Clearscope Pilot (2026): 3 Keywords on a Trial Quota — the Widest Term Coverage We've Measured, and a Grade Still on Zero
Pilot review — 3 keywords on a trial quota, tested 2026-08-20; a full run replaces these numbers when one exists.
Clearscope’s free trial allows reports and each keyword costs one, so we ran ids 1–3 of keywords-v1 and spent the quota. Two of our five tests never started. The correlation below is scored inside Clearscope’s own competitor table rather than against our frozen Google truth, so it does not sit beside the figures on the Frase and NeuronWriter pages.
Verdict
pilotClearscope works one keyword at a time. A report returns the terms the ranking pages use, in its own importance order, and a competitor table that grades the top 20 pages A+ down to F with their desktop and mobile positions. The term list is the strongest we have measured. The grade on top of it is not.
Against the terms the real Google (us) top 10 share, its top 30 reached recall@30 and precision@30 . No other tool we have run covered that much of the shared vocabulary, though no other run was this small.
The grade is another matter. Clearscope ranks the pages it grades, so we could ask whether its own grades matched its own ranking. Over 50 graded pages the mean Spearman came to +. All 3 keywords came out positive and all 3 came out small, +0.052 to +0.140. That is the easiest version of this test we run, and the grade still told us close to nothing about position.
Three keywords settle nothing. They do tell us which questions a paid run should answer first, and on this evidence, the list is the part worth paying for, not the grade.
you want the widest term coverage we have measured and a graded competitor set you can audit, and you will decide for yourself which terms belong in the draft.
you plan to write toward the grade, or $129 a month for 20 reports is more than the term list alone is worth to you.
Tested 2026-08-20 on the 14-day free trial, across 3 keywords of our frozen fixture. Every number on this page is read out of the run’s machine-readable report at build time. See how it compares on our comparison table.
Key findings
- This is a pilot. Clearscope's free trial allows 3 reports and each keyword costs one, so every number below rests on 3 keywords, run 2026-08-20. Read them as a direction, not a result.
- Its term list covered more of what ranking pages share than anything we have measured. Recall@30 0.422 and precision@30 0.201 against the terms the real Google (us) top 10 have in common, next to recall 0.298 for NeuronWriter over 20 keywords and 0.203 for Frase over 20. Different keyword sets, so not a head-to-head.
- Its content grade barely tracked the ranking Clearscope recorded itself. Across 50 graded pages, the mean Spearman between letter grade and position in its own competitor table was +0.108 (2026-08-20).
- That figure is not comparable with the ones on our other review pages. Frase and NeuronWriter were scored against the frozen Google (us) top 10; Clearscope was scored against its own SERP snapshot, which is the easier test. NeuronWriter under that same easier framing came to +0.105.
- One report buys a lot. 35–46 terms in importance order, plus a competitor table that grades the top 20 and carries desktop and mobile positions for every row. That table let us test the grade without pasting in a single page.
What works
- The widest term coverage we have measured: recall@30 0.422 across 3 keywords. Its list held about 42% of the terms the ranking pages share.
- The competitor table grades the top 20 itself and shows desktop and mobile position beside each row. You can check the grade against the ranking without scoring a single page yourself.
- 14–19 usable graded rows per keyword, thicker than the ten pages per keyword we get from the other two tools.
- The trial is real: 3 full reports over 14 days, no card on file. You can run this test yourself before paying anything.
All of it on 3 keywords. Pilot, 2026-08-20.
What doesn’t
- The grade sat on zero against the tool's own ranking: mean ρ +0.108 over 50 pages. On one keyword the pages at 3 and 4 both graded F; on another the only A+ sat at 7.
- Precision@30 was 0.201: about 20 in every 100 terms it hands you are terms the ranking pages share. The rest sit outside our top-30 truth, and sorting those is your job.
- Importance is an order, not a number. The list is ranked but carries no weight you can threshold on, so where to stop reading is your call.
- No API on the trial, and no refresh button we could find on a finished report, so a consistency test costs another report.
- The cheapest paid tier is $129 a month for 20 reports. That is the highest entry price of the three tools we have run.
All of it on 3 keywords. Pilot, 2026-08-20.
What Clearscope is
The unit of work is a report: one keyword, one country, one document. Clearscope calls it a Draft. Creating one returns three panels. The terms the ranking pages use, ordered by importance and annotated with how often those pages use them. An editor that grades your text as you write. A competitor table listing the top 20 results with a grade, a word count and both desktop and mobile positions.
The third panel is why this pilot has a correlation at all. Frase and NeuronWriter make you supply the pages you want scored. Clearscope grades the search results for you, so its grades and its ranking arrive in the same table and you can check one against the other on the spot.
