Clay Tech

"clay-works make things real"

translated from clazytech.com

Think About Method Later, Not First

Whether it’s product planning, business planning, promotion strategy, or anything else, one thing holds true across every case: think about the method later.

What happens if you think about method first

Discussing methods is a lot of fun.

If you’re an engineer, discussing implementation approaches, parts to use, UI design — these things make your brain boil with excitement. Throw in the full knowledge and experience of the team and argue back and forth, and it’s guaranteed to be exciting.

For business matters too, once you finish a spreadsheet with carefully calculated numbers for funding plans, staffing plans, and revenue forecasts, you get a huge sense of fulfillment even though the project itself hasn’t actually moved forward at all.

But in most cases, things don’t go that smoothly, and they don’t go as planned. At some point, the project is bound to run into a decisive either-or choice, or an issue where it’s unclear which answer is correct.

How should you judge things at that point?

A localized discussion almost never settles the matter. One choice ripples out to affect everything, and a single mistake can ruin the whole thing.

Now we’re in trouble.

“Why won’t this method work?” “Why won’t this layout work?” “Why won’t users bite on this feature?” “Why did we choose this user segment in the first place?”

Why, why, why comes flooding in.

And sometimes, after thinking and thinking and thinking, the conclusion ends up being to “flip the table.” The conclusion becomes: “We should scrap everything and start over.”

It’s an unfortunate outcome, but honestly, reaching that conclusion is still pretty good. Far more often, people choose to “force a conclusion out of the discussion on the spot despite a lingering sense of unease, and end up plowing ahead half-baked with their eyes closed.”

And that’s how the project dies.

I’ll leave what happens after that to your imagination. (I’ll explain it some other time if the opportunity comes up.)

Do the “whys” first

So to cut to the conclusion: throw out “Why?” and “Why?” as early as possible, as thoroughly as possible, to the point of being almost obsessive.

Let’s walk through an example to explain this.

The theme can be anything, but let’s say here we’re planning programming education material for elementary school students.

When it comes to programming material for elementary schoolers, Scratch is the obvious choice. There are more and more teaching materials, like micro:bit, where you can do hardware implementation with Scratch, so it seems like you could do all kinds of things starting from a Scratch base. Say, there’s an ESP32, which is trendy these days, and there are game emulators that run on M5Stack, so what if we used a wrapper that lets you implement that in Scratch, built a simple game with it, and made a curriculum where everyone plays together? Wouldn’t that be fun?

And with that, the planning is done.

This is a bad example. It’s all talk about products, services, and technology. It’s certainly exciting, but it’s completely unclear what problem exists and what gets solved. This kind of product trips up immediately the moment you actually try to “sell” it.

Sales rep: “What differentiates this product (curriculum)? What’s the pitch for selling it?”

You can’t give a clear answer to this essential question.

If the theme is programming education material for elementary schoolers, you really ought to first throw out the most fundamental “why” there is.

“Why do elementary schoolers need programming education in the first place?”

You might think, huh, all the way back to that?? But that’s exactly where real planning actually starts.

The fundamental “why” determines everything

I’m only using programming materials as an example here, so I don’t intend to dig all the way into that particular discussion in this post (that’s for another time).

It would also be fine to reach the conclusion here that “there’s no need to teach elementary schoolers programming.” The plan ends there, but if that’s the conclusion the team reached, it’s the team’s policy, and if it’s something the company’s management decided, it’s the business’s policy. Deciding “not to do” something carries just as much value as deciding “what to do.”

Let’s run a thought experiment assuming a different conclusion was reached here.

“Why do elementary schoolers need programming education in the first place?” – To maintain international competitiveness – To cultivate logical thinking skills

Listing too many would get unwieldy, so let’s think about these two for now. In an actual discussion, it’s preferable to list everything you can freely come up with. For example, “programming helps you pick up English” is also a good perspective.

Dig the “why” as deep as it goes

Let’s take up the “international competitiveness” argument. Apparently, even though Japanese math ability ranks internationally high up through the junior high school level, the number of PhDs in math and science fields is at a rather disappointing level (if you’re interested, look up the concrete numbers yourself). As an idea grounded in that kind of background, this seems pretty reasonable.

Now, this can still be dug further.

“Why does international competitiveness equal programming ability?”

Because an AI era is coming, and falling behind in AI technology means falling way behind industrially. If you can’t program, you’re not even on the playing field of competition, so it should be the single most important element of basic literacy.

“Is that actually true?”

Looking back at history, there was once an era when people hand-wrote HTML tags (the 1990s). That skill was valued at the time, but 30 years later, nobody gets the chance to type tags themselves anymore, and even if they could, it wouldn’t count as a skill at all — there’s zero competitive value in it. What matters far more is being able to properly think through the user story and experience (UX) of an entire website and bring in the right technology at the right time. Teaching this to elementary schoolers means it will be at least 10 years, maybe 20 years, before they can put it to effective use in the market. Can we really, correctly anticipate right now what will actually be needed in a market that keeps changing?

Suppose the discussion goes this way. Is the “competitiveness” side losing ground here? But it also seems a bit hasty to give up on the competitiveness argument at this point and swing sharply toward “well then, it must be logical thinking after all,” as mentioned earlier. Let’s keep digging.

Dig even further

We’d like to start converging soon, but let’s dig one more layer before heading back.

