Picture the last afternoon of a launch kickoff. The medical director has walked the room through the phase III data for the third time, the brand lead has rehearsed the core story, and three hundred reps have role-played the first call in small groups. The event gives the field a starting point for continued practice and learning.
That is the problem. The kickoff is the peak, and without rehearsal, what was learned there starts to fade. Our argument in this piece is simple: a kickoff is a moment, readiness is a curve, and most launch plans budget for the moment and hope for the curve. We'll say briefly what the memory research does and does not show, why a longer or louder kickoff won't fix it, and what a practical 30-day plan after the event looks like, based on the programs we have run with field forces.
What the forgetting curve does, and does not, tell a launch team
In 1885 Hermann Ebbinghaus taught himself lists of nonsense syllables and measured how much effort it took to relearn them later. In 2015, Jaap Murre and Joeri Dros replicated the experiment, published in PLOS One, with one participant and the same kind of material, and found a curve close to the original. Measured as savings (how much less effort relearning takes), savings after 31 days were about 4%.
What that shows is narrow and useful: unrehearsed material fades fast, and spacing the rehearsal helps. What it does not show is how much of a safety session or a trial summary a rep will retain by the next morning. Launch content is meaningful, reps have a professional reason to remember it, and nobody has measured that curve for your product. We covered the research in more depth in our earlier post on why field training fades after the cycle meeting, so we won't repeat it here.
The practical point stands without the percentages. Anyone who has asked a rep to quote the primary endpoint six weeks after kickoff knows that what was presented once does not stay fully available on its own.
A kickoff is built to peak, not to last
None of this is a criticism of kickoffs. They do things nothing else can: align the organisation, create energy, put leaders in front of the field and make the launch feel real. The trouble is that we ask them to do one more job they are badly suited to, which is to make knowledge stick.
Consider what a typical launch kickoff crams into two or three days:
- disease background and the treatment guidelines the product will sit within;
- the mechanism of action, usually with a slick animation;
- registrational trial design, primary endpoint, key secondary endpoints and subgroups;
- the safety profile, the label and what reps may and may not say;
- access, formulary status, dosing and the practicalities of a first prescription;
- competitor positioning and objection handling;
- the visual aid, the call plan and the POA for the first cycle.
Each block is delivered once, in sequence, to a tired audience. Then everyone flies home. It is the educational equivalent of a single enormous meal, followed by a month without food.
The natural reaction, when field feedback shows the story isn't landing, is to add more. Another module. A longer e-learning. A second day at the regional meeting. We understand the impulse, but more volume at a single point in time doesn't change the curve. It just raises the peak a little before the same fall.
A kickoff is a moment. Readiness is a curve. Most launch plans budget for the moment and hope for the curve.
Spacing beats volume, and short beats long
The better answer is almost boringly well established: spaced retrieval. Bring the key facts back at intervals, and make people actively recall them rather than reread them. Each successful recall, just as the memory is starting to fade, makes the next decline slower.
In practice, for a field force, that means three design choices.
First, make each touch tiny. A rep between calls will not open a 40-minute module. They will answer five questions in sixty seconds on their phone, especially if a colleague has just challenged them. Short is not a compromise here. It is the mechanism.
Second, make it retrieval, not review. Rereading the slide on the exacerbation data feels productive but does little. Being asked to choose the right figure from a set of plausible options, and getting it wrong in front of yourself, does a great deal more. The questions in our examples have three or four options; the number matters less than whether every wrong option is one a colleague might really pick.
Third, make it personal. The same question bank served to everyone wastes most people's time. What each person needs is the fix for their own mistakes, delivered close to when they made them.
We'd add a fourth, which is less about memory science and more about human nature: make it social. People answer more, and more carefully, when they are playing against real colleagues in a defined window than when they are working through a module alone on a Sunday evening.
A plan in three phases: before, during and after the kickoff
The kickoff doesn't have to carry the weight on its own. We think of it as the middle of a three-phase plan.
Before: find out what people already get wrong
A short diagnostic wave two or three weeks ahead tells you where the field starts. If most reps already know the disease area well but misunderstand how the new mechanism differs from the class they currently sell, you can cut the disease refresher in half and spend that time where it matters. The kickoff agenda gets shorter and sharper.
