Between 2011 and 2019 we ran 26 Time-Driven Activity-Based Costing workshops in Portugal and Brazil. 1,521 people registered for them, from 935 different companies. 566 paid and sat in the room. The other 955 raised a hand and never paid. Those numbers come from one spreadsheet, the registration list we kept for every session, and we extracted them in April 2026 while deciding what our training should look like next. Nobody had ever added them up before.
What do nine years of TDABC workshop data say about who pays to learn costing?
Across 26 workshops, 37 percent of the people who registered went on to pay. In the first workshop, in 2011, that figure was 76 percent. By 2018 it was 15 percent, and the two workshops of 2019 together sold 16 seats. The average price paid fell only 17 percent over the same period, so price does not explain the collapse. What changed was the moment of payment: registering stayed easy and free, paying became something a person had to ask their company for, and most never did. That reading is why our training now has a self-paced online edition priced for a decision one person can take alone.
What the registration list actually contains
The source is a single workbook, 33 sheets, one per session plus a few utility tabs, maintained from the first workshop in 2011 and last updated in July 2019. Twenty-seven of the sheets hold participant rows, and 26 sessions carry a year we can identify. Each row is a registration: name, company, whether the person paid, whether they cancelled, and in a handful of cases whether a follow-up meeting was booked. We extracted only what the sheet states. Where a figure is not in the workbook it is not in this article, and the section near the end lists what we could not measure.
Two totals frame everything else. 1,521 registrations and 566 paying participants, a conversion of 37.2 percent. The no-show and cancellation rate among those who paid was 2.0 percent, which is worth saying out loud because it rules out one explanation immediately. The problem was never people paying and not turning up. People who paid came. The whole gap sits between registering and paying.
Portugal and Brazil, side by side
Twenty-two of the 26 sessions were in Portugal, four in Brazil. Portugal accounts for 1,340 registrations and 512 paying seats, a conversion of 38.2 percent. Brazil accounts for 158 registrations and 54 paying seats, 34.2 percent. The Brazilian sessions were more expensive to run, because they involved travel, and the workbook is honest about not recording those costs consistently, so we cannot say what margin they made. What we can say is that conversion in Brazil ran four points below Portugal, not dramatically different, and that four sessions are too few to read a trend into.
The curve, year by year
The first session, in 2011, is the one every later session gets compared to: 131 registrations, 100 paying. Three in four people who signed up paid to attend. 2012 brought four sessions and 320 registrations, but only 109 paid, so conversion had already fallen to 34 percent. 2013 looks worse than it was: 298 registrations and 84 paying, but 125 of those registrations came from a single free session run with a software vendor, which by design produced no paying seats. Take that session out and the year converts at just under half.
From 2014 to 2016 the pattern is stable. Three or four sessions a year, 141 to 170 registrations, 62 to 74 paying, conversion between 37 and 53 percent. Then the volume goes. 2017: two sessions, 131 registrations, 38 paying. 2018: two sessions, 65 registrations, 10 paying, a conversion of 15 percent. 2019: two sessions, 72 registrations, 16 paying. In November 2018, in Porto, 51 people registered for a full-day session and two paid.
Over the same nine years the average price a participant paid fell by 17 percent. That is a real decline, but it is nowhere near the size of the fall in conversion, and it points the wrong way to explain it: the price was at its lowest in the two years when the fewest people paid.
Why we read this as a payment problem, not a price problem
A workshop registration in 2011 was, in practice, a purchase. The person filled in a form, the company was invoiced, the seat was taken. By 2018 the same form had become a free expression of interest. Registering cost nothing and committed nobody, so people registered the way they now click “interested” on an event, and the actual decision to pay moved somewhere else: to a manager, a training budget, a procurement step. Most of those decisions were never made, and the workbook shows them as 955 rows with no payment against them.
We saw the identical failure again in August 2026, at the checkout of an online workshop for Brazil, when the payment page asked Brazilian participants for a Portuguese tax number and charged them Portuguese VAT. The interest was there; the moment of paying was broken. The lesson generalises beyond our own training. A funnel that counts registrations is counting the wrong thing, and a conversion figure that looks like a price problem is often a friction problem wearing a price tag.
The people who paid always turned up: 2 percent no-shows across nine years. The seat was never the problem. The moment of paying was.
What we could not measure, and why it matters
Three things are missing from the workbook, and each is a lesson in its own right. The column for “follow-up meeting booked” is filled in on 7 of 1,521 rows, so we cannot say how many workshops led to projects; the honest answer is that nobody tracked it. There is no unique participant identifier across sessions, so we cannot count repeat attendees. And the Brazilian sheets are planning templates rather than records of costs actually incurred, so we cannot compute a margin for those four sessions. A training programme that ran for nine years without a way to link a seat to a later conversation was measuring attendance and calling it results.
The 935 distinct companies are the number we find most useful. 1,521 registrations from 935 companies means that 586 of them, well over a third, were a second or later person from a company that had already registered someone. That is a network, not an audience, and it is the reason the workshops are still running.
What changed in 2026
The in-person workshops continue, in Porto, in Lisbon and online, and they still sell to the people they always sold to: teams that want a day with the method and their own numbers in the room. What is new is a second door for the 955. A self-paced online course, eighteen sessions with exercises and an exam, priced so that one person can decide alone without asking anyone for a budget. It exists because the workbook told us that the audience was already there, had already raised its hand, and had simply never been offered a way to pay that did not require permission.
The online TDABC course is the self-paced edition of the workshop: eighteen sessions, exercises on a real model, a certificate at the end, and an option that works through your own company’s numbers with us.
Related reading: upcoming workshops, what you learn in a TDABC workshop and how this firm started.