simultaneously. The buyer’s guide content serves both Demand Match and AI Visibility. The structured data sprint serves both Narrative & Positioning and AI Visibility. This is a programme, not a separate workstream.
Structured data and entity signals (Weeks 4–6)
Update SoftwareApplication applicationCategory across all commercial pages. Implement FAQPage schema on top 5 commercial pages. Implement speakable schema on homepage and solution pages. Create and publish llms.txt. Deploy in a single development sprint. Estimated effort: 3–4 developer days. This workstream produces results fastest — AI systems incorporate structured data signals within weeks of crawling.
FAQ and citation-ready content (Weeks 4–16)
Add FAQ sections to all new category pages as they are built. Write 3 standalone FAQ blog posts targeting evaluation-stage queries (pricing, features, implementation time). Add FAQ sections to /solutions/mid-market-onboarding and /product/onboarding-automation. Add market-context pricing section to /pricing. Build the comprehensive buyer’s guide. This workstream produces results more slowly — 8–16 weeks for content to be indexed and incorporated into AI training cycles.
Entity signal consistency (Weeks 6–10)
Update G2 and Capterra product descriptions to include ICP-specific language and competitive positioning. Update Crunchbase category. Update CEO LinkedIn profile to include specific category language. Ensure all new press releases and media outreach use ‘mid-market’ and ‘200–2,000 employees’ consistently. This workstream is editorial and requires no development resources.
Monitoring and iteration (Ongoing from Week 8)
Run the 12-query AI visibility test monthly. Track Perplexity citation frequency. Monitor AI Overview appearances weekly using a saved Google Search for the 4 primary commercial queries. Set 6-month milestone: Onelo appears in at least 4 of 12 queries. Adjust content and schema programme based on which workstreams are producing citation results fastest.
AI visibility does not move quickly. The honest timeline for a company starting from zero with a focused programme:
A Blocking component has infrastructure that is failing. A Missing component has no infrastructure at all. Onelo has no AI visibility signals in place — no FAQ schema, no structured entity data, no AI-optimised content, no llms.txt, no presence on the citation sources AI systems draw from. This means the remediation programme does not start from a weak base and improve it. It starts from zero and builds. That is a longer timeline and a different type of work.
AI systems form their representations from signals that exist at the time of training and retrieval. Competitors who began building AI visibility signals in 2023 have 18 months of compounding advantage. Rippling appears in 10 of 12 tested queries. BambooHR appears in 9 of 12. These positions are not locked permanently — but they are increasingly expensive to displace the longer Onelo waits. Starting the AI visibility programme in parallel with Category Presence (week 4 of the intervention sequence) is the correct call precisely because it takes longest to compound.
This signal tests whether Onelo appears when a buyer asks a major large language model to recommend or list solutions for the onboarding software category. These are the queries a buyer might ask ChatGPT or Perplexity before they open a browser tab — the first moment of vendor discovery in an AI-assisted research flow. Appearing here means being in the consideration set before traditional search begins.
Onelo did not appear in any of the 6 category queries tested across ChatGPT (GPT-4o) and Claude. Rippling appeared in all 6. BambooHR appeared in 5 of 6. Deel appeared in 4 of 6. The gap is not marginal — Onelo is structurally absent from the AI category consideration set, not ranked lower within it.
Full keyword intent distribution — classified portfolio
The pipeline-to-traffic ratio makes the imbalance concrete: buyer-intent queries generate 94% of pipeline from 38% of sessions. Off-category queries generate less than 1% of pipeline from 15% of sessions. Every session on an off-category page costs the same in infrastructure as a session on a buyer-intent page, but produces 38x less commercial value.
RECOMMENDATION
Apply intent classification to the content brief process going forward. Before any new piece of content is commissioned, the brief must state which intent type it targets, which buyer persona it addresses, and which stage of the evaluation journey it serves. Content targeting informational intent should only be commissioned if it has a credible pathway to commercial navigation — i.e., the topic is one that a mid-market HR Director genuinely encounters before evaluating onboarding software, and the page is designed to capture that transition.
