Paste your URL into four SEO checkers and you get four report cards that don’t agree. One says your site is healthy, another lists 93 broken pages, and a third can’t audit it at all. We put the same small website through Semrush, SE Ranking, Ubersuggest and Mangools on the same day, then wrote down exactly what each one reported and why the numbers differ.
Why One Website Got Four Different Report Cards
On October 21, 2026, three crawlers scored our test site 89%, 90 and 94, and three authority metrics put it at 3, 7 and 15, per our test log. Same site, same day, default settings. So which number is the truth?
None of them, on its own. Each SEO tool checks a website with its own crawler, its own list of rules and its own link index. A score is that tool’s summary of its own rules, not a grade from Google.
Our test site is a small marketing site our team runs. It’s young, it has a few dozen linking domains, and it has the kind of rough edges most small sites have. That makes it a fair stand-in if you’re a blogger, a freelancer or a small business owner who wants to know what an SEO checkup actually tells you.
This article doesn’t pick a winner among the four suites. It answers a narrower question: when you check a website, what does each report mean, and which parts can you trust?
Key Takeaways
In our October 21, 2026 test log, three crawlers scored the same small site 89% (Semrush), 90 (Ubersuggest) and 94 (SE Ranking), while Mangools had no crawler audit at all. The scores measure different rules, so compare each tool only with itself, fix errors before warnings, and use Google Search Console for real traffic numbers.
Quick Verdict: Which Checker Answers Which Question
SE Ranking crawled the most of our site, 1,000 pages during a no-card trial, while Semrush’s free account stopped at 100 and Ubersuggest’s free plan at 150, per our October 21, 2026 test log. But “most pages” isn’t the only question you might have. Pick the checker by the question you’re asking.
| Your question | Checker to use | What it showed on our site |
|---|---|---|
| What’s broken on my site right now? | Semrush (free account) | Site Health 89%, 97 errors, broken internal links on 93 of 100 pages |
| Can a tool crawl most of my site at once? | SE Ranking (14-day trial) | 1,000 pages crawled, Health Score 94, 6 errors |
| Which on-page basics need work? | Ubersuggest (free plan) | On-Page SEO Score 90, 70 issues, desktop load time 1.81 s |
| How strong is my domain compared with rivals? | Mangools SiteProfiler (free plan) | Domain Authority 3, Trust Flow 19, Citation Flow 21 |
| Can AI search crawlers reach my pages? | Semrush (free account) | AI Search Health 95%, Blocked from AI Search panel |
Want to start without signing up for anything? Each vendor has a free door, and they’re not the same size:
- Semrush’s Free SEO Checker says “You can run one check per day for free,” and points you to Site Audit for a deeper crawl (Semrush SEO Checker). We ran the full Site Audit on a free account, not the one-off checker.
- SE Ranking’s free tools page offers “free daily checks” and says “No account required!” (SE Ranking free tools). Its full Website Audit needs the trial we used.
- Ubersuggest’s home page has an “Enter a website or keyword” box with a “Search free” button (Ubersuggest). Our free plan allowed 3 searches a day.
- Mangools SiteProfiler checks authority “with the well-known SEO metrics by Moz and Majestic” (Mangools SiteProfiler). Our free plan allowed 5 site lookups per 24 hours.
If you’re choosing one suite to pay for rather than one check to run, our comparison of four SEO suites tested on the same website covers keyword data, limits and prices. This page stays on the checkup itself.
How We Checked One Website With Four Tools
We checked one website with all four tools on October 21, 2026, using one work account, default settings and no credit card, and logged 4 site checks and 4 backlink lookups in total. Each tool got the same two jobs from our shared task list: audit the site, then look up backlinks for the same domain.
Here’s how the setup differed by tool:
| Tool | Plan we used | Site check we ran | Backlink lookup we ran |
|---|---|---|---|
| Semrush | Free account (pressed “Skip trial”) | Site Audit, default settings | Backlink Analytics, root domain |
| SE Ranking | 14-day trial, no card | Website Audit (started on its own) | Backlink Checker |
| Ubersuggest | Free plan | Site Audit from a new project | Backlinks Overview |
| Mangools | Free plan | SiteProfiler (no crawler audit exists) | LinkMiner |
Why does the plan matter? Because the free plan sets the crawl size, and the crawl size changes what the report can find. A 100-page crawl and a 1,000-page crawl of the same site are different samples.
We copied every number straight from the screen and kept masked screenshots. Three things we did not do: we didn’t fix anything between checks, we didn’t turn on JavaScript rendering, and we didn’t open paid plans. When a section below explains how a score works, it quotes the vendor’s own documentation rather than guessing.
