Complete FAQ
The real questions buyers ask before trialing Citable — each labeled with who typically asks it, so you can find your own question fast.
What is Citable?
Asked by: A founder or merchant hearing about Citable for the first time
Citable is the autonomous revenue engine for agentic commerce. We measure how AI agents decide what shoppers buy, act on what we find, and prove the results in orders.
When a buyer asks ChatGPT or Perplexity what to buy, the whole decision happens inside that conversation: the assistant compares options, weighs reviews, and recommends a shortlist. You see the order (or the lost sale) but never the why. Citable makes that decision measurable. Personas built from your real customers ask their questions across the assistants, so you see exactly how agents answer, why they recommend you or a rival, and what to fix. Then Citable fixes what agents read about your brand and joins every action to orders.
Why does 'measurable' matter? Marketing has always had analytics.
Asked by: A growth or data lead who already runs an analytics stack
Because analytics could never see a decision. It saw fragments.
A considered purchase used to scatter across ten platforms and die behind privacy walls, so marketing settled for correlation: last-click credit, attribution models, guesswork. Now the whole decision happens inside an AI assistant, in one observable conversation. That makes the decision itself a measurable object for the first time: which questions the buyer asked, what they were compared against, which doubt lost the sale. Your existing analytics instruments your website, and the decision has left the website. Citable measures at the agent, then joins what it finds to sessions and orders on your storefront.
Is Citable a GEO tool?
Asked by: A marketer evaluating GEO / AI-visibility tools
No. GEO tools are the analytics layer; Citable is growth software judged in orders.
GEO (Generative Engine Optimization) tools track how AI engines mention and cite brands, and they serve marketing teams with dashboards. They are not responsible for revenue. Citable works in a different shape, the one Klaviyo proved on email: own one growth channel end to end and be judged in the revenue attributed to it. For Citable that channel is agent commerce — the buying decisions that happen inside AI assistants. Monitoring is only the first layer of the loop; the product is the whole loop: measure the decision, act on the surfaces that influence the agent, and prove the results in your orders. If you're comparing GEO tools, the honest framing is that Citable includes what they do and is accountable for what they aren't.
How do I measure revenue from AI traffic?
Asked by: A growth lead who sees chatgpt.com in their analytics and gets asked what it's worth
Follow the chain: user → agent → conversion.
Citable joins three layers. It knows which questions buyers ask AI assistants in your category and how you're answered (the agent layer). It sees AI-referred visitors arrive on your site — chatgpt.com, perplexity.ai, and other assistant referrers — through your analytics integration (the traffic layer). And it connects those sessions to conversions on your storefront (the revenue layer). The join alone is becoming table stakes; the value is the layer above it — knowing why the agent recommended what it did, acting on that, and proving the action produced orders. One honesty note: AI agents shape the decision, but humans still check out on your storefront today. Citable claims no lift it has not measured; client brands typically see around 30% growth in AI-attributed revenue within three months.
How accurate is the data? AI answers change every time I ask.
Asked by: A skeptical founder or analyst who has watched ChatGPT give three different answers
Accuracy is the product. Citable treats measurement as an eval problem, not a single query result.
Citable runs the same prompt suite multiple times, deduplicates citations by canonical URL (accounting for redirects and tracking parameters), and versions personas as frozen snapshots so persona effects can be separated from model updates. When a model drifts significantly, Citable starts a new baseline rather than mixing old and new trend data. The result is a consistent signal you can act on — a real trend, not a screenshot that changes on refresh.
Is Citable safe to connect to my storefront and analytics?
Asked by: A brand owner approving a new vendor before connecting their store
Connectors are scoped, and nothing ships without your approval by default.
Integrations like Google Analytics, Shopify, and Reddit connect through scoped OAuth — you grant the specific access each connector needs, and you can disconnect at any time. On the action side, drafts land in a queue for your review: you edit and approve before anything is published, and community channels like Reddit are deliberately manual — Citable drafts, you post. On higher plans you can choose to delegate execution to the Citable Agent, but that delegation is your decision, not the default.