Alternative queries — ‘alternatives to Rippling’, ‘BambooHR competitors’, ‘tools similar to Deel’ — are among the highest-intent AI search queries a buyer can make. They represent buyers who have already evaluated a primary option and are actively looking for alternatives to compare. Appearing as an AI-recommended alternative to a competitor with a larger presence is one of the fastest paths to new buyer discovery. Absence means being invisible at the exact moment a buyer is most receptive to a new vendor.
Onelo does not appear as an AI-recommended alternative to any of its three primary competitors. In every alternative query tested, Rippling, BambooHR, and Deel recommended each other, Workday, Lattice, and occasionally niche tools — but never Onelo. This is commercially significant: buyers who find Rippling too broad or BambooHR too SMB-focused and are actively searching for a mid-market workflow-automation alternative are the perfect Onelo prospect, and AI systems are routing them elsewhere.
Traffic split by page type — last 90 days
The navigational 26% pipeline contribution comes almost entirely from the pricing page — which converts at 4.8%, the highest rate on the site. Buyers who navigate directly to pricing are self-qualified and highly motivated. This is a useful reference point: it demonstrates that the conversion architecture works when buyers are qualified. The problem is getting more qualified buyers to the site through organic in the first place.
RECOMMENDATION
The 14% commercial traffic share cannot be meaningfully improved by changing existing blog content. It requires two parallel actions: (1) building the category landing pages specified in Component 03, which will add dedicated commercial pages to the site that attract buyer-intent traffic by design; and (2) improving internal navigation from informational posts to commercial pages so that the 3.2% of informational visitors who do navigate commercially are better supported.
When Onelo is mentioned by name in an AI query, how accurately do AI systems describe it? Accuracy failures range from vague generic descriptions (‘an HR software tool’) to active misclassification (‘a small business payroll platform’). In both cases, the AI system is actively working against Onelo’s positioning — either by failing to communicate the differentiation, or by communicating the wrong positioning to a buyer who might otherwise have been interested.
The representation accuracy picture is mixed: ChatGPT provides a vague but directionally correct description when prompted directly. Claude acknowledges limited knowledge. Gemini declines to describe the product reliably. Perplexity does not surface Onelo in prompted responses. Most concerning: one ChatGPT response described Onelo as suitable for small businesses — a direct misclassification that contradicts the ICP and could actively deter the right buyers.
Funnel stage coverage map
The decision stage is the most striking gap: 89 ranking queries generating 58% of organic pipeline from only 8% of sessions. These are the highest-value queries in the portfolio, and Onelo holds ranking positions on only a fraction of what exists in the market. Every decision-stage keyword that Onelo does not rank for is a buyer in active vendor evaluation that the organic channel is invisible to.
Decision-stage keyword gap — what Onelo is missing
A keyword gap analysis comparing Onelo’s ranking portfolio against the decision-stage queries that Rippling and BambooHR rank for reveals the specific queries where qualified buyers are currently invisible to Onelo.
Decision-stage queries where Rippling ranks but Onelo does not (sample):
RECOMMENDATION
Develop a decision-stage content programme as the primary content investment in Phase 2 of the intervention sequence (weeks 12–24). The programme should prioritise three content types that are consistently high-value at the decision stage in B2B SaaS:
Beyond general accuracy, this signal assesses whether AI systems correctly understand Onelo’s specific audience and use case — the 200–2,000 employee mid-market, the HR Director and COO buyer, and the workflow automation depth that differentiates the product. A system that describes Onelo correctly in general terms but attributes it to the wrong audience or use case will route the wrong buyers toward it and potentially route the right buyers away.
