I spent about a year paying for keyword tools before I figured out I already had most of what I needed sitting in Search Console. Finding low-competition affiliate keywords doesn’t actually require Ahrefs or Semrush — it requires knowing which reports in GSC to dig into and which numbers to ignore. This isn’t a “GSC is secretly better than paid tools” post, because it isn’t. It’s a “here’s how to squeeze real opportunity out of free data” post.
Quick Answer
- Pull the Queries report in Performance, filter by impressions 10–500 and position 8–20
- Those are pages Google already ranks for something — just not well enough yet
- Cross-reference with Pages to see which posts are pulling in “almost” traffic
- Click-through rate below average for that position usually means the title is the problem, not the keyword
- Repeat monthly — GSC data shifts fast, especially after Google’s broader updates
Why This Works (and Why Most Keyword Tools Don’t)
Most keyword research tools estimate competition using backlink count and domain authority averages across the top 10 results. That’s a guess. GSC tells you what’s actually happening with your specific site, for queries Google has already decided you’re somewhat relevant for.
There are a few reasons this matters more for affiliate content specifically:
- Affiliate niches are saturated with content, not links. A lot of “high competition” scores from tools come from raw link counts, but in product-review and comparison spaces, thin AI-generated content with weak on-page structure ranks anyway. GSC shows you where you’re already close, regardless of what a Domain Rating score says.
- Long-tail buyer-intent queries rarely show up in keyword tools at all. Things like “best budget tripod for hiking under 5 lbs” might get zero search volume in a paid tool’s database, but show up in your Queries report with real impressions, because someone actually typed it.
- Position 8–20 queries are the cheapest wins on the internet. You don’t need new content. You need to fix what’s already half-working. Tools that estimate difficulty don’t know your existing rankings — GSC does.
And here’s the part that surprised me when I first started doing this seriously: a chunk of my best-converting affiliate pages were never built around the keyword that actually drove sales. They ranked for something adjacent that Google decided was relevant, and the conversion data backed it up. So, no, you can’t always plan for this. But you can find it after the fact, which is its own kind of keyword research.
Common Scenarios Where This Helps
- You’ve published 50+ product comparison posts and have no idea which ones are “almost” ranking
- You’re in a niche where every keyword tool says everything is “high difficulty”
- You just got hit by a core update and need to know which pages dropped from page 1 to page 2 (these are gold — they already proved they can rank)
- You’re trying to expand an existing pillar page without guessing at subtopics
Step-by-Step: Finding the Keywords
Step 1: Set the Right Filters in Performance
Open Search Console, go to Performance > Search Results. Set your date range to the last 3 months — anything shorter and you won’t have enough data, anything longer and seasonal stuff muddies the picture.
Then filter:
- Impressions: greater than 10 (filters out noise) but less than 500 (filters out queries you’re already winning)
- Average Position: 8 to 20
This isn’t a perfect science — your mileage may vary depending on niche size. A huge site might want to push the impression ceiling up to 1,000.
Step 2: Sort by Impressions, Not Position
This is the part people get backwards. A query at position 9 with 15 impressions a month isn’t worth your time. A query at position 14 with 800 impressions is. Sort the table by impressions descending, then scan down for anything sitting outside the top 10.
Step 3: Cross-Reference with the Pages Tab
Switch the dimension to Pages instead of Queries for the same filtered range. Find pages with decent impressions but a CTR noticeably below what’s typical for their position (roughly 2–3% at position 10, dropping fast after that). Low CTR at a mediocre position usually means your title tag isn’t doing its job — not that the keyword is hopeless.
Step 4: Check Query-to-Page Pairing
Click into a specific page, then look at which queries are driving its impressions. Sometimes you’ll find a page ranking for three or four related terms you never deliberately targeted. Pick the one with the most commercial intent (anything with “best,” “vs,” “review,” “price,” or “buy” tends to convert better in affiliate content) and build that into your H1 and intro if it isn’t already there.
Step 5: Filter by Country if You’re Running International Affiliate Links
If you’re monetizing through Amazon Associates or region-specific programs, add a country filter before doing any of the above. A query performing well globally but poorly in your target country (say, the US for Amazon US links) isn’t actually a good opportunity, even if the blended numbers look fine.
