Affiliate Site Analytics Without Google Analytics: What to Track and With What
Google Analytics on an affiliate site is a lot of dashboard for four numbers. GA4 needs a cookie banner, a tag manager, custom events for outbound clicks, and a reporting interface built for apps and funnels; what a review-site owner wants to know is which pages get search traffic, which ones send readers to Amazon, and which ones earn.
Those four numbers come from two free sources and one small tool, and a page a month of looking at them.
This post sets out the four numbers that matter and why the rest don't, where each one comes from, the lighter tools that replace GA4 for a content site, what to do about the cookie banner, and a one-page monthly view that answers the only question that matters: what to write and fix next. It builds on the Search Console post, which covers the first source in detail.
The short answer
Track impressions and clicks per page from Search Console, outbound affiliate clicks per page from a lightweight analytics tool, and earnings per page group from Amazon tracking IDs. Skip GA4 unless you run ads or a newsletter funnel. Put the three sources on one page once a month, sort by the gap between clicks and outbound clicks, and that list is the work.
The four numbers
| Number | What it tells you | Source | Decision it drives |
|---|---|---|---|
| Search impressions per page | Google is showing the page | Search Console | Which pages are close to ranking; what to refresh |
| Search clicks per page | Readers are arriving | Search Console | Title and snippet work; which topics to deepen |
| Outbound affiliate clicks per page | The page sends readers to Amazon | Lightweight analytics (outbound link tracking) | Placement and verdict fixes; which pages leak readers |
| Earnings per page group | The clicks turn into orders | Amazon tracking IDs | Which page types and topics to write more of |
Everything else, from session duration to scroll depth to bounce rate, describes behaviour on pages whose goal is to keep the reader. An affiliate page's goal is to send the reader away through a link, and the metrics built for engagement read that success as failure.
Why GA4 is the wrong size
GA4 is built around events and conversions for apps and ecommerce funnels. For a content site it means: a consent banner for EU and UK visitors, a tag manager or plugin to fire outbound-click events, custom reports to see them per page, a script that adds to every page's load, and an interface where the four numbers above are three clicks deep each.
It can be made to work. It's a day of setup and a monthly chore to produce what a smaller tool shows on one screen.
It's worth it when you run paid traffic, email funnels or on-site sales, where its event model earns its complexity. For a review site earning from Amazon links, it's the wrong size, and the site speed post lists its script among the things that slow pages.
The lighter tools
| Tool | Outbound clicks | Cookie banner needed? | Cost | Fit |
|---|---|---|---|---|
| Plausible | Yes, one toggle | No in most cases | Paid, small | Simplest hosted option |
| Fathom | Yes | No in most cases | Paid, small | Same idea, different vendor |
| Umami | Yes, via events | No in most cases | Free, self-hosted | If you'll run a small server |
| WordPress stats plugin (Jetpack, Koko, etc.) | Some | Varies | Free tier | Zero setup; check it tracks outbound clicks |
| Search Console alone | No | No | Free | Fine until you need outbound clicks |
The requirement is one: outbound clicks per page, shown next to page views. Any tool that does that without a consent banner is the right size. The cookieless tools achieve it by not storing anything on the visitor's device, which is what makes the banner unnecessary under most readings of the EU rules; check the vendor's own statement for your jurisdiction.
Connecting pages to money
No analytics tool sees the order; Amazon does, and reports it by tracking ID. One ID per page type (guides, reviews, comparisons) or per topic is enough to see which groups of pages earn, and a temporary ID on one page answers a question about that page.
The tracking IDs post covers the setup. The join between the two sources is by hand: the analytics tool says a guide sent 300 readers to Amazon this month, the tracking ID for guides says the group ordered 40 items, and that ratio across groups is your conversion by page type.
The one-page monthly view
A spreadsheet with one row per guide and five columns: impressions, search clicks, outbound clicks, outbound rate (outbound ÷ search clicks), and the tracking ID group it belongs to. Fill it once a month from the three sources (twenty minutes once the export is routine), then sort twice:
- By impressions, descending, where clicks are low. Pages Google shows and readers don't choose: title and snippet work, or position 11 to 20 pages that need links.
- By search clicks, descending, where outbound rate is low. Pages readers arrive at and don't leave through a link: verdict placement, button placement, out-of-stock picks.
The top five of each sort is the month's work. A page with high outbound rate in a tracking-ID group that orders well is the model for the next guides on that topic.
What not to bother with
- Real-time dashboards. Nothing on a content site changes in an hour.
- Heatmaps and session recordings. Occasionally useful for a layout question, never as a monthly habit.
- Bounce rate and time on page. See above.
- Demographics. You already know who the reader is; it's in the keyword.
- Rank trackers for every keyword. Search Console's average position per page is enough, and it's free.
Doing it with AffSERP
AffSERP's side of this is the tracking ID: each site has its own, applied to every link the X Best Review tool writes, so the earnings column of the monthly view splits by site without any extra setup, and a per-run ID is available when you want one batch reported on its own.
The page-level numbers still come from Search Console and whichever lightweight tool you pick; the guides the tool produces have the verdict and section buttons placed where outbound clicks happen, which is the fix the second sort usually calls for.
FAQ
Do I need Google Analytics for an affiliate site?
No. The numbers that decide anything on a review site are search impressions and clicks per page (Search Console), outbound affiliate clicks per page (a lightweight analytics tool or link counter), and earnings per page (Amazon tracking IDs). GA4 can give you the middle one, with a lot of setup, a cookie banner and reports built for apps rather than content sites.
What's the best lightweight analytics for a WordPress affiliate site?
A privacy-friendly page-view tool with outbound link tracking: Plausible, Fathom, Umami (self-hosted and free), or a simple WordPress stats plugin. Any of them gives page views and outbound clicks per page in one screen, without a cookie banner in most jurisdictions, and loads a script a tenth the size of GA4's.
How do I know which posts make money?
Amazon tracking IDs per page type or per topic, read in the Associates Earnings report. That's the only source that connects a page to an order. Analytics tools see the click leave; Amazon sees the sale. Combine the two and you have clicks per page and orders per group of pages, which is enough to decide what to write next.
Is bounce rate useful for an affiliate site?
No. A reader who lands on a guide, clicks the Amazon button and never comes back is a bounce and a sale. Most engagement metrics describe pages where the goal is to stay; on an affiliate page the goal is to leave through a link. Track outbound clicks instead.
Keep reading