Opening a finished report again, switching between the editor, the term map and the competitor table, costs nothing. Creating a report costs one from your monthly allowance. On the trial that allowance is 3.
Do its terms match what ranks?
best so farClearscope publishes a real term list, so this is the same measurement we ran on NeuronWriter rather than the substitute we had to use on Frase. We took its top 30 terms in its own importance order and matched them, on unigrams and bigrams, against the top 30 terms our ground truth pulled from the body text of the pages that actually rank. Same tokenizer, same cutoff, same ground truth as the other two runs.
About 20 of every 100 terms it hands you are terms the ranking pages share, and its list holds about 42% of those shared terms. The second figure is the one worth noticing. It is the widest coverage we have measured, on the smallest sample we have run.
| Measure | Value |
|---|---|
| Precision@30, mean over 3 keywords | |
| Recall@30, mean over 3 keywords | |
| Terms offered per keyword | 35–46 |
| Keyword | Terms offered | Terms taken | P@30 | R@30 | Hits |
|---|---|---|---|---|---|
| cost of living comparison | 46 | 30 | 0.164 | 0.333 | 10 |
| gpu comparison | 35 | 30 | 0.197 | 0.467 | 14 |
| ai voice generator | 40 | 30 | 0.241 | 0.467 | 14 |
| Mean, 3 keywords | — | — | 0.201 | 0.422 | 38 |
← swipe the table sideways for the rest of the columns
The spread is narrow for once. Recall ran from 0.333 on cost of living comparison to 0.467 on ai voice generator, against 0.100 to 0.500 for NeuronWriter over 20 keywords. With 3 keywords that could easily be luck, and only a full run settles it.
Do not read the precision figure against Frase’s as a contest. Frase publishes no term list, so its number is a stand-in built from keyword research, and this run covers 3 keywords against its 20 and NeuronWriter’s 20. Same definition, different inputs.
You can rebuild the ground truth yourself. Our Top 10 Term Finder runs the same extraction on any keyword you type: fetch the live top 10, pull the body text, keep the terms those pages share. Compare that against whatever your tool hands you.
Does the grade track rank?
negativeOn Frase and NeuronWriter we correlate the tool’s score with the real Google (us) position out of our frozen ground truth. Clearscope grades the SERP itself, and on a 3-keyword pilot we used that: its letter grades against the positions in its own competitor table. Same statistic, easier test, different answer key. The number below belongs beside no other number on this site.
We mapped the letters to numbers (A+ = 12 down to F = 0), took the desktop position from the same row, and ran a Spearman correlation per keyword. Positive means better-graded pages ranked higher. The mean across 3 keywords and 50 graded pages was +0.108.
Related: does a content score predict Google rank? That page carries this pilot as a separate block for the same reason.
| Keyword | Graded pages | ρ in its own table |
|---|---|---|
| cost of living comparison | 17 | +0.131 |
| gpu comparison | 14 | +0.052 |
| ai voice generator | 19 | +0.140 |
| Mean, 3 keywords | + |
← swipe the table sideways for the rest of the columns
The tables show why. On gpu comparison the single A+ page sits at position 7. On cost of living comparison the pages at 3 and 4 both graded F. Those two are short tool pages, a few hundred words each, and a grade built on coverage cannot explain why they rank where they do.
NeuronWriter published enough data for us to run this same generous framing on it. Its score against the ranks it crawled itself came to +0.105. Two tools marking their own homework, two answers within a hair of zero. On the harder test, against the real Google top 10, NeuronWriter came to −0.052 and Frase to -0.115. Clearscope’s quota ran out before we could put it there.
Price and quota
| Item | Recorded |
|---|---|
| Cheapest paid plan | Essentials, $129 a month for 20 reports |
| Trial | 14-day free trial, 3 reports, no card required |
| Spent on this pilot | 3 reports covering 3 keywords, $0 out of pocket |
| API | none we could find on the trial; every capture here came through the UI |
| Re-opening a finished report | free, and unlimited — only creating one draws on the allowance |
← swipe the table sideways for the rest of the columns
Prices read off the account’s own billing screen on 2026-08-20 and written up in data/reports/recon/clearscope-2026-08-20.md. This is the one figure on the page that comes from somewhere other than the run’s report, because a trial that spends nothing records no price. Check it against the vendor before you buy.
What we couldn’t test
Consistency (T3). Does Clearscope give the same page the same grade twice? We do not know. A finished report has no refresh we could find, so re-testing a keyword means creating a second report, and the 3 the trial allows were already spent on the 3 keywords above. We have this measurement for Frase and for no one else.
Factual accuracy of generated content (T5). Clearscope will write a first draft for you. We never asked it to, for the same reason: a draft costs a report. Nothing on this page says anything about what it writes.