Comparison: Signal Reliability for Affiliate Keyword Decisions
| Signal | Where to find it | Reliable for affiliate intent? | Notes |
|---|---|---|---|
| Impressions in 8–20 range | Performance > Queries | Yes | Best raw signal for “almost ranking” |
| CTR vs. position benchmark | Performance > Pages | Mostly | Skewed if title already has clickbait |
| Average position trend (30 vs 90 days) | Performance > Queries (compare) | Yes | Catches recent algorithm shifts early |
| Paid tool “keyword difficulty” score | External tool | Not really | Doesn’t account for your existing rankings |
What Actually Worked For Me
So my first attempt at this was way too manual. I was exporting CSVs from GSC every week and trying to spot patterns in a spreadsheet, which took hours and told me almost nothing useful — mostly because I was filtering by position alone and ignoring impressions entirely. That’s not the move. A query at position 6 with 4 impressions a month isn’t a keyword opportunity, it’s noise.
What actually moved the needle was switching to the impressions-first filter I described in Step 2, and then — this is the part I almost skipped — going back through old posts that had dropped in rankings after a core update instead of only looking at posts that were climbing. Those dropped posts already had topical authority Google recognized at some point. A coworker mentioned almost as an aside that “recovered” pages often convert better than brand-new ones because they’ve got some age and internal link equity already, and that turned out to be true for two of my biggest affiliate earners this year. I didn’t plan that discovery — I just happened to be looking at the right report at the right time.
Advanced Fixes and Edge Cases
Branded query contamination. If your site name shows up in a query, GSC will sometimes blend branded and non-branded performance for similar phrasing. Exclude your brand name using the query filter (set to “doesn’t contain”) before trusting impression numbers on competitive generic terms.
Mobile vs. desktop position splits. Affiliate content often performs very differently across devices, especially for comparison tables that render badly on mobile. Filter by device before deciding a keyword “isn’t working” — it might just be working on one device and tanking on the other.
Seasonal flattening in averages. A 12-month average position can hide a keyword that spikes hard for six weeks a year (think holiday gift guides) and is dead the rest of the time. Pull a month-over-month comparison instead of trusting the default range for anything with obvious seasonal intent.
Indexing lag on new affiliate disclosures. Not entirely related to keyword research, but worth flagging — if you recently restructured affiliate disclosure blocks or added schema, give it 2–3 weeks before trusting fresh GSC data on those pages. Google’s crawl and reporting lag can make a page look like it’s underperforming when it’s really just stale data.
Prevention Tips (Keeping the Pipeline Going)
- Re-run this filter monthly, not quarterly — affiliate niches move fast and 90 days of stale data leads to chasing keywords that already peaked
- Don’t delete old “almost ranking” content just because it’s not converting yet; update it first
- Keep a running spreadsheet of position-8-to-20 queries so you can track which ones move after you act on them
- Pair this with internal linking from your highest-authority pages toward the “almost ranking” ones — it’s often a faster lever than rewriting content
FAQ
Does Google Search Console show actual search volume? No. It shows impressions, which is close but not the same thing — impressions are tied to your specific rankings, not total search demand.
How much data do I need before trusting this method? At least 3 months, ideally with no major site changes in that window. Less than that and you’re basically guessing.
Can I use this for keywords I’m not ranking for at all? Not directly — GSC only shows queries where you got at least one impression. For truly new keyword discovery you’ll still want some outside research, even if it’s just Google’s autosuggest and “People also ask” boxes.
Why do some of my position 8–20 queries have zero clicks? Usually a weak title or a snippet that doesn’t match search intent. Sometimes it’s just genuinely low volume and the math rounds to zero — check impressions before assuming it’s a CTR problem.
Is this better than paid keyword tools? Not better, different. It’s free, it’s specific to your site, and it catches things tools miss. But it can’t tell you about keywords you’ve never touched at all.
Editor’s Opinion
honestly this is one of those things that feels almost too simple once you see it working. i ignored GSC for the longest time because it felt like “just analytics” instead of a research tool, which was a mistake on my part. it won’t replace a paid tool completely if you’re doing serious competitor research, but for finding the stuff already half-working on your own site? nothing beats it, and it’s free, which is nice when you’re running a dozen sites and don’t want a dozen tool subscriptions.
