Attention Isn't Given.
It's Spent.
Learners aren't losing the ability to pay attention.
They're making rational decisions about where to spend it.
That's not a capacity problem. It's a design problem.
Attention Is a Resource. Not a Reflex.
Herbert Simon saw it coming in 1971. A wealth of information creates a poverty of attention. His point was precise: information doesn't just compete for your learners' attention. It consumes it. And attention, like money, gets allocated — not lost.
Daniel Kahneman confirmed the mechanism two years later. His research showed that people direct their focus based on perceived importance – not randomly, not based on elapsed time, but on whether the next moment seems worth it.
This isn't attention deficit. It's rational economics.
The Numbers Everyone Gets Wrong
Two numbers often drive the attention conversation in L&D right now. One is fabricated. One is real and widely misread.
The real one: 47 seconds. Gloria Mark at UC Irvine tracked screen-switching behavior for two decades. In 2004, people averaged 2.5 minutes on a task before switching. Today: 47 seconds, replicated across five independent studies. Philip Guo analyzed 6.9 million edX video sessions and found engagement collapses at 6 minutes — videos over 12 minutes got roughly 3 minutes of actual watch time.
The fabricated one: 8 seconds, shorter than a goldfish. A 2015 Microsoft report cited a source that didn't exist. It spread because it felt true. It wasn't.
These numbers get called attention limits. They aren't. They're exit decisions.
A learner who switches screens at 47 seconds has made a judgment: not worth it. The one who stops a video at 3 minutes has made the same call. Attention followed value until value ran out. Then it went somewhere that offered more.
That behavior has a name. L&D built a model around it in the 1980s. Then mostly forgot it.
L&D Had the Answer. Forgot It Anyway.
John Keller's ARCS model — Attention, Relevance, Confidence, Satisfaction — was built directly on Expectancy-Value Theory. The equation is blunt: motivation equals expectancy times value. Either factor hits zero, engagement hits zero. Not a drift. A cliff.
Research on relevance interventions confirms it.
When learners understand why content matters to them personally, engagement jumps across every measure.
Keller wasn't describing good design. He was describing the entry fee.
The field had this for forty years and reached for a 10-minute timer instead anyway.
Marketers didn't have that luxury. Their feedback loops — click rates, bounce rates, conversion data — made getting attention wrong immediately visible and immediately painful. That forced one rule above all others: earn attention before you spend it. L&D never faced that reckoning. So myths filled the void.
The Harder Question
Before you build anything, ask this:
why would a learner believe the next 45 seconds is worth their attention?
Not whether the content matters. You know it does. Whether they'll feel it in the opening moment, before you've done anything to earn their focus.
That's your actual design problem. Not length. Not format. Not which authoring tool.
For eLearning: lead with stakes, keep it short (ideally under 6 minutes, if possible), cut anything that doesn't earn its place. Your learner is running a live cost-benefit call. Win it early or don't win it.
For ILT and VILT: the research is consistent — attention in live learning tracks the instructor, not the clock. Relevance, interactivity, stakes. Those are the levers. A timer is not.
