Why does a low CPL number from one month mean so little?
Cost per lead moves for reasons that have nothing to do with marketing skill. A single viral post, a seasonal dip in ad auction competition, a competitor pausing spend, any of these can drop CPL for two or three weeks without any underlying change to the campaign's actual quality. A provider who screenshots that window and publishes it as a benchmark isn't lying exactly. They're reporting noise as signal.
The reverse happens just as often. A genuinely well-built campaign can have a rough month from algorithm learning-phase resets, a platform policy change, or a seasonal spike in competition, and look worse than a mediocre campaign that happened to catch a cheap week.
What's the difference between a benchmark and a track record?
A benchmark is a snapshot. A track record is a pattern held under changing conditions: different seasons, different ad platform algorithm updates, different competitive pressure. myTTConline's 2.99x return held for 21 consecutive months, which means it survived multiple algorithm updates, seasonal enrolment cycles, and whatever competitors were doing during that window. That's a different kind of evidence than a single number pulled from one reporting period.
Training providers comparing agencies or in-house results against published benchmarks are almost always comparing a snapshot against a snapshot, with no visibility into how either number behaved the month before or the month after.
A return that survived multiple algorithm updates and a full seasonal cycle.
How should a training provider actually read a marketing case study?
Ask for the range, not the peak. A case study that only shows the best month is showing you survivorship, not performance. Ask what the number looked like in the worst month of the reporting period, and whether the campaign structure changed mid-period to produce the good numbers, or held steady throughout. myTTConline's number is reported as a 21-month consecutive run specifically because a single strong month would say nothing about whether the same system would work through a slow season.
Does this same pattern show up outside vocational training?
Yes, and it's useful precisely because it isn't vocational-specific. Resicert, a franchise brand in Australia and New Zealand, cut cost per registration from A$43.75 to A$18.39, a 58% drop, across a network of 50-plus franchisees. That number is also reported as a sustained result across the franchise network, not a single location's best week. Different vertical, same underlying discipline: the number that matters is the one that holds when you widen the window, not the one that looks best when you narrow it.
Why do so many education marketing benchmark posts mislead training providers?
Most benchmark content is built from ad platform aggregate data or from a handful of anonymous accounts, averaged into a single figure with no context on cohort size, course type, or funding structure. An average CPL across "vocational training" as a category tells a plumbing college and a coding bootcamp almost nothing useful about each other, because their funnels, price points, and enrolment cycles don't resemble each other closely enough for the average to mean anything.
Worse, published benchmarks rarely disclose what percentage of leads counted were qualified versus browsers. A campaign reporting an artificially low CPL by counting every form fill as a lead, with no Green/Amber/Red qualification split, will always beat a campaign that reports only qualified lead cost. Comparing the two numbers side by side isn't a benchmark. It's an apples-to-rocks comparison.
Comparing an unqualified lead count against a qualified one is not a benchmark. It is an apples-to-rocks comparison.
What should a provider measure instead of chasing a CPL benchmark?
Measure consistency of cost per enrolled student across a rolling multi-month window, not the lowest CPL achieved in any single period. A provider hitting an average that holds steady across a full academic year, through both strong and weak intake months, has a system. A provider with one great month and three mediocre ones has a coincidence dressed up as a system.
This is also why single-month case studies should raise more questions than they answer. Ask how long the number has held. Ask what happened the month before the screenshot was taken. If nobody can answer either question, the number being pitched is decoration, not evidence.
Is a slightly higher but stable CPL better than a lower, volatile one?
Usually, yes. A training provider planning cohort staffing, facility capacity, and funding drawdowns needs predictability more than it needs the theoretical floor on cost per lead. A campaign that reliably delivers at a steady cost lets a provider plan intake sizes with confidence. A campaign that swings between cheap and expensive months makes planning a guessing game, even if its average looks fine on paper.