Beyond general accuracy, this signal assesses whether AI systems correctly understand Onelo’s specific audience and use case — the 200–2,000 employee mid-market, the HR Director and COO buyer, and the workflow automation depth that differentiates the product. A system that describes Onelo correctly in general terms but attributes it to the wrong audience or use case will route the wrong buyers toward it and potentially route the right buyers away.
Branded BoFu keyword performance
The ‘Onelo vs BambooHR’ position 8 on a blog post is a missed opportunity. A dedicated comparison page with structured content and comparison schema would likely rank position 2–4 for this query — and would convert at significantly higher rates than a blog post.
Non-branded BoFu gap — alternative queries Onelo should own
These are queries where buyers are evaluating a competitor and could be captured by Onelo with a well-built alternative page. All are currently unranked by Onelo.
[Link to spreadsheet: Full non-branded BoFu keyword gap analysis — all ‘alternatives to [competitor]’ and ‘[competitor] vs’ queries with volume >100/month, sorted by ICP alignment score and estimated traffic value.]
RECOMMENDATION
Build three types of BoFu pages as priority content in Phase 2:
Perplexity AI is a search engine that generates answers with cited sources — it is closer to traditional search than to a chatbot, and its citations directly drive traffic to the cited pages. Appearing as a cited source in Perplexity responses is commercially valuable both as a discovery mechanism and as a trust signal (being cited implies the content is authoritative). This signal tests Perplexity presence specifically because its citation mechanism is more transparent and more directly traceable than ChatGPT or Claude.
Onelo does not appear as a cited source in any of the 6 Perplexity queries tested. The sources Perplexity consistently cites for onboarding software queries are: G2 category pages, Capterra comparison pages, Rippling’s website, BambooHR’s website, and high-authority HR publication articles. Onelo’s G2 and Capterra profiles are cited as part of the category pages, but Onelo’s own website content is never the primary cited source.
Intent alignment audit — top 10 landing pages
The pattern is clear: pages where alignment is high convert at 3–4x the rate of pages where alignment is low. This is not a conversion rate optimisation problem — it is an alignment problem. A/B testing button colours on misaligned pages will not close this gap.
RECOMMENDATION
For each of the 6 misaligned pages, define the primary query intent driving their traffic and rebuild the above-fold experience around that intent — not around what the company wants to say. This is the core principle: the page answers the buyer’s question first, then presents Onelo as the solution.
AI systems are optimised to answer questions. They preferentially cite content that is structured as question-and-answer pairs because it directly matches the format of their output. A site with extensive FAQ content, question-format blog post titles, and structured Q&A sections gives AI systems the exact building blocks they need to generate responses that cite the site. A site without this content format is structurally harder for AI systems to cite even if the underlying content is good.
Of Onelo’s 94 blog posts, 3 use question-format titles (‘How to…’ or ‘What is…’). Zero use FAQ structured content with schema markup. Zero commercial pages have FAQ sections. This is a near-complete absence of the content format that AI systems preferentially cite. Rippling, by comparison, has FAQ sections on 34 pages with FAQPage schema markup — which is why Rippling is cited in AI responses for queries that Onelo’s content could theoretically answer.
Commercial page organic traffic — current and potential
Combined commercial page traffic: ~4,320 sessions/month. If category page build moves the top 4 commercial pages to positions 5–8 for their target queries, this figure should reach 12,000–15,000 sessions/month within 12 months — a 3x increase without changing the number of commercial pages.
RECOMMENDATION
The category page build (Component 03) is the primary lever for increasing commercial page organic traffic. Each new category landing page is effectively a new commercial page — built specifically to rank for buyer-intent queries and drive direct organic sessions to commercial content.
Structured data provides machine-readable signals about what a company is, what it does, who it serves, and what content on the site is designed to answer specific questions. For AI visibility specifically, structured data is one of the most direct ways to communicate entity information to AI systems — it bypasses the ambiguity of natural language and provides explicit declarations that AI systems can incorporate with high confidence.
