Precision, coverage, traffic numbers, pricing — answered the way we answer them in calls. If your question isn't here, write to us and it will be by next week.
Accuracy is measured, not asserted. Before any full run, we screen a sample your team reviews, tune the signal definitions, and repeat until precision clears the bar you set.
In a large-scale production run on our own classification platform, a reviewed sample of a 1.2M-domain run confirmed roughly 96% precision. Your pilot measures the number for your vertical and ICP.
Keyword filters can't tell a travel blog from an online travel agency — both talk about flights. Our signal vocabulary can: an "own commerce" signal detects sites selling inventory themselves and disqualifies them structurally.
Combined with signals like blog-like structure and individual-or-duo operation, the ICP describes a publisher shape, not a topic. That removes the biggest class of false positives in this market.
A precise set of signals from a 20-signal vocabulary: some required, some preferred, some disqualifying. Every signal is extracted per domain with a confidence score, and fit is computed from that logic.
Each of our 30 industries ships with five typical advertiser ICPs as starting templates. See the method page for a worked example.
From reproducible public proxies only: presence in the our real-world browsing popularity dataset and rank groupings. Anyone on your team can re-verify them from public sources.
We never publish estimated visitor counts — nobody outside a site's own analytics knows those numbers, and pretending otherwise fails our standards.
The source database covers the global web, and language is one of the extracted signals — English-language screening is the default configuration. Country focus is delivered per domain as an identity field.
Non-English or geo-specific universes are configured as custom scopes and calibrated the same way. Tell us the target markets and we'll scope it in the pilot.
Every delivery goes through a dedup and canonicalization pass: mirrors, redirects, and country variants collapse into one canonical domain per publisher.
You get one row per real site — so counts are honest and your outreach team never contacts the same publisher twice.
Yes. Custom signals — "accepts sponsored posts", "already works with affiliate networks", "covers business travel" — are defined in writing per engagement.
They go through the same reviewed-sample calibration as the standard vocabulary, so their precision is measured before the full run.
Runs are executed live against current site content, and a "fresh content within 6 months" signal is part of the vocabulary — you can require actively maintained sites.
After delivery, optional quarterly refresh plans re-screen your universe and ship deltas: new publishers, dead domains, pivots. Monthly cadence is available through the annual monitoring plan.
A structured CSV, one row per canonical domain: identity fields, per-signal values, confidence scores, ICP verdict and fit score, traffic proxies, plus any custom fields.
It imports directly into partner CRMs and outreach stacks — no proprietary viewer, no lock-in.
No. We don't scrape or resell personal contact data — that's a hard line in our standards, and it keeps your program on the right side of GDPR.
The deliverable identifies the right sites with evidence; outreach then uses the channels each publisher itself publishes.
We are an EU-based EU company, and deliveries are built from publicly published site-level characteristics rather than personal data.
Data minimization is the default: no scraped personal contact details, no profiling of individuals — domains and their editorial signals.
We configure the complete run for your vertical and ICP, then stop it once 10 publishers qualify with confident classifications. You get those 10 with all fields, free.
You also get a measured projection of the full-run yield and quotes for every scope — so plan selection is grounded in your data, not guesswork. Start here.
Plans run from $999 for a niche vertical with standard fields to $25,000 for multi-category, multi-language platform mandates, with annual monitoring from $1,000 per year, scaled to project size.
Price is driven by two dimensions: vertical size and field complexity. The pricing page has the full breakdown — and the pilot tells you which tier you actually need.
Alpha Quantum, an EU-based data company operating a 120M+ domain classification database used by 300+ enterprise organisations in telecom, adtech and cybersecurity.
Affiliate Targeting productizes that infrastructure for partner teams — more on the about page.
We don't publish client names — reference engagements are described by category and verifiable numbers only, and public samples are paraphrased and anonymized.
The same discretion will protect you as a client. The pilot exists precisely so the data, not a logo wall, makes the case.
10 qualified publishers for your ICP, every field and confidence score included, plus a projection of the full run — free.
Request a Free Pilot