Case study

Corporate pilot: Soft skills training with 80 programmers

By WiseWorldCorporate pilot, 80 engineers, 3 months

Software developers collaborating in a team, soft skills training case study

A three-month WiseWorld pilot inside an engineering team put soft skills into the hands of 80 programmers, five minutes a day, no workshops, no homework. Here's what the data tells us about engagement, the skills that came up, and what made it stick.

Path completion
297%

Of baseline expectation

3-month retention
83%

Still active in final month

Training volume
~400h

No workshops scheduled

Skills tracked
44

Across 6 dimensions

The team & the friction

An engineering team that ships fast, collaborates across functions, and was tired of one-and-done workshops. The brief was simple: improve how they worked together without disrupting sprints or adding recurring meetings.

Team setup

  • 80 engineers
  • Cross-functional dependencies
  • Fast sprint cycles

The friction

  • Limited meeting time
  • Constant context switching
  • Hard to sustain training

The brief

  • Build a habit, not an event
  • Short, frequent practice
  • Measurable progress

Same calendars. Same sprint pressure. One shared experiment in growing without a workshop.

What had been tried

Workshops had given a burst of energy and almost no lasting practice. Skills faded within weeks, there were no measurable outcomes, and appetite for the next round was low. So we tested the opposite shape: short, daily, self-paced.

The old way

Two intense days, then silence.

  • 1–2 day events, packed and exhausting
  • Skills faded within weeks
  • No measurable outcome
  • Low appetite for the next round

The bet

Five minutes. Every weekday. For three months.

  • 5-minute AI scenarios slotted into the workday
  • Self-paced - no new meetings, no homework
  • Goal-based learning paths per person
  • Activity tracked across 44 skill dimensions

The bet, designed

A three-month program of weekday micro-practice (~5 minutes per day), targeting 44 soft-skill dimensions across 6 main categories. Two things to lock down before the data arrives: how it's designed, and how we'll know it worked.

Design

  • FrequencyWeekdays, self-paced
  • ModalityAI real-life episodes
  • PersonalizationGoal-based paths

Measurement

  • ActivityDialogues completed
  • RetentionFinal-month activity
  • ProgressPath completions

What people chose to work on

Given the choice, half chose growth at work and a third chose something more personal. The split says a lot about what people actually want from soft-skills practice when nobody is assigning it.

Set your learning goals - AI identifies the skills you need
~50%

Career & work growth

Leadership, tricky situations, visionary thinking, career development

~32%

Personal life

Feeling calmer, more confident, better focus, work-life balance

~18%

Connection

Social skills, deeper conversations, teamwork, communication

Skills that rose to the top

Out of 44 available skills, six rose to the top - the everyday muscles of working with humans: how to listen, decide, speak up, and bounce back.

Training episode - Practice targeted soft skills
1

Active listening

Communication

2

Public interaction

Communication

3

Confidence

Leadership

4

Decision making

Problem solving

5

Time management

Work ethic

6

Resilience

Work ethic

Notice what's missing: nobody asked for personality overhauls. They wanted small, specific, daily-life skills.

Show-up & output

Daily activity held near 47% across the full window. By month three, 83% of the team was still active. Together it added up to ~400 hours of practice and learning-path completion at nearly three times the baseline.

Daily activity
47%

Active on a typical weekday. Steady, not a spike.

3-month retention
83%

Active in the final month - when most courses lose people in week two.

Training volume
~400hours

Across 80 people. Zero workshops scheduled.

Weekly rhythm
5.2episodes / person

The quiet rhythm that did the work.

Skills per user
~15of 44

Average breadth covered.

297%of baseline learning-path completion

Nearly three times the typical rate - not from pressure, but from a format people could actually keep up with.

When people got stuck

During the episodes, participants could ask Wise, the AI soft-skills coach, for a hint. They did so 2,396 times over three months - about 30 per person. A safety net, not a teacher.

AI hints requested
2,396

≈ 30 hints per person across 12 weeks

What the number means

People want to figure things out themselves - but they want help available when they need it.

Not a teacher hovering over their shoulder. A safety net that catches them only if they reach for it.

AI Coach chat - Get real-time guidance and learning hints

What managers saw

Soft skills used to be the hardest thing to talk about in a review. With the team PowerWheel, managers finally had a picture - a shared view of progress across all 44 skills.

  • Vague feedback became data.

    “Be a better communicator” became a tracked, visible skill.

  • Team strengths became shared.

    Managers could see where the team was thin - and where it had grown.

  • Bias dropped. Signal rose.

    An unbiased view of progress, refreshed in real time.

Team PowerWheel showing strengths across 44 skills

Team coverage - sample of 8 skills (0–100)

Three rules for the next team

Three rules turned a tired format into a habit people kept - and the data backs each one.

01

Keep it short.

Five minutes survives a sprint. Fifty doesn't.

Daily activity ≈ 47%

02

Let people choose.

Goal-based paths beat one-size-fits-all every time.

Goals split 50 / 32 / 18

03

Make progress visible.

People keep going when they can see themselves move.

Completion: 297% of baseline

The bottom line

Engineers will invest in growth. They just won't invest in workshops that disappear the moment they end.

80

engineers

5

minutes a day

83%

stayed the course

297%

of baseline