This page is the thing it describes. You arrived on an ad. You are being handed hours of our actual thinking for five dollars. Nobody has asked you for a call. That is the whole mechanism — and it only pays when all sixty steps line up behind it.
DOOR ONE
Read all sixty — $5
177 pages. Nothing held back, including the sixteen that went wrong.
DOOR TWO
Or we run all sixty for you
Nothing to build it. You pay per lead, at a price locked before we start.
ONE AD ACCOUNT, 1 JAN – 1 SEP 2026·46 CAMPAIGNS·127 AD SETS·2,128 ADS·ALL OF IT OUR OWN MONEY
THE ACTUAL PROBLEM
Nobody buys expensive, complicated work from someone they don’t trust yet.
And trust is not a message. It is hours of exposure — enough time with how you think that a stranger stops needing to take your word for it. You cannot buy those hours, you cannot schedule them, and you certainly cannot fit them into an ad. That one fact is why each of these fails.
✕
Ads fail
A click is seconds long. It cannot carry hours. We have bought more than twenty thousand of them across our own products, and the only ones that ever became a customer arrived through something the reader could sit down with.
✕
Outbound fails
You pick the company, the person and the day, so the timing is entirely yours and none of it is theirs. Budget, approval and need are almost never in your window.
✕
Agencies fail
They sell you confidence, and confidence costs nothing to assert. The guarantee is written over their fee — never over the thirty thousand of your money that went through the ad account.
THE ONE THING THAT WORKS
Referrals — because somebody already spent the hours on your behalf. The entire problem is that you cannot schedule a referral. A book is the only asset we have found that manufactures the same thing on demand, for five dollars.
WHAT WE RAN, AND WHAT SURVIVED
Forty-six campaigns. Four of them worked.
Years of cold email, LinkedIn at volume, and every kind of paid funnel — ads to a call, free tools, lead magnets, short offers and long ones. Then a book. Same account, same pixel, same people, same money:
4 CAMPAIGNS · 1,411 CLICKS
A $5 book, run through all sixty steps
99
SALES · 7.0%
12 CAMPAIGNS · 2,679 CLICKS
A book, without the sixty steps behind it
5
SALES · 0.19%
30 CAMPAIGNS · 20,131 CLICKS
Everything else we have ever advertised CONTEXT, NOT A LIKE-FOR-LIKE
4
SALES · 0.02%
Read the middle row again. That is the whole argument.
Twelve of those campaigns were also selling a book. Between them they bought 2,679 clicks and sold five copies. The four with all sixty steps behind them sold ninety-nine — thirty-seven times the rate, on the same kind of product, out of the same account.
So the book is not the mechanism. Publishing one and hoping is how you get the middle row. The sixty steps are the entire difference between the two — which is what this page is actually about.
Basis: campaign-level export from our Meta account, 1 January to 1 September 2026 — link clicks and platform-counted purchases on every row, so the comparison is like for like. Meta over-counts purchases against our Stripe ledger by roughly a quarter, and it does so on every row equally, which is why the ratios hold while the absolute counts are the platform’s rather than our bank’s. Reconciling those two is step 53. The third row is context, not proof: it spans eight of our own products, and three of them cannot fairly be read as failures — one runs its own tracking and may never have reported to this pixel, one converts inside a free chat rather than at a checkout, and several were traffic or lead campaigns that were never optimised for a purchase at all. Judging those on sales would be reading a test against an intent it never had, which is step 58. The like-for-like comparison is the first two rows — both selling a book, same account, same pixel.
WHY SIXTY, AND NOT THE FOUR THAT SOUND IMPORTANT
Every one of these numbers moved. None of them fed us a customer.
The output is not a click, or an engaged reader, or even a book sale. It is a person with the budget and the need to buy the thing you actually sell. Between an impression and that person sits a chain — and a chain has no best link. Pull one and its number moves beautifully while the link after it quietly breaks. Here is each one we pulled, and what it bought us.
01
GETTING NOTICED
Click-through rate
22% — the highest we have ever posted, and it sold nothing. The fifty best click-through rates in the account bought 7,708 clicks between them and produced not one sale. The ad that has sold the most sits at 2.87%.
02
BUYING THE CLICK
Cost per click
Our cheapest clicks cost thirteen cents and bought nothing. The clicks that actually sold cost nine times more. Optimise this one and you systematically select against the people who can afford you.
