返回资讯中心
外部精选
软件工程
#开发者实时流

It's not a strategy if there are no tradeoffs

All day, everyone at the company makes decisions, independently:

Hacker News 最新Curiositry2026/10/16 · 2026/10/1629 分钟阅读

以下正文同步自 Hacker News 最新,版权归原站所有,已转换为易读排版。

All day, everyone at the company makes decisions, independently:

How to respond to this customer. How much effort to put into this feature. What we build next. How we position a weakness on the website. Where next month’s budget goes. Which marketing channel we bet on. What ideas we even bother to consider.

Everyone had better be making independent decisions. That’s how you scale⁠—whether three people or three hundred; even if it’s just you. It’s how you manage autonomous teams, pushing power down to the individual (but not too far).

But even perfect individual decisions can easily be inconsistent between teams: decisions can be locally rational yet contradict decisions made elsewhere. If everyone is moving in different directions, the company goes nowhere, even if everyone is productive and efficient.

Even autonomous teams require alignment. Not alignment on tactics (those change weekly), nor on platitudes (those change nothing), but alignment on a small set of macro-level decisions. These are the strategic choices.1

1 A term popularized by Roger L. Martin, who has much more to say in his 2013 book Playing to Win: How Strategy Really Works and in articles like this. Indeed, he defines the whole concept of strategy as a set of choices: “Strategy is an integrated set of choices that uniquely positions the firm in its industry so as to create sustainable advantage and superior value relative to the competition.”

When the daily decisions conform to the strategic ones, “more people” actually can produce “more results.” Customers stop seeing dissonance between the promises marketing made on the homepage and what engineering built. Decisions made independently across the company reinforce each other, which means they compound instead of diluting each other.

Most “strategy” documents fail because they repeat platitudes instead of making decisions that require trade-offs and sacrifice. They read like moral aspirations: “Customer-first.” “Move fast over bureaucracy.” “Quality over shortcuts.” “Great design over mediocrity.”

Those aren’t wrong per se; they’re just non-decisions. No one says, “Our strategy is to have low quality and ugly design,” yet mediocre products exist. Why? Because companies chose other priorities⁠—price, speed, extensibility, power, complexity⁠—and the trade-offs showed up as low quality or uninspired design. These underlying choices⁠—saying “no” even to things that are universally desirable like quality and design⁠—are the hallmarks of real strategy.

Your unique set of choices makes you “the best” according to your target market. A strategy that never says “no” is not a strategy. It should say “no” most of the time; that’s a signal that you are actually focusing.

Great strategies include strategic choices having these properties:

Next I’ll explain each property, share a few facilitation tips, and end with a table of examples to spark your own discussions.

“Good design” isn’t a real choice because “bad design” was never a smart alternative.

“All-in-one” is a real choice, because an alternative is “A small core, extremely extensible”; both are smart. You know this because many successful companies, with great products, have made each choice.

Notably, “both” is not a sensible choice. Either the product is self-contained or it isn’t. Either you design the UX as a complete experience, or you design it to accept third-party UX in ways the original designers couldn’t predict. Either you sell “the only thing you’ll ever need to buy,” or you sell “the platform with an ecosystem for everything.” Either it “does a few things very well,” or it “does everything but you piece it together.” Either support can answer questions about the whole product, or support has to route you to plugin vendors who own the pieces you’re asking about.

You know you’ve made a real, strategic choice when there’s a long list of second-order effects that permeates the entire company.

This is another form of the Opposite Test. Once again, ensuring that “the opposite” isn’t trivial or nonsense, is how we know we’re saying something real and important. It’s true of marketing too.

A platitude feels good because it has no downside. How could “great design” be bad? Customers love it, developers are proud of their work, and it reduces support.

If it’s all good news, it isn’t a real choice, or you aren’t grappling with the actual consequences.

“Great design” comes with costs. You need talented designers, which means a larger team. You need engineers who are excited to craft the UX with as much care as they scale the database and want to chase down every last UI bug before adding the next shiny new feature. You invest in design systems. Features and changes take longer because they are required not only to work, but to be well-designed. You say “no” to features customers want if it means the design would worsen. Even then, not all customers agree on what “great design” means, so you thrill some people while irritating others.

A strategic choice is a package deal: you don’t get advantages without weaknesses.

When you list the consequences, writing them into the strategy, the decision becomes usable by the whole company. You reduce individual interpretation and you help everyone align their daily decisions.

This also helps you ensure the entire strategy is internally consistent.

Use the list of choice-consequences to check for contradictions. Conflicts often hide in the second-order effects.

For example, “small team” and “feature-rich” might look unrelated, but keeping a team small usually means keeping the product small and simple, because the team has to build it, maintain it, support it, and evolve it. That’s in conflict with “feature-rich.”

