I Gave Away 18 Free Tools (And Wrote Rules So They Can't Lie to You)

TLDR

There are 18 free tools on this site now: calculators, live checks on your real domain, a speed-to-lead stopwatch, and a couple I do by hand. A lot of “free tools” on agency sites are lead magnets with made-up numbers, built to scare you into a call. I wanted the opposite. So every tool runs on three rules: every outside number has a real, dated source, every assumption sits where you can see and change it, and the tool is allowed to tell you you’re fine. This is how they got built, the research day where seven AI agents argued about my site, what we deliberately didn’t build, and what I honestly still don’t know about whether any of it works.

Why a Web Designer Gives Away Calculators

The honest version is printed at the bottom of the tools page: “some of you will find a problem worth hiring me to fix.” That’s it. No drip campaigns, no bait-and-switch.

But a free tool only earns trust if the math survives someone checking it. The fastest way to lose a prospect is to hand them a scary number that falls apart the moment they think about it. So the real question wasn’t “what tools should I build.” It was “what tools can I build honestly.”

A good free tool does three things:

1. Answers a real question a business owner is already asking, like “what’s my slow site costing me” or “do I actually own my domain”
2. Uses their own numbers, not an industry average they don’t recognize
3. Leads to exactly one next step, the one offer that fixes the problem it found

The Day Seven Agents Argued About My Website

In September I ran what I call a 100x on the site: seven AI research agents in parallel, each looking at the tools from a different angle, then six builder agents to ship what survived. I told them to steal whatever made sense from the tools on my other product and look for what was missing.

The interesting part wasn’t where they agreed. It was the tally.

• Five of seven said every tool should end on the one paid offer that fixes what it found, not generic “contact us” buttons.
• Four of seven said every tool should open from a link with a specific business’s numbers already filled in, so I can send a prospect their own result.
• Five of seven noticed that one of my offers, event websites, had no tool feeding it at all.
• Three of seven said not to build llms.txt or schema generators, because Google’s own documentation says no special files or markup are needed for its AI features.

And one agent dissented hard. It argued that star-rating, missed-call, and ticket-fee calculators are things software vendors give away to sell their software, so building them was pointless. We built all three anyway, on a different bet: they aren’t SEO plays against the vendors, they’re honest, sendable results for one specific business. The write-up names the dissent and says it plainly: if nobody uses them, that agent was right.

The four releases that came out of it all merged in a little over two hours that afternoon.

Every Number Needs a Receipt

Here’s the rule that makes the tools worth anything: an outside statistic can only appear on the site if it’s in a verified registry, with the source and the year. When a primary source can’t be found, the stat goes on a separate list, the folklore list, and it never ships.

That folklore list is humbling, because a lot of it is stuff every marketing site repeats. The widely quoted “1 second of delay costs 7% of conversions” couldn’t be traced to a primary source, so it’s on the list. A famous research article about responding to leads within the hour is real, but the numbers people quote from it disagree with each other by about ten times, so we don’t quote a number from it until someone reads the full paper.

What survives is the real stuff: the MIT and InsideSales study on lead response time from 2007, Google’s 2017 mobile speed research, and BrightLocal’s 2026 review survey. Each is cited on the page it’s used, with the year.

Then we turned the same scrutiny on my own tools, and found three that were overstating:

• The Missed-Lead Calculator used reach percentages for each response speed that weren’t from any study. They were my estimates. Now they’re a labeled slider, and the page says so in plain words.
• The Slow-Site Cost Calculator was counting every lead as a sale, which made its answer about three times too high. It has a close-rate input now.
• The Automation Score had a label that confused hours a task takes with hours you’d get back. Now it shows both, separately.

Each fix made a tool less impressive. Each one made it more worth sending.

Tools That Tell You Not to Hire Me

This is my favorite part, and some agency owners will think I’m crazy.

Rent vs Own compares your monthly platform bill against a one-time custom build with no hosting fees. If you’re on a cheap plan, it tells you to stay where you are. The break-even for my full build only makes sense once you’re paying more than about $125 a month.

The Ticket Fee Calculator compares nine ticketing platforms against an owned site with Stripe, using fee data read from each platform’s own pricing page. When Ticket Tailor, Tito, or Zeffy come out close to or cheaper than a custom checkout, it says so.

