The free lessons are the part of these tools people come back to. From October 12 through November 5, 2026, 117 browsers opened two or more MechanicRank Learn lessons, 754 lessons in all, across 46 countries. The quick calculators on clickflame.com logged 8 runs, and 6 of those came in one afternoon. The two sit on different sites with different traffic, so this is a usage log and not a test. If you plan to build a free tool, count second visits and real use, not page views.
What Changed
- MechanicRank Learn logged 1,144 lesson opens from 507 browser IDs in 62 countries, but 239 of those IDs opened one lesson and never did anything else.
- The stronger number is 117 repeat readers, 754 lessons and 46 countries, with 55 of the 117 in the U.S.
- 81 browsers used Study Mode 210 times, in 28 countries.
- The Clickflame SEO course shrinks from 196 IDs to 14 once one lesson IDs from China come out, and 4 browsers finished 13 lessons.
- The quick tools drew 11 to 22 page views each and 8 logged runs, most of them likely tests. Two of them can't log a run as built.
- A lesson has a next lesson. A calculator gives one answer and it's done. That is the best guess for the gap, not a proven cause.
Two Kinds of Free Tool, Two Very Different Counts
Put a free lesson library next to a free calculator and the gap is hard to miss. From October 12 through November 5, 2026, the study guides on MechanicRank Learn logged 1,144 lesson opens. The quick tools on clickflame.com logged 8 runs.
Both counts come from Clickflame's own usage log. A small script on each page sends one line when a page loads or a tool gets used. It keeps a random ID for each browser and no personal details. This count runs on Pacific time from October 12, the first full day of the log.
So the lessons won, by a mile? Not so fast. That first number is the weakest one in the log. If you're thinking about building a free tool for your own business, the way you get from that number to a real one matters more than the tool.
The Raw Count Flatters the Result
The raw MechanicRank line reads 1,144 lessons from 507 browser IDs in 62 countries. Then look at what those IDs did. 194 from China and 45 from Singapore opened exactly one lesson each. Not one of those 239 opened a second lesson. Not one used Study Mode. They turned up around the clock, most days of the month.
A log can't prove what those visits were. Some could be people. One lesson and gone, 239 times in a row, looks automated, so they don't count as readers here.
This isn't a strange edge case. Imperva, which sells bot protection, says automated traffic made up more than 53% of web traffic in 2025. Google Analytics drops known bots on its own, and it won't show you how many it dropped. A homemade script gets no such filter. You have to do that work yourself.
The same check hits the SEO course harder. Its raw line says 271 lessons from 196 IDs. Take out 182 one lesson IDs from China and 14 browsers are left, with 89 lessons in 4 countries. 4 of those browsers finished 13 lessons. That's a start. It isn't a crowd.
| Free tool | Raw count | After one lesson IDs come out | Deeper use |
|---|---|---|---|
| MechanicRank Learn | 1,144 lessons, 507 IDs, 62 countries | 905 lessons, 268 browsers, 60 countries | 117 repeat readers, 81 in Study Mode |
| SEO course | 271 lessons, 196 IDs, 5 countries | 89 lessons, 14 browsers, 4 countries | 13 lessons finished by 4 browsers |
| Quick tools | 119 page views, 60 IDs, 8 countries | Not split, the counts are too small | 8 logged runs |
What the Repeat Readers Did
The number that holds up is the second lesson. 117 browsers opened two or more MechanicRank Learn lessons, and those browsers account for 754 of the 1,144 opens. They came from 46 countries. 36 opened five or more, 17 opened ten or more, and 27 came back on a different day.
Count the second visit, because the first one can be anything.
Study Mode tells the same story from another side. It's where a reader drills the key points from a lesson. 81 browsers used it 210 times, in 28 countries. None of the 239 one lesson IDs touched it, and 59 of the 81 are also repeat readers.
The worldwide part is real, and it's also thin. 55 of the 117 repeat readers and 478 of their 754 lessons were in the United States. The other 62 browsers are spread out. One browser in Gambia opened 34 lessons. Two in Sri Lanka opened 21, two in Mexico opened 20 and two in Turkey opened 17. Many countries on the list have one or two readers.
Engine repair drew the most repeat use, at 171 lessons from 37 of those browsers. Diagnostics came next, at 91 lessons from 20. That is the part of this log worth being proud of. Readers are working through compression tests and scan tool modes for free, on their own time.
Keep the limits in mind. One person on a phone and a laptop counts as two browsers. Clearing a browser makes a new ID. Test visits from the team were not filtered out, so the real figures could be somewhat lower.
The Quick Tools Have Small Numbers So Far
The quick calculators and checkers have not caught on yet. Nobody should read these numbers as anything else. The free tools page is live, with more on the SEO tools hub. People do land on the tool pages. Very few have logged a run.
| Page | Page views | Browser IDs | Logged runs |
|---|---|---|---|
| Local SEO page grader | 22 | 20 | 4 |
| Missed call calculator | 20 | 19 | 1 |
| Free tools hub | 19 | 19 | Not a tool |
| Review link generator | 18 | 18 | 2 |
| Free SEO audit page | 16 | 14 | Can't log one |
| Schema generator | 13 | 12 | 1 |
| Google search preview | 11 | 11 | Can't log one |
Even these counts need care. A run gets logged the moment someone types into or submits a form in the main part of the page, so it means someone started, not that anyone got an answer. The audit page sends you to another site to run the audit, so it never logs one. The search preview has fields but no form, so it can't log one either. Zero on those two means not measured. A new Full Site Audit went live late on November 5, so it isn't in this table. A few views also landed one second apart across different tools, which looks like a scan and not a reader.