But perhaps they’re not in conflict after all if we make yet another strategic choice: a small team can maintain a small core that is designed for third-party extensions. Then, an ecosystem of plugins and APIs makes the entire product feature-rich after all! Now the choices interlock: “small team,” “small core,” and “extensible ecosystem.” Each drags along its own implications, which surface more choices, and so on. This exploration is how you arrive at a coherent set of decisions.

Even better than “not conflicting” is “reinforcing.” “Extensible” helps the product be “feature-rich.” A small team forces the core to stay small, maintaining space for the ecosystem to thrive. Each decision makes the next easier and stronger.

When strong choices reinforce each other, the whole company starts to feel like it has a personality⁠—an opinion about How The World Should Be. Some customers won’t share that opinion, but the ones who do will love you all the more. Love is how you win.

And everyone at the company can move faster. You stop re-litigating trade-offs. Marketing and sales stay aligned with the product. People have more fun, because coherence feels like competence⁠—and competence feels good.

When you’re facilitating the creation of strategic choices, you will find that people constantly write weak pseudo-choices that fail the Opposite Test. They already know which side they prefer, so they’d rather make the other side obviously bad⁠—so their choice “wins”⁠—instead of engaging in the struggle of making firm, clear decisions.

There’s probably a valid perspective hiding under that weakly-worded choice, so your job is to extract the underlying choice that captures that perspective. This can help:

What are some successful and beloved companies that chose the “bad” side?

What is the underlying reason why people love that company despite that flaw? What is the thing they love so much, that even though it has this bad consequence, they still love that product?

That “thing they love” is actually the other side of the strategic choice.

For example, “well-designed UX, not bad UX” is a fake choice. So ask: what are some reasons that some beloved products have a poor UX? We looked at one reason already: the product is a small core with an ecosystem of plugins and extensions. That flexibility lets third-party software “do anything,” which lets customers “do anything,” and that’s genuinely wonderful… but at the cost of UX coherence.

A real-world example of this is note-taking apps. All-in-one apps like Apple Notes and Bear are beautifully cohesive in part because they control the entire experience⁠—no plugins, not even an API. Meanwhile, competitors like Obsidian offer an incredible array of downloadable functionality. A common complaint is that the UX becomes a hodgepodge, but people love the expansive choice of features and the ability to build their own, custom, ideal information management system. By asking “Why is Obsidian so popular and beloved despite its UX being so messy,” you arrive at the real strategic choice, which is “all-in-one vs infinitely extensible,” in which design is one important consequence.

Since both sides are desirable, we are tempted to start pulling in facets of the other side. You’ll build an all-in-one product, but customers love that a competitor has plugins, so you’ll start exploring what it would mean for you to add plugins. Or: “We should raise the price; we’re leaving money on the table.” Or: “We should slow down and polish this; it’s embarrassing.” Those temptations feel compelling precisely because the other choice is also smart. Strategy exists so that anyone in the company can confidently say: Yes, that’s a good idea in the abstract, but no, it doesn’t fit the bundle of trade-offs we have chosen.

Still, sometimes it’s reasonable to do both. Suppose you’ve chosen “lowest price,” which often implies lower quality, because low prices mandate cost constraints. But sometimes you can improve quality without increasing cost⁠—better design choices, better suppliers, advantages of scale, great employees. Shouldn’t you take those wins?

The rule is: When in conflict, adhere to the strategic choice. When not in conflict, do whatever is best. When quality and cost don’t conflict, take both.

But usually they do conflict. In those moments, you can’t satisfy everyone. Some customers want one side of the trade-off; others want the other. If you don’t decide consistently, you land in the only truly losing position: not enough quality and design to earn delight and loyalty, but also not enough power and breadth to win complex use cases. You lose both markets, because you claimed “it’s a balance” and created something no one is excited to use.

The Agile Manifesto epitomizes this attitude. “Working software over comprehensive documentation” doesn’t claim documentation is bad; it says that when time is limited (and it always is), you invest more in one than the other. That’s not dogma; it’s a mechanism for consistency under pressure. They even underline their intent: “While there is value in the items on the right, we value the items on the left more.”

Use the same pattern for strategy. If your strategy is low price, you don’t introduce a high-price offering “to get some of those customers.” But if your low-price strategy implies lower quality, you happily take higher quality whenever it doesn’t threaten the low price decision.

It remains a strict choice, not a balance.

You could incorporate a few of the following into your strategy, and use them to prompt even better ideas:

The weaknesses of your choices are painful to accept. The advantages of the opposite choices sound wonderful. Welcome to strategy, where you make hard choices.

If the choice wasn’t hard to make, it wasn’t actually a choice.

Make the hard choices now, relieving everyone else of that burden, and allowing them to make smart decisions independently.

Click here for AI Skills that help implement some of these ideas.

正文由 探How 从来源站点 RSS 同步,内容未经改写;遇到排版缺失或需要图片、视频时请以原文为准。

打开原文
去社区讨论 →