There is no peer-reviewed study of post-conference follow-through as such — anyone who quotes you one is inventing it. What exists is stronger: converging evidence from memory science, goal psychology, and training research. Here is what it actually says.
Across hundreds of studies, people's intentions explain only about 28% of what they actually do — and roughly half of people who fully intend to act never do. A conference manufactures intentions at industrial scale, then sends them home unsupported.
Sheeran (2002); Sheeran & Webb (2016) — meta-analyses of the intention–behaviour relationship.
Training managers in 150 organisations estimated the share of employees applying what they learned: 62% immediately after training, 44% at six months, 34% at one year. Stated honestly: these are practitioners' estimates, not observed behaviour — and they slide in one direction only.
Saks & Belcourt (2006), Human Resource Management — survey of training professionals in 150 organisations.
For meaningful material, nobody can honestly give you one percentage: across real studies, people forgot anywhere from nothing to three-quarters of material within two days. What is consistent is the shape — the loss is steepest early, then flattens. Which is why the follow-up has to start early.
Thalheimer (2010), Work-Learning Research — review of 69 experimental conditions; forgetting-curve shape replicated since Ebbinghaus (1885) by Murre & Dros (2015).
Turning a goal into a specific if–then plan — when this situation arises, I will do that — measurably raises follow-through. The honest current estimate is a modest effect: real, replicated across hundreds of tests, and still standing after correction for publication bias. Modest and real beats dramatic and invented.
Gollwitzer & Sheeran (2006); Sheeran, Listrom & Gollwitzer (2024) — meta-analyses of implementation intentions.
Across 138 experiments, monitoring progress toward a goal promoted actually reaching it — and the effect was largest when progress was physically recorded and reported to another person. Not resolved in private. Reported, to someone, on a schedule. That finding is the backbone of this program.
Harkin, Webb, Chang et al. (2016), Psychological Bulletin — meta-analysis of 138 experimental studies.
Students who were tested on material recalled 61% of it a week later. Students who spent the same time re-studying it recalled 40%. Being asked beats being re-shown — which is why our follow-ups ask what you did, instead of re-sending you the slides.
Roediger & Karpicke (2006), Psychological Science — retrieval practice vs. restudy, one-week recall.
A new behaviour takes about 66 days to become automatic, on average — with a wide, honest range, from 18 to 254 days. That is why this program runs on a six-month arc. Anyone promising you a transformed practice in thirty days is selling you the calendar, not the change.
Lally, van Jaarsveld, Potts & Wardle (2010), European Journal of Social Psychology.
No study proves a cliff at 24, 48, or 72 hours — and we won't pretend one does. But every strand of evidence points the same way: memory loss is steepest early, intentions are strongest before they destabilise, and the first review of something you need next week belongs within a day or two. So the first action lands while memory, motivation, and context are all still warm. Days, not weeks.