The Missed-Lead Calculator says it right in its FAQ: “If the calculator says you’re fine, believe it.”

Honest verdicts cost me some leads. That’s the design. A tool that always finds a problem is an ad. A tool that sometimes says “you’re fine” is one you’ll believe the day it says you’re not.

Built to Be Sent, Not Just Found

Here’s something the research found that changed how I think about these: the older tools weren’t ranking in search for anything. Most of the tool names get almost no searches at all.

So the tools aren’t mainly SEO pages. They’re outreach assets. Every tool can open from a link with a specific business’s numbers already filled in. When I reach out to someone, I can send them their own result, not a pitch. Every result has a “copy a link” button, and those links are tagged so I can tell when one gets opened. Every result ends with a next step, and every tool has one button to the one offer that fixes what it measured.

That same research pass caught something embarrassing in the booking flow: on a phone, the embedded calendar could only reach 9 of 36 available time slots. Phones now open the booking page directly. Nine of thirty-six. That’s the kind of leak nobody finds until someone actually looks.

What We Didn’t Build

The list of things we said no to is as useful as the list we built:

• llms.txt and schema generators. Google’s documentation says no special files are needed for AI search features. Developers search for these tools. Business owners don’t.
• Another DMARC checker or website grader as an SEO play. Big brands give these away and hold the top results. We did build a live email deliverability check, but as an outreach tool: “I ran your domain, and you have no DMARC record.”
• An accessibility scanner. The FTC ordered a company in that space to pay $1 million over its compliance claims. Not a claim I want to make with a free widget.
• Benchmark-style tools with industry averages. When we stripped out every number we couldn’t source, there was nothing left.
• A no-show cost calculator. The people searching for it are mostly medical practices, not the businesses I build for.

The Newest Three: Speed to Lead

The latest additions are all about the first minutes after someone reaches out, because that’s where most service businesses leak.

• Speed-to-Lead Test: send your own business a test inquiry the way a customer would, start the clock, and stop it when a real reply lands. An automatic “we got your message” doesn’t stop the clock, because it doesn’t give the lead a next step. The timer keeps running if you close the tab, and you can time a competitor the same way.
• After-Hours Gap: mark the hours someone actually answers on a map of your week. It shows how many of your leads land in the gap and how long they wait. With Monday to Friday, 9 to 5, a lead who reaches out Friday at 5 PM waits until Monday at 9 AM: 64 hours.
• First-Reply Builder: four ready-to-paste first replies, a form auto-reply, a missed-call text, an after-hours text, and a same-day follow-up, each checked for a real next step and for whether it fits in a single text.

All three run on the same rule as the rest. The only outside numbers they use are the ones in the registry, like the MIT finding that web leads called within 5 minutes were 100 times more likely to be reached than leads called at 30.

What I Don’t Know Yet

Here’s the honest part.

When the research day wrapped, I asked for a confidence split, and I think it’s still right: about 85% confident that the tools work exactly as described, because each one was tested and the math was redone by hand. About 40% confident they’ll win clients, because that depends on things the tools can’t control.

The biggest unknowns:

• The site had no analytics when these shipped. So nothing about which tool earns a call could be measured yet.
• Traffic is the real bottleneck. A great tool nobody sees is a great tool nobody uses. The first twenty prefilled links I send are the real test.
• Prices change. Fee tables and platform prices go stale, which is why the registry has a re-check date on it. A tool that quotes last year’s Eventbrite fee undercuts the whole point.

I’d rather tell you that than pretend I’ve got a conversion rate I don’t have.

Steal This for Your Business

You don’t need eighteen tools. You need one good one. Here’s the recipe:

1. Pick the question your best customers ask before they buy. “What will this cost me?” “Do I need this?” “Am I behind?”
2. Answer it with their numbers. Industry averages are wallpaper. Their own numbers are a mirror.
3. Source every outside number, or leave it out. Keep a folklore list of the stats you couldn’t verify, and never use them.
4. Put every assumption on the screen, where the visitor can change it.
5. Let it say no. The tool that sometimes says “you’re fine” is the one people trust.
6. Make it sendable. A result someone can open from a link with their numbers filled in is worth more than any page ranking you’ll get.

I appreciate you reading through how these came together. The best way to judge them is to run one on your own business. Start at the tools page, and if one of them finds a problem worth fixing, book a call from the home page.

Build first, learn fast.