A note on money: Indie Stack Lab may add affiliate links to some tools later. When a link is an affiliate link, it is labeled. That did not change what we tested or what we wrote down.
Health Scores Compared: 89%, 90 and 94 From Three Crawls
Three crawlers scored our site within five points of each other, 89% in Semrush, 90 in Ubersuggest and 94 in SE Ranking, but they crawled 100, 150 and 1,000 pages to get there, per our October 21, 2026 test log. Close scores, very different inputs.
| Tool | Score shown | Pages crawled | What the label said |
|---|---|---|---|
| Semrush | Site Health 89% | 100 of 100 (free cap) | “Top-10% websites” benchmark at 92% |
| Ubersuggest | On-Page SEO Score 90 | 150 (free cap, then an Upgrade prompt) | “High” |
| SE Ranking | Health Score 94 | 1,000 of 1,000 | “Excellent” (recommended 90+) |
| Mangools | No health score | No crawler audit | Not applicable |
Do the scores mean the same thing? No, because each formula weighs different things:
- Semrush says Site Health “is based on the number of your total errors and total warnings that were found on the pages crawled on your site,” and that “Errors have more impact on your Site Health Score than warnings” (Semrush KB). Notices “don’t impact your overall site health score” (Semrush KB).
- SE Ranking says its score weighs each category, each issue’s priority and “the share of detected pages affected by the issue” (SE Ranking blog).
- Ubersuggest labeled our 90 “High” and put it on top of a 70-item issue list. Its site audit guide explains each check, such as the low word count rule, so read the list before the number (Ubersuggest help center).
One rule from Semrush changes how you read a crawl cap. Its Site Health “doesn’t depend on the number of pages crawled on the site, since the frequency of each error’s occurrence is what accounts for the score” (Semrush KB). So a 100-page sample can score similarly to a full crawl. It just can’t show you problems on pages it never reached.
What surprised us was SE Ranking’s 94 next to 2,230 issues. That stops looking odd once you see 2,195 of those were Notices, the lightest priority. Mangools is the odd one out: it has no crawler, so it can’t give a health score at all. Its SiteProfiler is a domain-metrics page, and comparing it with a crawl isn’t like for like.
For the full walkthrough of one of these audits, see our write-up of what a free Semrush Site Audit found. Our side-by-side test of three site audit crawlers goes deeper into crawl limits, schedules and issue categories.
What Each Checker Flagged as the Biggest Problems
The three crawlers didn’t agree on our biggest problem: Semrush led with broken internal links on 93 pages, Ubersuggest with 36 poorly formatted URLs, and SE Ranking with 6 errors out of 2,230 issues, per our October 21, 2026 test log. That disagreement is more useful than any single score.
| Checker | Top issues on our site | Issue totals |
|---|---|---|
| Semrush | Broken internal links (93 pages), low text to HTML ratio (89), low word count (36), content not optimized for AI Search (4), 4xx errors (1) | 97 errors, 159 warnings |
| Ubersuggest | Poorly formatted URL (36 pages), title too short (18), low word count (6) | 70 SEO issues |
| SE Ranking | Errors, warnings and notices by priority | 6 errors, 29 warnings, 2,195 notices |
The broken-link disagreement
Here’s the catch. Semrush flagged broken internal links on 93 of its 100 pages and counted 2 pages as Broken. Ubersuggest crawled 150 pages and reported 0 Broken. Which one is right?
Both can be. They crawl different page sets, start from different links and run different checks. A broken link on a page, which Semrush counts, is not the same as a broken page, which Ubersuggest’s status count shows. When nearly every crawled page shares one link problem, the cause often sits in a shared part of the template, such as a menu or a footer. We didn’t open the detail list in this round, so that’s the first thing we’d check next.
Content warnings need judgment
Low word count showed up in both Semrush (36 pages) and Ubersuggest (6 pages). Ubersuggest says it “flags pages with fewer than 200 words as having a low word count,” and that pages like “Contact Us” or “About” often trip it without hurting SEO (Ubersuggest help center). Should you pad those pages? Usually not. Decide page by page whether more text would help a reader.
Ubersuggest’s guide also lists limits worth knowing before you trust a clean result: it audits HTML pages only, it won’t follow a nofollow internal link, and firewalls or Cloudflare can mistake its crawler for scraping. A short report can mean a healthy site, or a crawler that got stopped early.
Reading SE Ranking’s 2,195 notices
SE Ranking sorts issues into Notice, Warning and Error, and its blog says “More critical problems have more influence on the final score than minor issues” (SE Ranking blog). Six errors is a short list. Start there, and treat the notices as a backlog, not an emergency.
