Rolling out new software often means employees finish training and pass knowledge checks, but weeks later, they hit roadblocks when real work begins. The problem usually lies in the gap between how people retain information and when they actually need to use it.
Why Software Training Fades Fast
Memory research, starting with Hermann Ebbinghaus’s work in the 1880s, shows that most new information slips away quickly unless it’s used or reviewed. Ebbinghaus’s experiments, published in 1885, involved memorizing lists of meaningless syllables and tracking how much effort was saved when relearning them after different time gaps. His results showed a steep drop in recall at first, then a slower decline. Historical summaries report about 58% savings after 20 minutes, roughly 34% after 24 hours, and around 21% after 31 days. These numbers reflect how much easier it is to relearn, not the percentage of information remembered or lost.
In 2015, Jaap M. J. Murre and Joeri Dros at the University of Amsterdam replicated Ebbinghaus’s forgetting curve in PLOS ONE, confirming the pattern of rapid early forgetting followed by a slower decline. Their study used the original "savings" method and reinforced that unrehearsed learning becomes harder to recall as time passes. Still, these lab results don’t translate into a universal workplace figure like "employees forget 70% of training," and shouldn’t be used as a direct measure of software-training retention.
Software skills fade for several reasons. Training often covers many steps across different screens, more than most people can keep in mind. Many tasks are rare, such as quarterly reports or exception handling, so the training may be months old when needed. Software interfaces change, making memorized steps outdated. Employees also juggle several applications, which can blur how similar tasks work in each system.
Designing Support for the Moment of Need
Instead of asking employees to memorize every step, learning and development (L&D) teams can focus on support delivered right when it’s needed. The Five Moments of Need model, created by Conrad Gottfredson and Bob Mosher, highlights five situations: learning something new, learning more, applying knowledge, solving problems, and adapting to change. Formal training covers the first two, but the last three happen during real work, when employees need quick, specific help. This model encourages pairing formal instruction with searchable procedures, embedded guidance, job aids, or help built into the application.
In-app guidance delivers support inside the software, offering step-by-step help at the point of use. Other useful tools include job aids, checklists, and short reference videos linked in the workflow. These resources let employees find what they need without digging through long courses or documents. On-demand guidance matters because software procedures are often conditional, involving certain screens, permissions, exceptions, and changing interfaces. Being able to pull up the right step at the right time matters more than recalling an entire training session from memory.
What to Teach and What to Support
Not every detail belongs in formal training. L&D teams can focus on the system’s purpose, core concepts, decision-making skills, and the most common daily tasks. Step-by-step procedures, rare tasks, field-level details, and information likely to change with updates work better as on-the-job support tools. This approach keeps training sessions shorter and more focused, helping employees remember key ideas and know where to find help when they need it.
Reinforcing Key Knowledge
For information employees must remember, spaced reinforcement works better than repeating full training sessions. Short follow-up prompts, scenario-based exercises a week after training, and brief manager check-ins help refresh knowledge before it fades. These activities should be quick and targeted, taking just a few minutes each time. The memory research by Ebbinghaus and later by Murre and Dros shows that review and relearning are essential for long-term retention, especially for knowledge not used right after training.
Measuring What Matters
Tracking attendance or course completion doesn’t show if employees can actually do key tasks. More useful measures include the time it takes for new users to complete tasks on their own, error rates on important forms, the number of support tickets for basic questions, and whether employees use the features the software was meant to provide. Watching these metrics in the weeks after a rollout helps L&D teams spot where more support is needed and improve future training. Operational measures like time to complete tasks, error rates, help-request volume, successful completion of rare workflows, and performance after interface changes give a clearer view of real-world proficiency.
Getting Started with Better Support
To start improving software adoption, review the most common support questions that came up after your last system rollout. For each question, consider if it could have been answered within the application at the moment of need and if it relates to a rarely performed task. This review can guide the creation of targeted support resources for future launches. Employees adopt new tools at different speeds; some will need ongoing support long after early adopters are comfortable. Recognizing these patterns helps L&D teams decide how long to keep support resources available.
Software training still matters for providing context and building confidence, but expecting employees to memorize every workflow isn’t realistic. By focusing on core concepts in training and offering just-in-time support for everything else, organizations can help employees use new technology more effectively when it counts.
