Sonar
The Sonar model

We borrow what works.
We're honest about what doesn't.

Sonar fuses four lenses — Prosci PCT, ADKAR, the layered Saarinen model and the emotional curve — into one continuously-updating picture. Each lens earns its place. The bits that don't hold up under research, we say so.

Prosci Change Triangle (PCT)

What it is

Four corners — Sponsorship, Delivery (project), People (adoption) and Success — that have to stay roughly in balance.

How Sonar uses it

Sonar renders the triangle as a live SVG. When one corner weakens, the shape visibly skews — imbalance is visible the second it appears.

Evidence: Prosci's 'Best Practices in Change Management' study (multi-year, 2000+ projects) shows projects with strong active sponsorship are ~3× more likely to meet or exceed objectives.

ADKAR funnel

What it is

A sequence of individual states — Awareness, Desire, Knowledge, Ability, Reinforcement — that any person must traverse for change to stick.

How Sonar uses it

Sonar runs an ADKAR funnel per cohort, highlights the FIRST broken block, and routes the recommendation to that block (not three blocks downstream).

Evidence: Prosci ADKAR® model, Hiatt 2006. Widely validated across digital transformation studies (McKinsey 2023, Deloitte State of AI).

Layered (Saarinen) model

What it is

Role (chair) → Team (room) → Function (house) → Org & Strategy (city). Each layer's needs and signals are different.

How Sonar uses it

Each participant rates their OWN layer AND the layer above (mirroring how Aiware measures AI leadership). This catches 'starved from above' and 'hollow below' patterns that single-layer surveys miss.

Evidence: Adapted from Eliel Saarinen's nested-context principle. Layered measurement aligns with how transformations actually break — typically at the seam between two layers.

Kübler-Ross / Bridges (signal, not script)

What it is

An emotional overlay — shock, denial, frustration, experiment, integration / endings, neutral zone, new beginnings.

How Sonar uses it

Sonar uses curve position as a SIGNAL, not a rigid stage model. The most useful flag is 'cohorts stuck in the neutral zone' — that's where transformations regress.

Evidence: Bridges, 'Managing Transitions' (2009). We deliberately avoid treating Kübler-Ross as a strict sequence — that misreads the original grief research and ignores the modern critique (e.g. Stroebe et al.).

What about Kotter's 8 Steps?

What it is

A famous prescriptive sequence — create urgency, build coalition, vision, communicate, empower, short-term wins, sustain, anchor.

How Sonar uses it

Useful as a checklist of executive moves, but treated by Sonar as a menu — not a sequence. Real transformations don't unfold in 8 neat steps; they oscillate.

Evidence: Kotter 1995/2014. Honest caveat: there is little controlled evidence the strict 8-step order outperforms other framings. We use it as a vocabulary, not a script.

Why merge them?

No single model captures the whole picture. PCT shows the shape, ADKAR shows the funnel, the layered model shows where the break is, and the emotional curve shows how people are coping. Sonar fuses these into ONE composite Change Adoption Score so you can act on a single number — without losing the texture underneath.

See how the loop runs →