Authority Scores: Why the Same Site Scored 3, 7 and 15
Our site scored 3 on Moz’s Domain Authority, 7 on Semrush’s Authority Score and 15 on SE Ranking’s Domain Trust, all on October 21, 2026, per our test log. Every one of those is a 1-to-100 scale. So why the spread?
| Metric | Vendor | Where we saw it | Our score |
|---|---|---|---|
| Domain Authority | Moz | Mangools SiteProfiler and Ubersuggest | 3/100 (both) |
| Page Authority | Moz | Mangools SiteProfiler | 26/100 |
| Authority Score | Semrush | Backlink Analytics | 7, “Low authority” |
| Domain Trust | SE Ranking | Backlink Checker | 15/100 |
| Page Trust | SE Ranking | Backlink Checker | 7/100 |
| Citation Flow / Trust Flow | Majestic | Mangools SiteProfiler and LinkMiner | 21 / 19 |
The two 3s match for a simple reason. Ubersuggest says it sources backlink data “via the Moz API,” including “Domain Authority and Page Authority” (Ubersuggest help center), and SiteProfiler shows Moz and Majestic metrics. Two tools, one source.
The other scores use different recipes:
- Semrush Authority Score is built from “Link Power,” “Organic Traffic” and “Spam Factors (Natural Profile),” and Semrush says it “is best used for domain comparison, and not for determining good/bad on an absolute scale” (Semrush KB).
- SE Ranking Domain Trust assesses “both the number and quality of backlinks and referring domains, using PageRank principles,” and SE Ranking calls it “a comparative metric, meaning its value depends on your niche and competition” (SE Ranking).
- Moz Domain Authority “predicts how likely a website is to rank” compared with competitors, and Moz adds that it “is not a Google ranking factor” (Moz).
- Majestic says Trust Flow “represents the authority of a link” and Citation Flow “represents the popularity of a link,” and that the two “are designed to be used together” (Majestic FAQ).
Notice that Semrush folds organic traffic into its score, while Domain Trust and Domain Authority lean on links. A young site with a few dozen linking domains and little search traffic will land low in a traffic-weighted score. That’s a likely reason our 7 sat below our 15, though no vendor publishes enough of its formula for us to prove it.
So what do you do with your number? Look up the two or three sites you actually compete with in the same tool, and compare. A 7 next to rivals at 10 is a small gap. A 7 next to rivals at 50 says links come before more content. Our explainer on why a young site’s Domain Authority starts low and how DA is built covers the DA side in more depth.
Backlink Counts for One Domain: 1,860 to 2.7K
Four backlink tools found between 40 and 77 referring domains and between 1,860 and 2.7K backlinks for the same domain on the same day, per our October 21, 2026 test log. The referring-domain counts sat close together, except one.
| Tool | Referring domains | Backlinks | Extra detail on the screen |
|---|---|---|---|
| Semrush Backlink Analytics | 46 (+12%) | 2.7K | “Live Update: 1 backlink found today” |
| SE Ranking Backlink Checker | 77 | 2.2K (2,235 analyzed) | 61 domains new in the last 30 days |
| Ubersuggest Backlinks Overview | 48 | 1,860 | 1,618 of them No Follow |
| Mangools LinkMiner | 40 | 2.4k active | “One link per ref. domain” filter on |
Why don’t they match? Each tool crawls the web with its own bot and keeps its own index. Ubersuggest, for example, says its report “displays a selection of the most relevant backlinks to a website based on each link’s quality and traffic potential” and that it updates backlink data monthly (Ubersuggest help center). A selection updated monthly will differ from an index that logged a new link that morning.
Which number should you watch? Referring domains, not backlinks. Our site had roughly 1,900 to 2,700 backlinks from only 40 to 77 domains, depending on the tool. That means a handful of sites link to us many times each, which inflates the backlink total. The domain count tells you how many different sites vouch for you.
Ubersuggest’s nofollow split is worth a second look, too. 1,618 of its 1,860 backlinks carried nofollow in our lookup. That’s not a problem to fix. Profiles, comments and directories often add nofollow by default, so the split mostly tells you where links come from.
SE Ranking’s 77 stood apart, and its own numbers hint at why: 61 of those domains were marked new in the last 30 days. It may simply have picked up recent links faster. We can’t confirm that from one lookup. For tools built only for this job, see our comparison of four backlink checkers on one domain, and for a zero budget, our roundup of what free backlink checker plans actually show.
