How to Track Outbound Clicks on a Comparison Site

A comparison site earns nothing on the page. It earns on the click that leaves the page. That makes outbound click tracking the single most important measurement you set up — and the one most sites get subtly wrong.

Why outbound clicks are the real metric

Pageviews tell you what got found. Outbound clicks tell you what got chosen. On a comparison site those are frequently different pages, and optimising for the first at the expense of the second is how traffic goes up while revenue stays flat.

The pattern shows up clearly in fragmented markets, where the whole value of the site is that it organises supply nobody else has organised. Rental marketplaces are a good illustration: Rentay aggregates listings within a single national market, while RentsCompare compares providers across countries. In both cases the visitor arrives on a category page, scans several options, and leaves on one specific outbound link. If you only measure the landing page, you never learn which option won — or why.

What to capture on every click

Aim for enough context to answer "which listing, on which page, in which position, for which visitor intent" without storing anything personal.

• Destination — the merchant or provider, not just the raw URL.
• Source page — the exact path the click came from.
• Position — rank in the list, and whether it was above the fold.
• Component — hero CTA, table row, card, sticky bar. These convert very differently.
• Applied filters — category, location, price band. This is where intent lives.
• Click ID — a unique identifier you can match against a later conversion.

Position and component are the two people skip, and they are the two that make the data actionable. Knowing a provider gets clicks is interesting; knowing it gets clicks only from position one is a finding.

How to implement it without losing clicks

The technical risk is that your tracking call does not finish before the browser navigates away. Older implementations fired an asynchronous request on click and lost a meaningful share of events.

1. Use a redirect endpoint on your own domain. Record the click server-side, then issue a 302 to the destination. This is the most reliable method and it survives ad blockers.
2. If you must track client-side, use the beacon transport so the request is queued by the browser and completes after navigation.
3. Never delay navigation with a timeout to "let tracking finish". It costs conversions and users notice the lag.
4. Keep the redirect fast and cacheable at the edge. A tracking hop that adds 300 ms is a tracking hop that costs clicks.
5. Make the redirect endpoint noindex and disallow it in robots.txt so crawlers do not follow every outbound link.

Connecting clicks back to revenue

A click count without a revenue number is a vanity metric. The join key is the click ID: generate it at the redirect, pass it through in the network's sub-ID parameter, and read it back when the conversion report arrives.

Once that loop closes you can compute earnings per click at the level that matters — per page, per position, per component — and that is what turns reporting into decisions. A listing with a high click rate and low EPC is costing you the slot it occupies.

Five mistakes that break reporting

• Counting bot clicks. Filter known crawlers at the redirect, or every comparison table will look wildly popular.
• No deduplication. Double-clicks and back-button retries inflate counts. Dedupe on click ID within a short window.
• Tracking the URL instead of the entity. URLs change with campaign parameters; provider IDs do not.
• Ignoring position. Without it you will credit a provider for performance that belongs to the slot.
• Letting the redirect be crawlable. It pollutes analytics and wastes crawl budget.

Summary

Track the click that leaves, not just the page that loads. Capture destination, source, position, component and intent. Use a server-side redirect so nothing is lost in the navigation, and carry a click ID through to the conversion report so every click has a value attached. Once that is in place, comparison-page optimisation stops being guesswork.