The "lost clicks per month" metric: how to prioritize SEO fixes by money left on the table

Lost clicks per month tells you where your SEO opportunities are; it doesn't tell you which one to fix first. To prioritize by the money left on the table, convert each page's lost clicks into recoverable revenue — lost clicks × the page's own revenue per click × a realistic recovery rate — then rank the whole backlog by that dollar figure. Do that and a product page bleeding 380 clicks a month will often outrank a blog post bleeding 2,900, because one click on a buying query can be worth twenty on an informational one.

We ran into this every time we imported a store's Search Console export into Contexta: the pages with the scariest lost-click counts were almost never the pages worth fixing first. Sorting by clicks pushed long, high-traffic guides to the top and quietly buried the product and category pages that pay for themselves. The correction wasn't a better click metric — it was translating clicks into revenue before ranking anything.

What is "lost clicks per month" and why can't you prioritize by it alone?

Lost clicks per month is the number of monthly clicks a page is leaving on the table: its impressions multiplied by the gap between the CTR you'd expect at its position and the CTR it actually earns. It's an excellent unit for finding underperformers, and the exact way to compute it lives in the workflow for fixing CTR losers. It's a poor unit for deciding order, because it silently assumes every click is worth the same — and on a store, no two clicks are.

Take two queries. A click on "buy waterproof hiking boots" and a click on "how to clean hiking boots" each count as exactly one lost click in Search Console. One is a shopper a step from checkout; the other is a reader who may never buy anything. Ranking by lost clicks treats them as identical work with identical payoff.

That's the flaw. The metric measures attention you're missing, not income you're missing — and those two numbers point in different directions more often than most SEO reports admit.

How do you turn lost clicks into money left on the table?

Multiply each page's lost clicks per month by its revenue per click, then by a recovery rate you actually believe. Revenue per click is the page's own money divided by its own clicks — either the page's conversion rate times your average order value, or simply the page's total revenue divided by its total organic clicks over the last 90 days in your analytics. That one number is what collapses three separate problem lists into a single comparable pile.

Those lists — CTR losers, page-2 pages, cannibalization — are the ones we build straight from a raw export in turning Search Console data into ranked action lists. The money layer sits on top of all three at once, so a title rewrite, a page-2 push and a cannibalization merge finally compete on the same axis instead of living in three tabs you never reconcile.

The spread in revenue per click is the whole reason this works. Informational blog pages often earn cents per click; high-intent product and category pages earn dollars. When one input varies by 20× and lost clicks rarely varies by more than 5× or 10× across your top pages, revenue per click ends up deciding the ranking — which is exactly what you want, since revenue is what you're actually optimizing.

Why does ranking by money reorder your backlog?

Ranking by money reorders the backlog because revenue per click swings harder than lost clicks do, so it dominates the sort. Here's a store we looked at with six candidate fixes, ranked by recoverable revenue rather than by raw lost clicks — watch where the biggest click numbers land.

1. 1

   Product page — "winter parka" title rewrite

   380 lost clicks/mo · $9.80 rev/click · ~$2,600 recoverable

2. 2

   Category page — "waterproof boots", page 2 to 1

   1,240 lost clicks/mo · $2.10 rev/click · ~$1,040

3. 3

   Product page — "trail runners", cannibalization merge

   210 lost clicks/mo · $6.40 rev/click · ~$740

4. 4

   Blog post — "how to waterproof boots", CTR loser

   2,900 lost clicks/mo · $0.35 rev/click · ~$710

5. 5

   Buying guide — "boot sizing", snippet + title fix

   520 lost clicks/mo · $1.20 rev/click · ~$410

6. 6

   Blog post — "best hiking socks", page 2

   1,600 lost clicks/mo · $0.60 rev/click · ~$380

Sort that same list by lost clicks and it flips almost entirely. The "how to waterproof boots" post, with 2,900 lost clicks, sits at number one — yet by money it's fourth. The "trail runners" merge, dead last on clicks with just 210, climbs to third once its high-intent value is counted. One of those two orderings sends your next work session at a $2,600 product page; the other sends it at a $710 blog post that took the most impressions to look important.

The cannibalization case is worth calling out: merging two product URLs competing on "trail runners" recovers few raw clicks but valuable ones, which is why finding and fixing cannibalization by intent often outperforms its click count. Money surfaces those quiet, high-value fixes that a click-sorted list hides near the bottom.

What recovery rate should you assume for each kind of fix?

Never assume 100% — the recovery rate is the fraction of the CTR (or ranking) gap you'll realistically close, and it changes with the fix type and with whatever SERP feature sits above you. Assuming you'll capture the entire gap is the single fastest way to make a backlog look like money you'll never collect. Pick conservative rates, stay consistent across pages so they rank fairly, then re-measure and calibrate.

A rough, honest sense of the ranges we work with:

• Title and meta rewrites on a plain SERP tend to recover the highest share, because you're changing the click on a position you already hold — not the position itself. But if an AI Overview answers above you, the ceiling is capped no matter how good the headline, so cut the rate hard there (as of mid-2026, overviews visibly compress top-position clicks on informational queries).
• Page-2 pushes — moving positions 11–20 onto page 1 — recover less and slower, because you're moving the ranking itself, which takes 30–90 days and isn't guaranteed. The mechanics are in breaking posts out of page-2 purgatory; price these with a lower rate and a longer horizon.
• Cannibalization merges are variable: a clean merge can inherit combined signals within weeks, but you can also lose the retired page's residual clicks, so net recovery is rarely the full gap.

The number itself matters less than applying it consistently. A uniform, cautious 40% beats a per-page guess that flatters your favourite pages.

Isn't "recoverable revenue" just the "traffic value" number in Ahrefs or Semrush?

No — third-party "traffic value" prices a click at the keyword's advertising CPC, not at what your store actually earns from that click. Those tools estimate what the same traffic would cost to buy on Google Ads, which can badly over- or under-state your real revenue per click because it ignores your conversion rate, your margins and your basket size.

The mismatch cuts both ways. A keyword with a $4 CPC might convert on your store at 0.5% on a $60 order — about $0.30 of real revenue per click, a fraction of its "traffic value." Meanwhile a low-CPC branded or long-tail query can be your best converter and gets valued at almost nothing by a CPC model. Rank your own fixes on ad-market prices and you'll chase expensive keywords that don't pay you.

Traffic value is a fine competitive benchmark — it tells you what a rival's visibility is worth on the open ad market. It's the wrong input for prioritizing your own backlog, where the only price that counts is the revenue your own pages already produce per click.

How do you build this ranked backlog without doing the math by hand?

Pull impressions, clicks, CTR and position per page from Search Console, join each page to its revenue per click from analytics, apply a recovery-rate assumption, and sort — which is a spreadsheet job for a small site and a tool job once you have hundreds of URLs. The tedious part isn't the arithmetic; it's rebuilding the export, the joins and the three problem lists every month so the ranking stays current.

Contexta's Problem Map does the first half of that automatically: it imports your Search Console data and ranks every page by lost clicks per month across CTR losers, page-2 pages and cannibalization, so the biggest click opportunity already sits at the top instead of the loudest one — and, because that list is short, it doubles as a map of the small set of pages actually driving your growth. Turning that into a money ranking is the one input you bring — your revenue per click per page — layered on the lost-click backlog it builds for you. We're honest that the default sort is clicks; the revenue refinement is yours to add, because only you know what a click on each page is worth.

Either way, the discipline is the same. Find opportunities in clicks, rank them in money, assume you'll recover only part of each gap, date every change, and re-measure a cycle later. The pages worth your next hour are almost never the ones with the scariest click numbers.