“Why is basic literacy necessary at all?”

If you have access to Excel, is it fine not to be able to do addition? If there’s spelling autocomplete, is it fine not to know the difference between “L” and “R”? If you can use CAD, can you design without knowing drafting technique? If there’s automatic translation, can you serve as an interpreter and close business deals across regions with different languages?

Suppose here you feel something like “it’s not logical, but something doesn’t sit right.” You still feel it’s better to be able to do mental arithmetic at a minimum, and that as a businessperson, you ought to be able to manage at least minimal conversational English on your own.

By this point, you may have reached something fundamental. Once you reach a layer where judgment is made not through logic or calculation but through reason and morals (or in some cases, intuition), digging further doesn’t accomplish much. (For a philosopher, this might actually be where the real work begins, but we’re practitioners here, so.)

Still, there are things you’ll struggle with once you’re out in society if you haven’t acquired at least a minimum of basic literacy, or things you’ll be glad you did once you’re out in society. So young people should acquire this as part of their intellect, you conclude.

The logical path by which you arrive here varies case by case, and the conclusion itself varies by person, by team, by company, so that part isn’t a major issue in itself.

Thinking about what already exists

The vaguely visible fundamental hypothesis:

“Acquiring, in childhood, the kind of literacy one ought to have as part of one’s intellect should ultimately help raise ability and boost international competitiveness, and programming may well fall under that category.”

Once you arrive at this, you need to think more analytically about how existing things measure up.

Again, whether this hypothesis is correct isn’t the topic here, so don’t get too hung up on it. We’re continuing to talk purely about the thinking method here. Also, from here on there will be somewhat more extreme claims (deliberately so), but that’s because a project can’t become sharp and distinctive without a certain amount of conviction or belief.

So, now, about the things that came up in the earlier bad planning example.

I think Scratch, with its approach of not being tied to a particular language and letting you define program logic by manipulating the UI, is extremely well thought out and wonderful. (But it won’t do just to keep praising it, so let me push back.) Couldn’t the following be needlessly raising the bar for basic literacy? – Requires one PC or tablet per student – Students have to learn how to use a keyboard at this stage And couldn’t the following be considered serious gaps? – Not touching on matrix operations at all doesn’t help develop AI talent – Skipping the process of writing a design document risks instilling unplanned thinking

A few hypotheses about pain points have come up here. At this point, you want some kind of backing for concerns like these. For example (note: the following is not based on fact):

“Once a typical public elementary school buys computers out of its budget, it can’t replace them for 10 years.” “China ranks among the world’s top in AI-related papers, but only about 5% of children in the early elementary grades have ever touched a keyboard — not so different from Japan.” “The latest neuroscience research has revealed a major difference, between people accustomed to matrix operations and those who aren’t, in whether they can process tasks in parallel mentally — that is, in whether they can handle complex tasks once they become working adults.” “A worldwide longitudinal study spanning the last 40 years has statistically shown a correlation in which learning programming in the early grades lowers the rate at which people go into jobs requiring long-term planning, such as architecture, urban development, or policy.”

And so on. The above is pure fantasy, but if even one of these turned out to catch on as fact, wouldn’t something like,

“This can’t go on as it is! We have to change the world!”

start to well up?

That’s Pain, and that’s Passion.

Get hold of that, and the plan is as good as made.

In the end, we still haven’t thought about method

We’ve come a fairly long way by this point. Suppose, as a result, a planning concept like the following takes shape.

Programming is important as basic literacy, but preparing equipment is a major burden. This isn’t a problem that can be solved at the level of national or local government budget control. Also, learning how to use a PC isn’t really the main point — present-day teens are already building a world where they handle everything on their smartphones. We don’t know what the future holds. Given that, if we develop an educational tool that lets students acquire programming-style thinking like Scratch does, without using a computer, and introduce it to schools cheaply, it could merge with existing programming education curricula and become a strong product that can compete globally. It could also use its low price, durability, and lack of need for a power source to help spread programming education in developing countries.

I think that’s a good product plan.

Up to this point, there’s been zero thought given to implementation method. But it’s still a good plan. You can picture whose pain gets solved and how it makes the world happier once this is realized.

A curveball idea like the following might also be good.

In the current education curriculum, matrix operations are taught under linear algebra — that is, only as a category learned by people who go on to a specific field after entering university, not something general. This is the root cause behind Japan losing international competitiveness across industry as a whole, not just in math, science, and AI. So children should be properly taught matrix operations, macros, and functions from an early age, but it will take an extremely long time before this gets included in the public school curriculum guidelines. So first, propose a curriculum to private cram schools that don’t focus heavily on entrance exams, and introduce it along with a self-developed tool that can measure learning outcomes. Once the business grows, establish a dedicated cram school and vocational school of our own. Ultimately, aim to gradually bring in school boards and get it adopted into the official education curriculum.

This is a fairly outlandish idea, and honestly I don’t know whether it’s right or whether it would really succeed, but I think it’s an interesting plan. It makes you think, “I want to know how this turns out.”

Once a plan has been properly set up like this, all sorts of ways to actually realize it just start coming to mind.

Summary

That’s it — this was a story about the most basic of basics.


Originally published in Japanese at https://clazytech.com/2020/10/377/. Translated with LLM assistance and reviewed before publication.