During: capture the peak
A synchronised battle during or immediately after the event, with everyone playing in the same window, does two things. It gives you a clean measurement of what the room actually took in, by team, country and topic. And it gives people their first retrieval practice while the material is fresh, which is exactly when it is cheapest to strengthen.
After: reinforce what each person missed
This is the phase most plans skip, and it's where the curve is won or lost. For the following weeks, each person receives short waves built on their own gaps, plus the correction for each mistake. The questions that come back wrong again are the ones that need a manager, not another quiz.
What it looks like in practice
Two programs from our case library show the pattern at different scales.
The first was an APAC MedTech vascular program that ran seven congress challenges over 26 months, each in a window of about three days. Rather than one big event, the field met new evidence in short, repeated bursts. Across the seven waves, 24,183 knowledge gaps were found at first attempt and 49% were verified closed, meaning the same person later answered the same question correctly. The gaps still open were not treated as failure. They became the coaching list and the brief for the next wave.
The second was a respiratory launch in Japan with 257 players. Rather than one big block, the content was split into two waves: disease first, then product. Over 32 days the field recorded around 80,000 answers and 4,503 gaps, and 80% of those gaps were closed. Readiness rose from 75% to 95%. Separating disease and product also meant managers could see which kind of knowledge each person was missing, which is a far more useful coaching conversation than "you scored 70".
Across the programs we run, closure varies a lot, and the same mechanics used well or badly produce very different results. One pattern we see repeatedly is that programs tied to a real business moment (a launch, a kickoff, a certification, a congress or a cycle meeting) close more of their gaps than always-on programs. That is an association, not proof that the moment causes it. But it suggests the kickoff is not the enemy of retention. Used as an anchor, it can help.
Repetition fixes most gaps, but not the dangerous ones
Here's the honest limit of everything above. Spaced retrieval is excellent at closing gaps that come from simply forgetting. It is much weaker against gaps that come from believing the wrong thing.
When we track individual questions across programs, some close well, some close halfway, and some stay stubbornly open. On a share of the example questions we track, very few of the people who were wrong ever corrected it. Some of those people were never retested, so not every one is a proven misconception. But where people saw the question again and kept the same wrong answer, it is rarely a memory failure. It is usually a misconception, usually about mechanism, patient selection or how a trial result should be read, and they tend to survive quizzing because the rep's mental model keeps reproducing the error.
In one global immunology kickoff with 1,330 players, a single misconception about the mechanism of action, held by 125 reps, barely moved over several rounds of repetition. The rest of the program closed 53% of its gaps. This one needed a different tool: a medical colleague, a manager on a ride-along, or a short live session built around exactly that point.
That's why the after-kickoff plan needs two outputs, not one. The first is a curve that stays high. The second is a short, specific list of misconceptions that need human coaching, sent to the people who can do it.
Who owns the curve?
In most organisations, the kickoff has a clear owner: usually the launch team or commercial excellence, with L&D and medical supporting. The weeks after it often have no owner at all. Brand moves on to the first-cycle materials, L&D moves on to the next curriculum, and first-line managers are busy with territory plans.
We'd suggest giving the first 30 to 60 days after kickoff an explicit owner and a simple metric. Not "completion", which only tells you who clicked, but the share of critical gaps closed, by team and country. When that number sits on the same slide as early sales indicators in the monthly launch review, people start paying attention to it.
It also changes the conversation with MLR. Reinforcement questions are built from the approved materials, so the bank can be reviewed as a set rather than piece by piece. Reuse is simple. Rewording is not free: a reworded question, or an explanation adapted for a local market, is new material and needs to go back through the same governed review, with a record of who approved which version. Plan that review time into the 30 days.
Plan the curve before you book the venue
If you're planning a kickoff in the next few months, try one exercise at the next launch meeting. Ask the team what the field should know cold on day 30, not day 1. Then ask what, specifically, will happen between the plenary and that date to make it true. If the answer is "they'll have the materials", you have a plan for the moment and not for the curve.
This is the work we do at Atrivity: short, spaced, competitive retrieval on the phone, with Gap Fixing that sends each person the fix for their own mistakes and readiness data managers can act on. But the principle doesn't depend on any tool. Measure the peak, protect it with spaced retrieval, and hand the stubborn misconceptions to people who can coach them.
The kickoff will still be the best day of the launch. It just shouldn't be the last day anyone remembers what was said.