What does Citable actually do — is it analytics or execution?
Asked by: A buyer burned by dashboards that reported problems without fixing them
Both. It's a full see → explain → act → measure loop, not just a dashboard.
Most tools show you how you appear in AI engines today. Citable goes further: it diagnoses where and why you're losing, recommends the specific actions to close those gaps (content to write, sources to earn, fixes to ship), generates the content to execute, and re-tests after 30 days to confirm the engines' behavior shifted. The core unit is the Action Engine — instead of telling you "your visibility is low," it tells you what to do next and produces the work.
How does Citable prioritize which actions will actually move the needle?
Asked by: A small team that can only ship a few things per week
By demand x gap x consistency — starting with fixes that improve multiple AI engines at once.
We rank by which questions drive the most buyer demand for your category, where you're most consistently losing to competitors, and how stable that gap is across repeated runs and personas. If engines disagree on where you rank, we focus on "overlap wins" first — changes that lift you across multiple models simultaneously — then layer in engine-specific work based on where your buyers actually ask (Perplexity for research-heavy purchases, ChatGPT for mainstream recommendations, etc.).
If AI generates fewer clicks, why does it matter to track brand mentions at all?
Asked by: An SEO-minded marketer watching organic clicks decline
Because AI has become the decision layer. Clicks aren't the point anymore.
When a buyer asks an assistant "what should I choose for X," the answer shapes the shortlist before any website is visited. Being recommended consistently — even without a click — changes consideration, share of voice, and which brands get evaluated at all. And when the click does come, AI-referred buyers convert 42% better — up to 4x versus Google Search, per public industry data — which is exactly why the traffic is worth understanding rather than ignoring.
Who is Citable built for?
Asked by: A buyer checking whether their category fits
Merchants selling high-consideration ecommerce products — anything people research before buying.
The core categories are wedding, pet, wellness, beauty, consumer electronics, outdoor gear, and similar considered purchases. The deepest value is for Shopify and DTC brands, where Citable connects directly to the storefront and Google Analytics to tie AI visibility to actual orders. It also serves marketplaces and multi-SKU operators with product-level tracking, and agencies running AI-search services for merchant clients, with per-client workspaces and reporting on Grow and Scale. More broadly, it applies to any brand whose buyers research with AI, including B2B products shortlisted through AI research.
How is Citable different from Profound, Bluefish, Peec, Otterly, or AthenaHQ?
Asked by: A buyer with a monitoring-tool trial already open
Those are AI-visibility analytics tools built for marketing teams. Citable is in a different category: merchant growth software judged in orders.
Profound, Bluefish, Peec AI, AthenaHQ, and Otterly track brand mentions and citations across AI engines. They serve any brand's marketing team, and they are useful at what they do — but they are the analytics layer, not responsible for revenue, and none are judged on attributed orders. Citable's nearest neighbors are actually merchant-side tools like Wildcard and FERMÀT, software judged on merchant outcomes. Citable includes comparable monitoring as its first layer, then adds what an analytics tool never carries: explained gaps, an Action Engine that produces the content and fixes on a managed cadence, and attribution that connects AI traffic to orders. The difference in one line: marketing teams renew dashboards; merchants renew revenue. Citable is built to be renewed on revenue.
What is agentic checkout (UCP and ACP), and what does it change?
Asked by: A merchant reading about agents buying on shoppers' behalf
Agentic checkout protocols will let AI agents complete purchases on a buyer's behalf. Today they shape the shortlist; checkout is next.
UCP is the Google and Shopify-backed protocol; ACP is OpenAI-led. As these roll out, an agent will research, negotiate, and buy without the shopper ever browsing a website — which means website analytics sees even less than it does today. The practical takeaway for merchants: influence still flows into the agent from surfaces you control (your site and product data, reviews, third-party media), so the work of being chosen doesn't disappear — it becomes measurable at the agent instead of the pageview. Citable acts on those influence surfaces and measures at the agent, which is why the approach carries over as the protocols land.