Traffic Estimates: What the Tools Think the Site Gets
Semrush estimated our site at 21 organic visits a month and Ubersuggest at 79, per our October 21, 2026 test log. Both are models, not counts. Neither tool had access to our analytics.
| Tool | Traffic figures on the screen | Keywords figure |
|---|---|---|
| Semrush (Backlink Analytics summary) | Monthly Visits 329, Organic Traffic 21 | Not shown on that screen |
| Ubersuggest (Site Audit summary) | Organic Monthly Traffic 79 | Organic Keywords 36 |
| SE Ranking | Not checked in our test | Not checked |
| Mangools | Website Popularity Rank N/A, “Not enough history data.” | Not shown |
How are those estimates made? Both vendors describe the same basic recipe: take the keywords a site ranks for, apply an expected click-through rate for each position, and multiply by search volume.
- Semrush says its organic traffic “is estimated based on a website’s total keyword positions and the click potential of every position” (Semrush KB). Its Semrush Rank page adds that its database covers “Google’s first ten results pages” (Semrush KB).
- Ubersuggest lists keyword search volume, search rankings and “Click-Through Rate by Position,” and notes it “doesn’t include referral, social, email, or direct traffic” (Ubersuggest help center).
The Monthly Visits figure of 329 on the Semrush screen is a separate, larger number than its organic estimate of 21. Treat it as one more estimate, not a count of real visitors.
Our Ubersuggest project also kept the market it picked from our browsing location, not the US. Its 79 visits and 36 keywords may describe that market rather than US search. Is that a small detail? Not if you’re comparing it with a US-based number. Check the market setting before you read any traffic estimate.
For your own site, skip the guessing. Google Search Console’s Performance report shows “The number of times a user clicked your site from Google Search results” and how many times your site appeared (Search Console Help). It’s free and it’s Google’s own data. Our beginner’s guide to Google Search Console for SEO walks through setting it up. Use the estimates above for competitors, where you have no other option.
AI Search Checks Inside SEO Checkers
Semrush’s free audit gave our site an AI Search Health score of 95%, marked beta, and SE Ranking’s project page added widgets for AI Overview, AI Mode, ChatGPT, Gemini and Perplexity, per our October 21, 2026 test log. Ubersuggest’s project setup offered an AI prompt step, which we skipped, and Mangools’ account menu listed AI Watcher limits we didn’t use.
Semrush describes AI Search Health as a score built from three inputs: AI search checks in Site Audit, AI bot access in robots.txt, and technical readiness such as “schema markup, crawlable navigation, internal links, llms.txt” (Semrush KB). Four of our pages carried the “Content not optimized for AI Search” issue.
Next to the score, the Blocked from AI Search panel listed ChatGPT-User, OAI-SearchBot, Googlebot and Google-Extended. Semrush says the widget covers “8 major AI crawlers” (Semrush KB), so our view showed half of them. We didn’t open its detail view, so we can’t say more about our own site than what the summary showed.
What matters more is the type of each bot. Semrush’s guide sorts them into groups (Semrush KB):
| Bot group | Examples Semrush lists | What it does |
|---|---|---|
| On-demand fetchers | ChatGPT-User, Claude-User, Perplexity-User | Fetch content in real time when someone asks an AI assistant |
| AI search crawlers | OAI-SearchBot, Claude-SearchBot, PerplexityBot, Googlebot | Crawl ahead of time to power AI search results |
| AI training bots | Google-Extended, GPTBot, ClaudeBot | Collect content for model training; Semrush says these “do not affect your search visibility” |
Blocking Google-Extended, then, is a content-ownership choice, not an SEO mistake. Blocking OAI-SearchBot or Googlebot is a visibility choice with real costs.
SE Ranking took a different angle. Its project setup wouldn’t launch until we entered at least one AI prompt, and the project overview then showed AI widgets next to Google rankings. Our results were still collecting when we wrote this, so we have no AI answer data to report.
Can an SEO checker tell you whether ChatGPT cites you? Not really. An audit tells you whether AI crawlers can reach and read your pages. Whether AI answers mention you is a separate measurement. Our test of four AI search visibility trackers covers tools built for that, and our guide to what generative engine optimization is explains the wider picture. If Google’s own AI answers are your main concern, our guide to how Google AI Overviews work and who gets cited covers that side, and our guide on what gets a site cited in ChatGPT covers the other.
Speed and Rendering Signals in the Reports
Ubersuggest reported a desktop load time of 1.81 s for our site and SE Ranking showed “JS Rendering Disabled” on its audit, per our October 21, 2026 test log. Neither one is a full speed test, but both are signals worth reading correctly.
