What attendees actually remember after a conference

If you have spent any time in event marketing you have seen the number. Attendees forget 70% of what they heard within 24 hours. It shows up in vendor decks, in LinkedIn posts, and in at least one article a week about “return on attendance.”

I build software whose whole job is helping people keep what they heard at an event, so you would think I would love that number. I don't use it, because it doesn't say what people think it says.

Where the number comes from

The 70% figure traces back to Hermann Ebbinghaus, who in 1880 taught himself lists of nonsense syllables (“DAX”, “BOK”, “YAT”) and measured how long it took to relearn each list after a delay. The gap between first learning and relearning gave him his famous forgetting curve. In 2015 two researchers at the University of Amsterdam, Jaap Murre and Joeri Dros, repeated the experiment carefully, with one subject spending about 70 hours on 70 lists, and got a curve very close to the original.

The curve is real. What Ebbinghaus measured is another matter:

• The material had no meaning. Nonsense syllables are chosen precisely because you can't connect them to anything you already know. A keynote is the opposite.
• He measured savings, meaning how much faster relearning went. That is a different thing from what share of the list someone could repeat back, and “savings” of 30% does not mean “70% forgotten.”
• It was one person working alone, on purpose. There were no slides, no hallway conversation, no colleague asking “what was good?” on the train home.

The curve is still useful. Its shape holds up: most of the loss happens early, then it flattens, and that matters a lot for events. The precise percentage doesn't transfer to a 40-minute talk with a story in it.

What the curve does tell an organizer

Forgetting is steepest in the first day or two, and anything that brings the idea back during that window pays off disproportionately. Memory researchers call this the spacing effect, and it's one of the most replicated findings in the field.

In practice that means three things for an organizer:

1. The first 48 hours matter most. A recap email on day five is reaching people after the steep part of the curve. One on the evening of the session reaches them before it.
2. People keep what they connect to their own work, so volume doesn't help. A list of twenty bullet points per session is close to nonsense syllables. Three ideas with the example that made each one land are not.
3. Doing something with an idea is the strongest form of review. Writing one line about how you would use it beats rereading it twice.

What people actually walk out with

Ask attendees a week later what they got from a talk and you rarely hear the speaker's thesis restated. You hear one of these:

• A story or example, like “The airline that removed the confirmation step.” Stories carry the idea with them.
• A number. Specific figures are surprisingly sticky, which is also why invented ones are dangerous.
• A line they could quote, usually a short sentence the speaker said almost as an aside.
• Something they decided to do. This is the most valuable and the least often captured.

That list is also a decent spec for any takeaway, whether a person or software writes it. It should keep the concrete detail and, where it can, the speaker's own words. It should also leave room for the attendee to add what they will do.

Four things you can do this week

None of these need new software. They work with a shared document and an email tool.

1. Get the takeaways out the same day. Ask each moderator or a volunteer to write three takeaways per session, each with the example or number that made it memorable. Send them that evening, before the steep part of the curve.

2. Quote, don't paraphrase, where you can. A takeaway with the speaker's actual sentence next to it is both more trustworthy and more memorable than a summary. If you record sessions, pull the line from the recording rather than from memory.

3. Ask one question back. End the recap with “Which one are you going to use?” and make replying easy. Most people won't reply. The ones who do will remember the idea, because answering is retrieval practice.

4. Send one more touch around day seven. Don't make it a survey. Send one idea from the event and one sentence about why it mattered, ideally from an attendee's reply. Spaced review only needs one or two repeats to work.

Where Gemlo fits

I built Gemlo because doing the four steps above by hand, across a dozen sessions, is a lot of work for an events team that is already exhausted. A phone records each session. A few minutes after it ends, attendees get a short deck of takeaways, each one showing the line from the talk it came from. They keep the ones that matter and can add a note on how they will use it. That note is private; organizers and sponsors never see it.

If you would rather do it by hand, the four steps still work. Get it out early, keep it specific, and leave the attendee somewhere to put their own thinking.