Grade against real Google position. The headline test on our other review pages. Running it needs its competitor tables lined up against our frozen SERP truth across the full fixture, which needs more reports than a trial holds.
SERP overlap. We did not measure how much of Clearscope’s crawled top 10 matches the real Google (us) top 10, so the +0.108 above cannot be adjusted for how far its snapshot drifts from Google’s.
17 of the 20 fixture keywords. This pilot ran ids 1–3 of keywords-v1. Because the sample is this small and the answer key differs, Clearscope gets no head-to-head page against the other tools. We have nothing yet that can honestly go side by side.
How we tested
Every tool runs against the same frozen fixture and, for the term test, the same ground truth: the live Google (us) top 10 for each keyword, the body text of those pages, and the terms those pages share. Country, language and device stay constant across tools. The full procedure is on our methodology page.
The trial has no API, so we drove the app through its UI and read the term list and the competitor table off the rendered report. Before counting we dropped the rows the app marked as failed and the duplicate rows pointing at the same URL, which is why a table of 20 results yields 14–19 usable rows. Every raw capture is archived on our side, so these numbers do not depend on Clearscope keeping our account data.
The term test scores Clearscope against our ground truth. The grade test scores it against itself, and we say so everywhere the number appears. When a paid run exists, the grade goes up against real Google positions like every other tool here, and this page changes.
Last tested 2026-08-20, on the 14-day free trial, 3 keywords.
Frequently asked questions
How good are Clearscope's term suggestions?
The widest coverage we have measured, on the smallest sample we have run. We took the top 30 terms in its own importance order and compared them with the terms the current Google (us) top 10 actually share. Precision@30 came to 0.201 and recall@30 0.422 across 3 keywords, so its list covers about 42% of the shared terms. That recall is the highest we have measured, but 3 keywords is a pilot on a trial quota, not a result. Tested 2026-08-20.
Does Clearscope's content grade predict Google rankings?
We cannot answer that yet — this pilot covered 3 keywords, tested 2026-08-20 — and the number we do have points the same way as the other two tools. Clearscope grades the pages in its own competitor table, so we correlated its letter grades (A+ down to F, mapped to numbers) against the positions in that table: mean Spearman +0.108 over 50 graded pages and 3 keywords. That answer key is Clearscope's own SERP snapshot, not the frozen ground truth we use on Frase and NeuronWriter, so the figure is not comparable with theirs. On its own terms it sits on zero: the grade barely tracks the ranking the tool itself recorded.
Is Clearscope worth $129 a month?
This pilot cannot tell you. It can tell you what one report buys: 35–46 terms in importance order, and a competitor table that grades the top 20 with desktop and mobile positions beside each row. The cheapest paid tier is Essentials at $129 a month for 20 reports, checked on the account's own billing screen on 2026-08-20. The free trial runs 14 days and allows 3 reports, which is exactly what this pilot spent.
Why only three keywords?
Because the trial allows three reports and each keyword costs one. We spent 3 of 3 on ids 1–3 of keywords-v1, which left nothing for a consistency re-run or an AI draft. We publish the three we have rather than hold them back, and we label every number on the page as a pilot. A full run on the shared fixture replaces them.
How did you test Clearscope?
Against the same frozen fixture and the same ground truth as every other tool here: the live Google (us) top 10 for each keyword, the body text of those pages, and the terms those pages share. Clearscope has no API on the trial, so we drove the app through its UI and read the term list and the competitor table off the rendered report. T1 is scored against terms-truth-v1. T2 is scored inside Clearscope's own table, and we say so wherever that number appears. Full method on our methodology page.
We have no affiliate relationship with Clearscope and have not applied for one, so nothing on this page earns us a commission. This pilot ran on the free trial and cost us nothing. Outbound links to Clearscope are marked nofollow. Where a page on this site does carry an affiliate link, it says so at the top. A commission cannot change a test result: the fixture is frozen, the extraction is scripted, and the raw captures are archived. More on who we are and how we make money.
Alternatives we tested
Other tools in this category we have run against the same fixture and the same ground truth. Where two runs covered different keywords, the comparison page says so instead of averaging across them.
- Frase — A research-to-draft pipeline with no term list, and a content score that did not predict Google rank in our sample. Tested 2026-08-18
- NeuronWriter — A weighted term list and a SERP that tracks Google's closely, under a content score that did not predict rank in our sample. Tested 2026-08-21
- Scalenut — A weighted term list and a scored competitor table, under a content score that did not track rank against Google or against its own SERP. Eight-keyword run. Tested 2026-08-21
- Surfer SEO — A full 20-keyword run: its content score did not track Google rank under either URL match, and a day-two re-run swapped terms on three of five keywords while the scores held still. Clearscope vs Surfer SEO Tested 2026-08-24
All 5 on one page, a row per tool and a column per measurement: best content optimization tools.