A continuously refreshed graph of public buying-intent signals across the open web.
Wavly's Demand Graph scans 12+ public conversation platforms (Reddit, Hacker News, Stack Exchange, Dev.to, Lobsters, GitHub, Indie Hackers, Product Hunt, YouTube, X) on a 6-hour cycle. Each scan extracts threads where someone is explicitly asking for, comparing, complaining about, or shopping for products in your category. The graph is filtered through your ICP and scored for relevance + buying intent before any human sees it.
A single public post, comment, or thread that expresses buying intent against your ICP.
A signal carries the source URL, original text, author handle, posting timestamp, and engagement metrics. Each signal is scored on two axes: relevance (does the asker fit your ICP?) and intent (are they actually buying?). Sub-threshold signals are dropped before they reach your queue — your queue is the small, high-value subset.
A unique, attribution-tracked landing page generated for one specific signal in under 30 seconds.
When a signal scores high on relevance + intent, Wavly generates a personalized page that echoes the asker's language and pains. The page lives at thewavly.com/p/<slug>, carries a tracked CTA, and ties every signup it produces back to the originating signal. No competitor ships this — it's the unlock that turns a Reddit reply into a measurable conversion.
A 0–100 score for how likely a signal is to convert into a paying customer.
The intent score combines five inputs: (1) explicit buying language ("looking for", "alternative to", "recommend"), (2) thread engagement (replies, upvotes, age), (3) ICP fit, (4) anti-pattern penalties (hiring posts, self-promo, dead threads), and (5) the asker's posting history. The scoring formula is transparent and tunable — no neural net, no black box.
A repeating signal shape that empirically correlates with high conversion for your product.
After ~50 conversions, Wavly's pattern engine starts surfacing the language, subreddits, intent phrases, and posting times that historically convert best for YOUR ICP. Patterns become the basis for proactive scans + alerts: when a fresh signal matches a known winning pattern, it jumps to the top of your queue.
Software that runs continuously in the background and turns intent into pipeline — no human campaign needed.
The same way Vercel is the runtime for your deployed code or Stripe is the runtime for your billing, Wavly is the runtime for your distribution. You install it once (paste a URL), and it keeps running: scan → score → draft → page → reply → track → repeat. You stay in the approval loop; you don't run the loop.
A drafted, contextual public response attached to a signal, awaiting human approval.
Every Reply is generated by Scout (Wavly's reply LLM) using the original signal text, your product context, and the platform's tone (HN-formal, Reddit-casual, etc.). Replies are public answers — never bulk DMs, never link-dropping. You approve, edit, or reject in keyboard-first review; nothing posts without your green light.
Ideal Customer Profile — the filter that decides which signals are relevant to you.
Wavly extracts your ICP automatically from your product description: who the persona is, the pain points they have, the search keywords they use, anti-patterns to skip, and the communities (subreddits, tags, sites) where they hang out. The ICP is the source of truth for relevance scoring, subreddit filtering, and reply tone.