Correct On Paper, Useless On A Tuesday

The problem with telling people to eat better

A few months prior to this workshop, I had already run another run-of-the-mill nutrition workshop. Stand at the front, explain what people should eat, watch them nod and take notes and leave with more information than they arrived with. While it may work and nobody complains, it objectively sucks because nothing changes.

At another client's request, I also tried the hands-on version instead — a sandwich-making workshop. The ingredients were prepped by a caterer alongside a recipe for each table, everyone assembles their sandwich, and everyone feels good. I followed up two months later.

Nobody was making the sandwich.

Making a sandwich does not require extraordinary skill. The workshop failed because the actual work was done for them. Cooking isn't the skill people are missing. Rather, deciding what to cook, under real-life constraints, with what's actually in the fridge, is. Hands-on doesn't mean transferable. It just means entertaining.

Solving a mission with a fake client profile

In my years as a nutritionist, I learned something no one talks about: the moment you ask someone to design their own healthy meal, people become aspirational. At best, they'll keep up with their aspirations. At worst, they find the "correct" answer and perform it for you. So I gave them someone else's problem instead.

The pseudo-client I gave them was a 35-year-old software engineer, working long hours, protecting their sleep and their gym time, but surviving on instant noodles and delivery because actually cooking felt like one more decision they didn't have capacity for. (I asked questions about the audience prior to designing the workshop, so the profile matched roughly 80% of the room, but they were not explicitly told so.)

The assignment was to design a dinner this person would actually make. RM50 grocery budget ^[1], a real grocery delivery app, cost per portion, and it has to fit around the client's routine.

Distance from the judgment removed the necessity to be vulnerable with one's personal dietary habits and aspirations. Solving a stranger's problem means you don't have to perform virtue. Subsequently, people in the room were excited to problem-solve.

How I ran the workshop with little to no guidance

The workshop ran in four phases: Hook, Immerse, Apply, Connect.

Hook — Ingredient Forensics

Teams pulled apart food labels hunting for misleading claims. I don't expect everyone to source raw ingredients on their own, so being more informed of grocery products was a huge lever here. In Ingredient Forensics, there were numerous products dressed up to look healthier than they are (there's a special place in hell for "creamer" doing its best impression of milk, or margarine for butter). The only guidance they had was a very rough "read the ingredients list and cross-check the claims." They had to make the effort to learn marketing (the front cover) versus information (the ingredients label), and because it was a minigame with competitive stakes, they learned quick.

Immerse — the client and the constraints

This handed them the client profile, an RM50 budget, and an ingredient list they could choose from — familiar Malaysian carbs, proteins, and vegetables, deliberately limited to the Malaysian market so no one has to go to a specialty organic store to get an obscure powder. The point of the limitation was to prove the power of math (3 carbs, 3 proteins, and 3 vegetables gives you 27 unique dishes) and that using simple and familiar ingredients would more than suffice.

Apply — I stop teaching

Apply means I don't need to continue teaching, because the assignment meant groups needed to pull out their phones, open real local grocery delivery apps, compare prices, argue about trade-offs, and decide what was worth the money. I gave them a scoring rubric and portions per meal to calculate. However, they did not get a recipe from me, since handing someone a recipe is just giving the answer with extra steps, which defeats the purpose of the workshop. I wanted participants to practice finding, instead of executing, other people's answers.

Connect — the defense

Every group had to present their recipes, budget, and cost per meal, and I judged on whether the meal fit the client's life. In the end, I asked the million-dollar question: would you make this yourself?

The group that got the nutrition right and the assignment wrong

One group built a quinoa chicken rice meal and put roughly 40% of their RM50 budget into the quinoa alone. Their logic: more protein = better meal. In fact, their dish contained about 60g of protein ^[2], which sounds like an incredibly optimized meal on paper. However, context matters. In this assignment, 60g of protein was nowhere near what the client actually needed.

The client already had rice at home and was just done with a full day of work. He was not a macro-tracking athlete short on protein. The meal didn't need to be optimized. It needed to be something he could realistically cook in twenty minutes without looking up "how to cook quinoa" on a Tuesday night. That group scored in the 80s out of 150 because they solved a problem the client didn't actually have.

Meanwhile, one group built a full dinner for roughly RM2 a portion, spending about RM17 of their RM50. They didn't win either, because I asked why didn't they spend at least 70% of the budget — maybe to buy higher quality proteins (they only had tofu and eggs)? Or to buy more to meal prep for the week? They would've won if they managed to use their resources appropriately. However, they definitely answered: yes, this is something I, and the client, can actually make.

Finally, we had one group who injected a lot of humor into the session (by deciding the client would buy expensive bottled mineral water since they had leftover money), and created 2 different recipes with the ingredients they had, at around RM7 per meal. Although this group's macros and budget weren't perfect, I still scored them almost 120/150 because they understood keeping something sustainable required variety, some little treats (expensive mineral water), and using resources correctly — even though they appeared unserious, and they were shocked they won too.

What 32 engineers taught me about teaching

The average score for the assignment landed around 110 out of 150. Two groups pushed close to 120 by treating the client's actual life as the constraint that mattered instead of trying to optimize for the "most nutritious" or "the best" meal.

I don't have data on whether anyone kept doing this three months later; I won't pretend I do. What I can say is that the room shifted, mid-session, from passively listening about nutrition to arguing about it. The act of comparing prices, defending trade-offs, catching each other's blind spots taught them more than I'd ever hope to laser-pointing at a slide deck. One of them even told me the lesson felt "too simple." Good. That was the point. Simple only feels obvious after you've had to apply it under a budget and a deadline.

The regular lecture, and the hands-on workshop, taught me that if you stop handing people the answer, and hand them someone else's problem instead, you can get out of the way while they solve it. And they can solve it really, really well.