Hey folks, it’s your friendly neighborhood UX designer here in Topeka, Kansas—where the wind blows harder than most service handoffs. I’ve spent years knee-deep in customer journeys, sticky notes, and the occasional existential crisis at 2 a.m. wondering why that “simple” process redesign still feels like herding caffeinated cats.
Then in my Master's program at Boise State (Go Broncos!) I stumbled on Thomas Gilbert’s Behavior Engineering Model from the ‘70s (yes, the guy who basically invented human performance tech before it was cool). And holy service blueprint, Batman—it maps perfectly onto service design. Suddenly everything clicked: why some interventions feel like permanent magic while others are just expensive band-aids that keep falling off.
Gilbert didn’t mess around. He built a 2×3 matrix that screams: Fix the environment first. It’s cheaper, faster, and actually sticks. Only then do you tweak the humans (because let’s be honest, people are squishy and forgetful).
Here’s the original Gilbert quick-hit for context (no PhD required):
Environmental Supports (the “system” side – do this first!)
- Information: Clear expectations, feedback, guides.
- Instrumentation: Tools, processes, resources.
- Incentives: Rewards, consequences, carrots/sticks.
Person’s Repertory (the “human” side – use sparingly)
- Knowledge/Skills: Training, education.
- Capacity: Ability, selection, physical/mental fit.
- Motivation: Intrinsic drive, attitudes.
Gilbert’s big mic-drop: Most organizations obsess over training (the sexy, visible fix) when the real problem is a broken system yelling “I’m not designed for this!” at everyone.
So I did what any self-respecting UX nerd would do: I stole it, re-skinned it for service design, and turned it into The Service Design Intervention Matrix—my love letter to Gilbert with a side of snark.
The Service Design Intervention Matrix
(Or: “How to Stop Training People to Swim Upstream”)
I kept the six boxes but translated them into the language we actually use in workshops, stakeholder decks, and late-night Slack rants. Each category includes when it’s the best option (context + criteria) and a real-world “don’t be that guy” example.
1. Information Interventions (Environmental)
What it is: Redesign how expectations, feedback, and guidance flow through the service—dashboards, signage, error messages, journey maps that actually get used. Best when: Confusion is the enemy and you want permanent clarity for every future employee and customer. Cheap, instant ROI, zero retraining. Example: Instead of teaching baristas to “read the customer’s mind,” you redesign the POS system with one-tap modifiers and visual cues. Boom—fewer wrong orders forever. Pro tip: If your service blueprint has more question marks than arrows, start here.
2. Instrumentation Interventions (Environmental)
What it is: The big guns—process redesign, tool/tech upgrades, automation, physical space tweaks, API overhauls. Permanent structural changes. Best when: The workflow itself is the bottleneck and you need scalable, lasting efficiency. Ideal for high-volume services or when growth is killing you. Example: Airlines that keep training gate agents on “how to calm angry passengers” vs. the ones that redesign boarding with digital queuing and clear signage. One costs salaries; the other costs a one-time dev sprint. Pro tip: This is your “permanent change in the system” hero. Prioritize it over training 9 times out of 10.
3. Incentive Interventions (Environmental)
What it is: Rewire what gets rewarded—KPIs, bonuses, recognition systems, policy nudges, service-level agreements. Best when: People know what to do but have zero reason to care (or are actively punished for doing the right thing). Example: Call centers that stop measuring “average handle time” (which rewards hanging up on grandma) and start measuring “customer delight score.” Suddenly agents become humans again. Pro tip: Incentives are the silent killers of great service design. Fix them early or watch your fancy new journey map collect dust.
4. Knowledge/Skills Interventions (Individual)
What it is: Classic training, workshops, job aids, e-learning, onboarding playbooks. Best when: The system is already solid and there’s a genuine skill gap (new tech, new regulations, highly specialized roles). Use as a supplement, never the main dish. Example: Only after you’ve fixed the clunky CRM do you train reps on its advanced features. Otherwise you’re just teaching them how to lose slowly. Pro tip: Training feels productive. It’s also the most expensive way to paper over bad design. Gilbert is side-eyeing you right now.
5. Capacity Interventions (Individual)
What it is: Hiring the right humans, adjusting workloads, ergonomic setups, accessibility tweaks, role redesign. Best when: The people literally can’t physically or mentally keep up (burnout city, wrong skill mix, accessibility fails). Example: Instead of “more training” for overwhelmed nurses, redesign shifts and add AI triage tools. Capacity first, compassion second. Pro tip: Great for mature services where the system is good but the humans are maxed out.
6. Motivation Interventions (Individual)
What it is: Culture work, purpose storytelling, recognition programs, psychological safety initiatives. Best when: Everything else is fixed and people still don’t want to deliver great service (rare, but it happens). Example: Only after the tools, incentives, and processes are delightful do you run the “find your why” offsite. Otherwise it’s just corporate karaoke. Pro tip: Motivation is the cherry on top. You can’t sprinkle it on a broken sundae and call it gourmet.
The Golden Rule (Steal This for Your Next Deck)
Start left to right, top to bottom. Nail the three environmental boxes first. They deliver permanent, organization-wide wins at lower cost and drama. Only then dip into the individual side—and even then, treat them like seasoning, not the main course.
I’ve watched this model save projects that were circling the drain. One client was ready to blow $250k on enterprise-wide training. We spent $40k redesigning two core processes and one incentive rule instead. Training budget? Slashed to $15k for a quick job-aid refresh. Results? Measurable, sticky, and the CEO still sends me memes about it.
So next time your stakeholder says “We just need better training,” channel your inner Gilbert and reply: “Cool story. But have we checked if the cage is designed for the monkeys to succeed?”