Start with the load time. A single desktop figure tells you roughly how fast one load went for Ubersuggest’s crawler. It doesn’t tell you how real visitors on phones experience your pages. Ubersuggest’s own guide points to a Google PageSpeed test when a page loads too slowly for its crawler (Ubersuggest help center).
That’s where Google’s free PageSpeed Insights comes in. Google says it “provides both lab and field data about a page,” where lab data is “useful for debugging issues” and field data captures “true, real-world user experience” over the previous 28-day collection period (Google for Developers). We didn’t run it in this round, but its thresholds give you a yardstick:
| Metric | Good | Poor |
|---|---|---|
| Largest Contentful Paint (LCP) | 2,500 ms or less | Over 4,000 ms |
| Interaction to Next Paint (INP) | 200 ms or less | Over 500 ms |
| Cumulative Layout Shift (CLS) | 0.1 or less | Over 0.25 |
| Lighthouse category score | 90 or above | Below 50 |
Source: Google PageSpeed Insights documentation.
Now the rendering flag. SE Ranking explains that “Without JS rendering, auditing tools miss content, links, and interactive elements,” and that you can switch it on under the Parser Settings tab (SE Ranking blog). Semrush warns about the other side: turning on JS rendering “might affect the Health Score,” because the audit “will be able to check more content, and, therefore, find more errors and warnings” (Semrush KB).
So is your site built with a JavaScript framework? If yes, a crawl without rendering may be reading half a page. Turn rendering on, accept a lower score, and trust the new number more. If your site is plain HTML, like most small blogs, the default is usually fine.
How to Read a Website Check Without Panicking
Read a website check by fixing errors first and comparing each tool only with itself; in our October 21, 2026 test, the same site got 97 errors in one tool and 6 in another, per our test log. Here is the routine we’d follow after a checkup like ours.
- Pick one tool as your scoreboard. A Site Health of 89% and a Health Score of 94 use different formulas. Track the same tool over time instead of comparing tools.
- Fix errors before warnings. Semrush says it “is recommended that you address errors first” (Semrush KB). Within the errors, start with the issue that hits the most pages, and look for a shared cause in the template.
- Read content warnings page by page. A short contact page can stay short.
- Treat authority and traffic as context. Compare them with the sites you compete with, in the same tool. Use Search Console for your own traffic.
- Check the market and rendering settings. A wrong country or a disabled renderer can skew every number in the report.
- Re-check on the tool’s own schedule. Our Semrush audit defaulted to “Weekly, Every Friday,” SE Ranking scheduled its next update for Oct 28, 2026, and Ubersuggest says free users get “a detailed weekly report” with no on-demand recrawls (Ubersuggest help center).
How often should you run a full check? For most small sites, once a week is plenty, and that’s what all three free or trial crawlers defaulted to. SE Ranking’s help center says audits can run “weekly or monthly” (SE Ranking help center). Daily checks on a small site mostly show you the same list.
Not sure where to start with the tools themselves? Our beginner’s routine for getting results from SEO tools explains the steps from the first audit to keyword research, and our comparison of what four free SEO plans let you do lines up the free limits. For the tool-by-tool detail behind the numbers here, read our walkthroughs of what a free Semrush account includes, the SE Ranking 14-day trial screen by screen, what Ubersuggest’s free plan did on a real site and Mangools’ free KWFinder, SiteProfiler and LinkMiner.
One habit carries over to every checker: decide what you want to learn before you paste your URL. On a free plan, every check is a budget item.
References
- Semrush: Free SEO Checker
- Semrush Knowledge Base: How is Site Health Score calculated in the Site Audit tool?
- Semrush Knowledge Base: Reviewing Your Site Audit Issues
- Semrush Knowledge Base: What is Authority Score?
- Semrush Knowledge Base: AI Search Health in Site Audit
- Semrush Knowledge Base: Blocked from AI Search in Site Audit
- Semrush Knowledge Base: Semrush Rank
- SE Ranking: Free SEO tools
- SE Ranking blog: New Website Audit, Health Score and JavaScript Rendering
- SE Ranking: Domain Trust checker
- SE Ranking Help Center: How to use Website Audit
- Ubersuggest home page
- Ubersuggest Help Center: Ubersuggest’s Site Audit Guide
- Ubersuggest Help Center: Site Audit Limits for Free Users
- Ubersuggest Help Center: How Ubersuggest Traffic Estimations Work
- Ubersuggest Help Center: Ubersuggest Is Not Showing All of My Backlinks
- Mangools: SiteProfiler
- Moz: What Is Domain Authority?
- Majestic: Frequently Asked Questions
- Google for Developers: About PageSpeed Insights
- Google PageSpeed Insights
- Search Console Help: Performance report (Search results)