This signal is rated Fragile rather than Missing because some structured data exists — Organization and SoftwareApplication schema are implemented. But the implementation is incomplete in two important ways: the category classification is too broad to contribute meaningful AI positioning signals, and the AI-specific schema types (FAQPage, HowTo, speakable) that are most valuable for AI citation readiness are entirely absent. This is the fastest-path intervention for improving AI visibility — structured data changes can be deployed in a single development sprint and begin influencing AI systems within weeks.
Content-to-commercial navigation analysis
Top 5 blog posts generating commercial navigation (these are the model to replicate):
Bottom 5 blog posts by commercial navigation (high traffic, near-zero commercial engagement):
RECOMMENDATION
Apply the navigation pattern of the top 5 performing posts to the top 20 posts by traffic. The common factor in posts with high commercial navigation rates is a contextual inline link within the body text — not a sidebar widget, not a generic footer CTA, but a sentence that naturally bridges the informational topic to the commercial solution.
llms.txt is an emerging standard that allows websites to explicitly communicate to AI systems which content is most relevant for them to index and how to represent the company accurately. It is the AI equivalent of robots.txt — a machine-readable file that shapes how AI crawlers interact with the site. While not yet universally adopted, it is being implemented by forward-looking companies as a direct signal to AI systems, and its absence places Onelo behind the small number of competitors who have already implemented it.
No llms.txt file exists at onelo.com/llms.txt. This is not a critical gap given the file’s current limited adoption, but it is a signal of AI readiness that takes under 2 hours to implement and has no downside. In the context of a company that is Missing on 9 of 12 AI visibility signals, adding llms.txt is a quick win that demonstrates AI visibility investment to any AI system that checks for it.
Conversion rate by segment — segmented analytics
The buyer-intent segment is already performing above benchmark. The Demand Match intervention is not about improving conversion rate — it is about increasing the volume of buyer-intent sessions. Doubling the buyer-intent traffic share (from 38% to 65% of total sessions) while maintaining conversion rate would increase organic pipeline from ~$89K to ~$153K/month — a 72% increase with no change to the conversion experience.
RECOMMENDATION
Use this segmentation data to reframe how the organic programme is measured internally. The current dashboard tracks total organic traffic and total organic leads — metrics that obscure the real performance because they blend high-value and low-value sessions. Replace the top-line metric with buyer-intent sessions and buyer-intent conversion rate. These two numbers, tracked weekly, tell a more accurate story about organic channel health.
Google AI Overviews appear at the top of search results for an increasing share of queries, particularly informational and evaluation-stage queries. They summarise information from multiple sources and cite the pages they draw from. Appearing in an AI Overview for a commercial query is a highly visible trust and discovery signal — the AI Overview appears above all organic results and includes the vendor’s name and a link. Absence from AI Overviews for relevant queries means missing visibility at the very top of the search result page.
Onelo appears in zero Google AI Overviews for any of the 4 commercial queries tested where AI Overviews were triggered. Rippling appears in AI Overviews for 3 of 4 queries. BambooHR appears in 2 of 4. The structural reason is the same as Perplexity citation absence: Onelo has no content formatted for AI extraction. AI Overviews draw from pages with clear question-answer structures, FAQ schema, and content that directly addresses the query being searched.
Content gap map — consideration stage
These are consideration-stage queries with meaningful search volume where Onelo has no ranking content. All reflect the research behaviour of an HR Director at a 200–2,000-employee company evaluating whether to invest in onboarding automation.
Content gap map — decision stage
Decision-stage gaps are the most commercially urgent because buyers searching these queries have already made a category decision and are choosing a vendor.
These 4 decision-stage gaps collectively represent approximately 1,650 monthly searches. At a 5% conversion rate (consistent with buyer-intent segment performance) and a $25K ACV, capturing these queries would represent approximately $85,000 in monthly pipeline from 4 pages.
