On February 27, 1998, Apple announced that it would stop developing the Newton, its handheld computer, along with the two devices built on it. “To realize our ambitious plans,” said Steve Jobs, the company’s interim chief executive, “we must focus all of our efforts in one direction.”
Six weeks earlier, he had named the direction. Apple had lost more than $1 billion in its last fiscal year. “Returning Apple to sustainable profitability,” Jobs said in January, “is the Company’s number one objective for fiscal 1998.” The full story is in the case.
The goal that won’t happen by itself
Pick the one or two goals that matter most, and give them attention daily work won’t.
The phrase “Wildly Important Goal,” or WIG, comes from The 4 Disciplines of Execution (2012), by Chris McChesney, Sean Covey and Jim Huling of the consulting firm FranklinCovey. It is the first of their four disciplines, and the other three start from it.
Their starting point is a complaint most managers will recognize: “There will always be more good ideas than there is capacity to execute.” Meanwhile there is the day job, which they call the whirlwind. A WIG is not a list of everything that matters. In FranklinCovey’s words, it is “the most important objective that won’t be achieved unless it gets special attention,” because “your normal course of business won’t make it happen.”
They don’t ask teams to drop the daily work. They ask them to pick, alongside it, what they most want to improve, and to define it sharply: “a starting line, a finish line, and a deadline.” From X to Y by when. And they want “a single measure of success.”
Why so few? FranklinCovey’s claim is that a team’s chances of achieving two or three goals with excellence are high, and that the more goals it juggles, the lower they fall. It also reports, from its own research, that only 15 percent of employees know their organization’s most important goals.
Apple’s one direction
Apple didn’t use the 4 Disciplines, which came out fourteen years later. But its 1998 reads almost like an illustration of them.
The starting line was a loss of $1.045 billion. The finish line was a profit that would last. The deadline was the end of fiscal 1998. And much of what Apple did that year can be read as clearing space for that one goal. It went “from approximately 15 separate individual products to three main product families,” its annual report said. Research spending fell 38 percent, with “the cancellation of many research and development projects.” The Newton went.
Apple made money in every quarter of the year, $309 million in all. Its annual report spread the credit across several things, including a restructuring begun in 1996 and better inventory management. On this reading, the goal didn’t do the work alone, but it told the company what to stop doing.
Ten years later, Jobs described focus to Fortune in a way that matches the WIG idea closely. “People think focus means saying yes to the thing you’ve got to focus on,” he said. “But that’s not what it means at all. It means saying no to the hundred other good ideas that there are.”
What saying no costs
Those hundred other ideas are usually good. That’s what makes this hard. Apple’s previous annual report had warned that a simpler product line would leave it depending “on fewer products.” Its sales kept falling through 1998, to $5.9 billion. A WIG concentrates risk as well as effort.
Imagine a twelve-person bakery that names its WIG: raise weekday lunch sales from $800 to $1,200 a day by June. The owner shelves a wedding cake line and a new website. If lunch doesn’t move, the bakery may have lost half a year on the other two ideas as well. If it does move, the team will have a clearer idea of what worked.
There is a quieter way to fail. A goal can be announced once and then lost in the whirlwind. In the 4 Disciplines, naming the goal is only the first step. The next three are about the few actions that move it, a scoreboard everyone can see, and a weekly meeting where people report on their commitments.
Focus in the age of AI
An AI agent has its own whirlwind. Anthropic’s guide to its coding tool, Claude Code, says that most of its advice comes from one constraint: the agent’s context window, its working memory, “fills up fast, and performance degrades as it fills.” One failure it names is the “kitchen sink session,” where a person keeps adding unrelated tasks to the same conversation until the context is “full of irrelevant information.” Its fix is to clear the context between unrelated tasks.
Piling on demands has a measurable cost. In a 2025 benchmark called IFScale, researchers asked models to write a business report while following up to 500 instructions at once, each requiring a particular keyword to appear. Performance degraded as the instructions piled up, and even the best models met only 68 percent of them at the top density. That is one narrow kind of instruction, not real work.
The same guide also suggests a version of the finish line. “Claude stops when the work looks done,” it says, so give it “something that produces a pass or fail,” such as a test suite. That carries over the part of a WIG that is easiest to transfer: a clear target the worker can check without asking you.
The analogy stops there. An agent doesn’t get bored or lose morale. The reason to narrow its goals is mechanical, about what it can hold and check. For a team, the reason is also about where people put their energy.
In 2008, Jobs said Apple had been “pressured for years to do a PDA” and chose not to. “If we had gotten into it, we wouldn’t have had the resources to do the iPod.”