The 8 runs may not be strangers. Only 2 landed before the afternoon of November 5. The other 6 came that afternoon, the same day Clickflame was building a new audit tool on these pages. All 8 came from the U.S., and all 8 came within 4 seconds of the page loading.
Three of the grader runs landed in under a tenth of a second. Nobody types that fast. The grader can open from a link with the address filled in and run itself, which fits those three. So the count of strangers using these tools could be very small. The log can't say how small.
Why the Lessons Pulled Ahead
This wasn't a fair fight, so don't read it as a clean test. MechanicRank Learn lists 13 modules, 189 skill areas and 780 concepts. That's a lot of pages for a search to land on. The quick tools are six pages on a different site with a different crowd. More doors bring more visitors, whatever is behind them.
Three more guesses fit the pattern. None of them is proven.
- A lesson has a next lesson. A calculator gives you one answer and you're done. There's no reason to come back tomorrow.
- A reader with a goal stays longer. The study guides pair with a certification exam. Someone working toward that has weeks of reasons to return.
- The audience differs. A person learning a trade wants to read. A shop owner checking a number wants the number, and may not want to type in their own figures.
Don't expect most learners to finish, either. An analysis of MIT and Harvard online courses found 3.13% of all participants completed in the 2017 to 2018 school year, against 46% of those who paid for a verified track. Those are big university courses, so they aren't a direct match. They do show why 13 finished lessons in 25 days isn't a shock. Jakob Nielsen's old 90-9-1 rule for online communities points the same way. Most people only look.
The lesson so far is plain. Free teaching is what people have kept coming back to in this log, while a quick answer has drawn only a look. That could change as the tools get more visitors, and a later count will show it.
What to Count Before You Build a Free Tool
A free tool can eat weeks of build time. You can run this same check on your own site for free.
- Count a use, not a view. Log the moment someone gets a result. A page view only proves the page loaded.
- Test the count on every tool. Run each one yourself and watch for the line in your log. A tool that can't report looks exactly like a tool nobody uses.
- Group by ID. Sort every ID by how many actions it took. The ones with one action are your first suspects, not your audience.
- Look for a pattern in the one hit IDs. Check country and time of day. Then decide what you won't count, and write that rule down so next month's number means the same thing.
- Pick one deep action. For a course it's a second lesson or a study session. For a calculator it could be a saved result or a second run. Report that number first.
- Give the tool a next step. Ask what the person does after the answer. If the plain reply is nothing, you've built a page that ends the visit.
- Say what the count can't show. These 25 days say nothing about who hired anyone, bought anything or passed an exam. Don't let a usage number stand in for revenue.
The free lessons at the SEO course and MechanicRank Learn are open if you want to see what people are reading. The quick tools are open too, and every real run helps the count.
Evidence
Every number in this story comes from one of these. If a figure is ours, the paragraph says so.
- Imperva, Bad Bot Report 2026 (April 29, 2026)
- Google Analytics Help, Known Bot Traffic Exclusion
- Inside Higher Ed, Why MOOCs Didn't Work, in 3 Data Points (January 15, 2019)
- Reich and Ruipérez-Valiente, The MOOC Pivot, Science (2019), MIT repository record
- Nielsen Norman Group, The 90-9-1 Rule for Participation Inequality (October 8, 2006)
- MechanicRank, Learn Auto Repair Free
Clickflame builds the websites and runs the local search and ads for service businesses that live on the phone ringing. See the free tools, look at the work, or pick a time and we will look at your numbers with you.
Help Desk Answers
Do free tools work as marketing?
Some do and some don't, and the page view count won't tell you which. In this log, free lessons brought people back and quick calculators mostly didn't. The log covers 25 days on two different sites, so treat it as one data point.
How many people use the free tools?
The closest fair answer is 117 browsers that opened two or more MechanicRank Learn lessons, plus 14 browsers on the SEO course, through November 5, 2026. A browser is not always one person, and team test visits were not filtered out.
Why take out the one lesson visits from China and Singapore?
Because all 239 IDs opened exactly one lesson and none ever did anything else. Some could be real people. The pattern looks automated, and leaving them in would make the reach look bigger than the log can support.
How do I tell bots from real readers in my own usage log?
Group events by ID and look at what each one did after the first hit. Look for IDs with one event, all from one place, spread evenly across the day. Then count an action a bot has no reason to take, like saving progress or starting a quiz.
Are the quick calculators and the courses free to use?
Yes. The quick tools carry no charge and the SEO course is listed as free on its own page. MechanicRank Learn says no account and no paywall on its own page.
Is 25 days long enough to judge a free tool?
It's long enough to spot a broken count and a big gap. It isn't long enough to call a winner. Small tools can take months to pick up search traffic, so run the same count again before you cut one.