RECOMMENDATION
Prioritise the content gap programme in three tiers. Tier 1 (weeks 12–16): build the decision-stage pages identified above — comparison pages, alternative pages, and the implementation-time page. These have the highest per-session commercial value and the most direct impact on pipeline.
This signal is not assessable in the traditional sense for Onelo — because Onelo is absent from all AI Overviews tested, there are no citations to assess for quality. The signal is documented here because it becomes relevant once the AI visibility programme begins producing results, and because understanding what citation quality looks like for competitors provides the template for what Onelo’s citation content should be built toward.
Based on analysis of competitor citations within AI Overviews, the content that gets cited shares three characteristics: it is on a dedicated page for the specific query topic (not a general product page), it contains specific and quotable claims (statistics, timeframes, named features), and it is structured with clear H2 subheadings that match the query intent. Rippling’s /onboarding page is the benchmark — it is cited in more AI Overviews for onboarding-related queries than any other vendor page because it is architecturally optimised for citation.
Three cannibalisation instances — detail and impact
Cannibalisation instance 1:
Cannibalisation instance 2:
Cannibalisation instance 3:
RECOMMENDATION
Consolidate all three cannibalisation pairs. The consolidation process for each pair: (1) identify the stronger page using the GSC data (higher impressions and position); (2) merge the content of both pages into the stronger URL, incorporating the best sections of the weaker page; (3) set up a 301 redirect from the weaker URL to the consolidated URL; (4) update all internal links pointing to the deprecated URL.
This signal compares Onelo’s brand mention share in traditional search (branded queries as a proportion of category queries) against its brand mention share in AI-generated responses. A company with growing brand awareness in traditional search but zero AI mention share is building a future problem: as AI-assisted research displaces traditional search for initial vendor discovery, a brand visible in traditional search but absent in AI becomes increasingly reliant on existing brand awareness rather than discovery.
Onelo’s branded search volume is growing at 18% year-over-year in traditional search — a healthy signal (Component 07). In AI-generated responses, branded mention share is effectively zero. The gap between traditional search brand growth and AI brand absence is widening. As AI-assisted research grows from its current estimated 40% of B2B evaluation journeys toward majority share, this gap becomes a structural acquisition problem.
Three cannibalisation instances — detail and impact
Cannibalisation instance 1:
Cannibalisation instance 2:
Cannibalisation instance 3:
RECOMMENDATION
Consolidate all three cannibalisation pairs. The consolidation process for each pair: (1) identify the stronger page using the GSC data (higher impressions and position); (2) merge the content of both pages into the stronger URL, incorporating the best sections of the weaker page; (3) set up a 301 redirect from the weaker URL to the consolidated URL; (4) update all internal links pointing to the deprecated URL.
AI visibility trajectory measures whether a company’s AI presence is improving over time, declining, or stagnant — and compares that trajectory against competitors. A company that began its AI visibility programme 18 months ago and has been consistently building signals has a compounding advantage that is increasingly expensive to close. A company starting from zero today can still close the gap, but the required investment grows with each month of delay.
Onelo’s AI visibility trajectory is flat at zero — no improvement detectable because no programme has been in place. Competitor trajectories are all positive. Rippling’s AI presence is estimated to have been building since Q1 2023, giving it an 18-month compounding advantage. The gap is material but not insurmountable: AI systems update their training data and retrieval indices regularly, and a focused 6-month programme will produce measurable results within that window.
Three cannibalisation instances — detail and impact
Cannibalisation instance 1:
Cannibalisation instance 2:
Cannibalisation instance 3:
RECOMMENDATION
Consolidate all three cannibalisation pairs. The consolidation process for each pair: (1) identify the stronger page using the GSC data (higher impressions and position); (2) merge the content of both pages into the stronger URL, incorporating the best sections of the weaker page; (3) set up a 301 redirect from the weaker URL to the consolidated URL; (4) update all internal links pointing to the deprecated URL.