03
HOLDING THE PAGE
Engagement
A page variant lifted engagement sixteen points and sold fewer books. Separately, 66% of visitors never scrolled past the hero at all — we had written the good part further down, where they were never going to reach it.
04
TAKING THE MONEY
Average order value
$16.57, against a break-even floor of $20–40. We treated it as a traffic problem for months. It was a pricing problem wearing a marketing problem’s clothes, and no amount of better traffic was ever going to fix it.
05
READING THE RESULT
Reported return
The platform said 0.06 on an ad set that was doing 1.55 on settled cash. Three separate distortions, all stacking the same way. Believe the dashboard and you kill your best performer on a Tuesday.
06
CREDITING THE SALE
Attribution
$0 creditable, on a fully built, live, well-designed funnel — because campaign parameters never reached the payment record. Not less accurate. Impossible, at any spend, on any day.
07
SOMEONE WHO CAN ACTUALLY BUY
Revenue per book buyer
The last link, and the one almost nobody instruments. What a customer is worth over their lifetime, multiplied by how often a book buyer becomes one. Get this wrong and all six above can be perfect and it still does not pay — you will have built a beautiful machine for delivering people who were never going to buy from you.
WHAT IT ACTUALLY COST TO FIND ALL SEVEN
The media was eleven thousand dollars. That was the cheapest part of this by an order of magnitude.
A$11,770
TOTAL MEDIA, EIGHT MONTHS
The only line anybody could copy
You cannot buy this by spending money on advertising. We know, because that is what we tried first and it is the whole left-hand column above. What it actually cost was eight months, two people who do this for a living rather than as a side project, and three complete rebuilds of the machine underneath — each one because the build before it was wrong in a way we could only find by running it.
2,128
ads written, scored, launched and graded
797,335
impressions bought and read back
190
scripts in the pipeline that runs it
246
commits across three rebuilds of the funnel
That is the year, and it is the actual asset. Not the eleven thousand dollars.
It is also the entire reason the book is sixty steps rather than the four that sound important — every one of them exists because a link in that chain broke and somebody had to go and find out why. Door two exists so that you do not repeat it: you get the machine on day one, and you pay for leads.
And the seventh link is the only one that decides whether any of the other six were worth doing — so it is the first thing we run for a client. In writing, before anybody commits to anything, and we charge for it because it is real work.
SO HERE IS WHAT JUST HAPPENED TO YOU
We did not pitch you. We handed you the method and let you check it.
Four moves, in order. You are somewhere in the first two right now — and this is the identical sequence we would build and run on your business.
1
The proof, up front
Hours of our real thinking, free to read, no email. What would otherwise take several coffees and a lot of your evenings to demonstrate.
→
2
A $5 test of intent
Five dollars filters for the one thing nobody can schedule: the problem is live now and the money is already allocated. Curiosity does not pay five dollars.
→
3
A conversation that is not a pitch
By the time we speak you already know the mechanics, so you arrive with the detail questions. It is a check for fit and edge cases, nothing more.
→
4
Trust, already built
The understanding and the fit exist before anyone signs anything — built out of the book, not asserted by us in a deck.
If it worked on you, that is the demonstration. It is also the only sales argument we have, which is why the whole method is on sale for five dollars instead of locked behind a call.
WHO IS SAYING THIS
Measurement is our day job. That is the entire reason to believe any of it.
We are not an agency that learned analytics. We are a data and analytics consultancy that turned the instruments on our own marketing and wrote down what came back. Every agency sells a feel for what works and asks you to buy it before you can check it. We spent years proving that intuition does not survive contact with data — mostly by being wrong ourselves.
THE TWO PEOPLE WHOSE LEDGER THIS IS
RR
Ryan Richardson
Founder, Onwards Analytics
Builds the measurement systems. Wrote the instrumentation half of the book — parts VI to IX, which is where most of the money is lost.
DR
Dan Rodgers
Marketing lead, Onwards Analytics
Runs the offer and the creative. Wrote the market and positioning half — parts I to V. Every section names which of us is talking.
How you can tell us from an agency
✓We publish the losses with their figures. Nobody selling a template shows you the version that lost money.
✓Every number is reconciled against the payment ledger, never the ad platform’s own report of itself.
✓Fifty-seven failure modes are written down, one per step. Engineers document what breaks.
✓We rebuilt this funnel three times, and the third build still had bugs the first two taught us to look for.
