Affily’s first product
The same product, the same audience, the click at a third of the price.
Analyser pulls together the catalogue of several networks, measures real search demand and tells you which product is worth competing for — and in which country the click is cheap.
- 6,662active offers
- 2networks integrated
- 20 countries11 languages
- 748,702demand measurements
- 09/06/2026state of the database
The top-of-page bid for the same term, from the same measurement. In the US the sale has to pay for 21 clicks; in the Netherlands, with the full demand preserved, the same product costs less than a third per click.
Auction map
| Product | Country | Searches/mo | Top-of-page bid | Commission | Clicks per sale | Network |
|---|---|---|---|---|---|---|
| Product A | United States | 4,400 | US$ 5.68 | US$ 120 | 21.2 | Network A |
| Product B | United States | 22,200 | US$ 7.48 | US$ 160 | 21.3 | Network A |
| Product C | United States | 27,100 | US$ 4.09 | US$ 84 | 20.5 | Network A |
| Product D | Australia | 390 | US$ 5.76 | US$ 126 | 21.8 | Network A |
| Product E | France | 8,100 | US$ 3.51 | € 110.60 | 26.2 | Network B |
Commission stays in the network’s declared currency — the last row is from a network that pays in euro, and the conversion used to calculate clicks per sale uses a recorded exchange rate, with a date. Without a rate, the ratio is absent — never 1.00.
The problem
Three ways to get it wrong that the market has already normalised.
Affiliate marketers pick two or three products out of thousands every month — and pay out of their own pocket for the traffic that will test them. The wrong choice does not hand back a bad report: it hands back a Google Ads invoice.
The spreadsheet that rots
Every network has its own dashboard, its own CSV, its own vocabulary. Consolidation becomes a hand-copied spreadsheet that ages in a week and keeps no decisions: the next month the whole catalogue is reassessed as if nothing had been decided.
The wrong metric with the right face
Tools in this category show “domain traffic over the last 30 days”. That number measures visits to the advertiser’s site — largely paid traffic bought by affiliates themselves. It is a saturation gauge, not demand. Worse: because the metric belongs to the domain, seven different products show the same identical number.
The invented number
Where the expensive data does not exist, it gets estimated. A financial decision tool that displays an estimate without saying it is an estimate is worse than one that displays an empty cell.
How it works
From raw catalogue to live campaign, in six passes.
Each pass is isolated and resumable. None depends on the previous one having finished in full — stopping halfway never loses what has already been processed.
- 01
Collection
Pulls in the catalogue of each connected network. We plan to integrate more than 64 product sales platforms. Login is always user-assisted: we never automate a password or store a credential.
- 02
Term derivation
The title the vendor writes is a promotional headline, not a product name — in 938 of 1,247 offers in the initial measurement. The system extracts the real search term using the sales page domain as an anchor, and declares the provenance of each one.
- 03
Demand measurement
Queries monthly search volume, CPC, top-of-page bid and auction competition from Google — by term, by country and by language. Every figure carries the month it refers to.
- 04
Score
A single 0-to-100 verdict per offer, assembled from auditable deductions. Hovering the cell shows every deduction applied and every check that did not run.
- 05
Curation
Three manual states per offer. Nothing promotes itself. It is this record that makes the next month start where the last one stopped, instead of reassessing the whole catalogue from scratch.
- 06
Campaign
The work after the choice, inside the system: request affiliate link access (with the English email already drafted, ready to copy), store the link, record the published site, mark the Google Ads launch — and domain name ideas, with RDAP checking on demand.
The score
Every offer starts at 100 and loses points for a declared reason.
It is not a black-box score. There are five checks, each with a maximum deduction, and the screen shows which one ran, how much it took off and which one could not run. A check that does not run does not deduct.
- No identifiable product name
- −70
- Third-party brand in the term
- −45
- Demand below 100 searches/mo
- 0 to −35
- Commission below US$ 5
- 0 to −25
- Google evidence (percentile in the language)
- 0 to −100
Differentiators
Four things only it does.
Demand belongs to the product
Measurement is per search term, never propagated from the domain. Two products from the same vendor have different numbers because their demand is different.
Country and language in the same key
The same term in Canada in English and in French are two measurements. English is measured across all 20 countries — including where it is not dominant — because it is the axis that makes countries comparable.
“Potential product”, read from the SERP
Distinguishes a product from a common word. A term with 5 million searches can be nothing but a misspelling; another that looks generic can be a real product with a listing on a major marketplace. What answers that is the search result itself, not a dictionary.
The whole funnel in one place
From the catalogue of 6,662 offers to the product with the link saved, the site published and the campaign live. What the user builds stays separate from what the market says: having a published site is not evidence about the offer.
The screens
Two tabs and a detail page.
Offers
The consolidated table: ~30 columns with adjustable visibility, sorting and width, facet filtering, search, and a tooltip per column declaring its source and semantics. This is the discovery screen — it shows the whole catalogue and hides discarded offers by default.
Tracking
A work queue, not a reading list: who already has a link and needs a site, who is published, which commission is higher, where the bid is cheaper. Sortable by any column.
Offer detail
The monthly demand series by country and language, the search results with position and domain, the score breakdown, the auction by country, the campaign block and the domain ideas.
Numbers
The real state of the system on 09/06/2026.
- 6,662active offers in the catalogue
- 2,785distinct sales page domains
- 748,702demand measurements in the database
- 2,167terms with measured demand
- 14 monthsof continuous series, from Jul 2025 to Aug 2026
- 371(product, country) pairs in the auction map
- 64platforms mapped for integration
None of these numbers is rounded for effect.
Vocabulary
The terms, as the audience uses them.
- Offer
- A product available to promote on a network. The unit of everything.
- Network / platform
- Whoever intermediates the commission between the vendor and the affiliate.
- Search term
- The name the product is actually searched for on Google.
- Top-of-page bid
- What the click costs to appear at the top of Google in that country.
- Clicks per sale
- How many clicks one commission pays for. Below 20 is the working cut-off.
- Commission
- Always in the currency the network declares. Different networks pay in different currencies.
- Track
- Marking “I am going to sell this product”. A manual decision, never automatic.
- Evidence month
- Which month the displayed number refers to. Different months are never mixed in the same row.
Analyser ships first.
One email when it opens. No newsletter, no drip sequence.
We email you before launch. No spam. Leave whenever you want.