An agency that has never been wrong has never measured. We can show you sixteen times we were wrong, because we had the instruments to catch it.
WHERE THAT DISCIPLINE CAME FROM — TWENTY YEARS OF PROBABILITY AND BEHAVIOURAL ANALYSIS FOR ORGANISATIONS LIKE THESE
You can run the sixty yourself. Most people should not.
Forty-one of them are a machine’s job, twelve need your judgement rather than your time, and the year it takes to discover which is which is a year of the work you are actually good at. So the other option is that we build the whole thing and run it.
THE ENTIRE COMMERCIAL ARRANGEMENT
You pay nothing to build it. You pay for leads, at a price locked before we start. If our cost to acquire them goes up, that is ours to absorb, not yours.
No setup fee. No retainer. No percentage of ad spend. We do not bill for effort, because effort is our problem. You prepay a refundable lead balance and we draw it down only as leads actually land. That risk transfer is the product — everything else is just how it gets delivered.
WHAT WE DO
Write your book, build the funnel, run the media on our own accounts. All sixty steps on your business — you do the twelve that need you.
WHAT A LEAD IS
A named person who paid for your book and is on your list. Not an impression, not a click, not a form-fill from someone who wanted a PDF.
WHO VERIFIES IT
Everything up to the handoff happens in our systems, so a lead is countable in our payment ledger. There is nothing to adjudicate and no dispute path to negotiate.
WHAT YOU KEEP
The book, the site and every lead — including if you leave. Month to month, no minimum term, no exit fee.
WHAT DECIDES YOUR PRICE
How many people could ever buy from you. A wide market and a single-vertical one are different machines, not one machine at two prices.
WHO IT IS NOT FOR
Anyone whose arithmetic will not carry a lead price. That is most businesses, and we would rather tell you in five days than find out together in five months.
HOW FIT IS DECIDED — AND IT IS NOT DEAL SIZE
Revenue per book buyer: what a customer is worth over their lifetime, multiplied by how often a book buyer becomes one. Lifetime value, never the size of a single deal.
A firm selling million-dollar engagements that closes one a year from a huge pool is a worse fit than a smaller service that closes often and repeats. This is the arithmetic almost nobody runs before buying leads, and it is the only thing that decides whether a lead can be priced at all.
WHY WE CAN AFFORD TO CARRY THE RISK
Our own funnel, measured 1 January to 31 August: A$60.36 to acquire a book buyer, A$31 back per buyer from the book and the two add-ons.
Roughly half the media comes back before we bill a client a cent — which is steps 3 and 50, and the whole reason the offer can exist. Anyone who needs a retainer before touching your account is telling you their own front end does not work.
IT STARTS WITH THE ARITHMETIC, NOT A CALL
We run your numbers first — and we charge for it.
A written verdict inside five business days: yes at a named volume and price, or no with the numbers that make it a no. An honest no is the deliverable, not a failed sale. If it is a yes, what you paid comes off the engagement in full. Offered to book buyers at checkout — read the method first, then decide whether you need us at all.
5 days
TO A WRITTEN VERDICT
Rather just talk it through?
Thirty minutes with one of the two people who built this. We will tell you on the call whether your arithmetic looks like it carries a lead price — no deck, no pitch.
The full table of contents. If a step is on this list it is in the book, in enough detail to do it — nothing held back for an upsell. Tap any step.
Went wrong — numbers publishedNeeds your judgement, not ours
PART IDecide what you actually sell
PART IIMine the market
PART IIIExtract the point of view
PART IVWrite it
PART VProduce it
PART VIBuild the machine
PART VIIInstrument it — the part everyone skips
PART VIIICreative and launch
PART IXRead it and decide
ONE OF THE SIXTY, IN FULL, SO YOU CAN JUDGE THE REST
HARDALMOST NOBODY DOES THISAUTOMATED
45. Persist campaign parameters into the payment record
Three requirements, every one of which fails quietly if you get it wrong. Miss it and every sale becomes un-creditable to any campaign — not less accurate, impossible. This is the step that cost us the $0 in the receipts above.
TOOLS
Edge middleware, which every modern host provides. Cookies, never local storage — the server has to be able to read them. And your payment provider’s metadata field: the most underused feature in commerce tooling.
WHERE IT BREAKS
Overwriting a stored campaign with an empty one, which is the most common version. Metadata field limits, which are real and low. Redirects that drop the query string. Any checkout on a different domain.
HUMAN OR MACHINE
Fully automatable, and it should be. But verify it by hand before every launch by making a real purchase through a real ad. Five minutes — the single highest-value manual check in the book.
REASON IT OR BUY IT
Reason it, then verify with a purchase. The design is knowable; whether your specific stack drops the value somewhere along the way is not, and one test transaction settles it.
Every one of the sixty is written out like this. That is what the 177 pages are.
WHY YOU CANNOT JUST ASK A MODEL FOR THIS
41 steps run without a person. Getting them to run without a person is the expensive part.
They are not automated because you can ask a model to do them. They are automated because there is software underneath, and that software exists because it has been wrong a great many times and was corrected each time. None of that is in a prompt.
190
scripts
in the advertising pipeline alone: generate, score, upload, read, iterate
246
commits on the funnel
across three builds — the third still had bugs the first two taught us to look for
14
creative batches
each one generated, scored against a rubric and approved before a dollar moved
57
documented failure modes
one per step — the receipt for the iteration
And twelve that no software will ever do.
These need somebody who knows your field — which is you, and it is the reason this is not a service you can be entirely absent from. We bring the machine, you bring the judgement. Neither half works alone.
03 Revenue per book buyer
08 The one sentence
13 Avatar & awareness matrix
14 The contrarian spine
16 The receipts, incl. failures
18 The hard exclusions
32 Print-on-demand drop-ship
33 Retail as second channel
52 Blind-verify every launch
57 Reprice on evidence
59 Resist obvious additions
60 Floor problem or spike problem
DOOR ONE
Check us for five dollars.
The whole method, before you decide whether you want us anywhere near your business. PDF, sent to whatever email you use at checkout, usually within a minute.
The Sixty Steps
177 PAGES · NINE PARTS · SIXTY STEPS
$5
✓All sixty steps in the order they have to happen, each with its tools, its failure mode, and whether to automate it
✓The sixteen we got wrong, with the actual numbers attached
✓Fifty-seven documented failure modes — one per step
✓Sixty days to ask for your money back, and you keep the book either way
TWO OPTIONAL ADD-ONS — SEPARATE WORK, NOT CHAPTERS HELD BACK
Refundable on request · No email required to read this page · Nothing to schedule
QUESTIONS
Asked before you asked.
Because knowing the sixty and doing the sixty are not the same job. Most readers will do a handful and stop — not a criticism, just what happens when you already have work to deliver. The ones who go and build the whole thing themselves were never going to hire anybody, and we would rather find that out at five dollars than at the end of a sales process.
Its job is not to make money. It is to prove we can do this, to exactly the sort of person who will not take that on trust — and to filter for the ones whose problem is live right now. The book and its add-ons return about half of what the advertising costs, so the front end roughly pays for itself and the leads it produces are close to free. That is the same mechanism the book teaches, in steps 3 and 50.
A named person who paid for your book and is on your list. Not an impression, not a click, not a form-fill from somebody who wanted a free PDF. Everything up to the handoff happens inside our systems, so it is countable in a payment ledger rather than argued about in a monthly report.
It depends entirely on how many people could ever buy from you, because a wide market and a single vertical are different machines rather than one machine at two prices. We will not quote a number before running your arithmetic, and quoting one here would be the same guessing we are arguing against. That is what the paid teardown is for — a written verdict with a named volume and price inside five business days.
No. Your leads are yours, along with the book and the website, and you keep all of it if you leave. Month to month, no minimum term, no exit fee. We keep the machine that produces them and the front-end revenue that funds the advertising — and we say so in writing before you commit to anything.
No. Every step on that list is in the book in enough detail to execute. The two add-ons at checkout are separate pieces of work — one live funnel’s actual assets, and a working spreadsheet — not chapters removed from the book to create a reason to buy them.
Yes, on the book and on both add-ons, for any reason or none, for sixty days. Ask and we refund it, and you keep everything. At this price arguing with you would cost us more than the refund does.
Ryan Richardson and Dan Rodgers, of Onwards Analytics. We each wrote the parts we actually run — Dan the market and positioning half, Ryan the build and instrumentation half — and every section says which of us is talking. The figures throughout are from our own firm’s ad account and payment ledger.
Here are the sixty steps. Or we run them for you.
Read the whole method for five dollars and decide for yourself whether we are any good at this. That has been the point of the page the entire time.
