Ahead, Apart, Against

Andy Grove looked out the window in mid-1985. Great America's Ferris wheel turned in the distance. He asked Gordon Moore what a new CEO would do if the board replaced them.

Moore answered without hesitation: "He would get us out of memories."

The budget: ahead

A gap between a plan and the work can look like failure. It can also make you wonder whether anyone meant the plan. But Intel's work had moved toward an answer its leaders still struggled to accept.

Middle managers had been shifting silicon wafer capacity toward more profitable products, including microprocessors. An operating rule favored the products earning more gross margin per wafer. Each allocation moved the business, without a new strategy from the top.

By the time the leaders decided to leave DRAM, Grove writes, only one of eight fabrication plants still made memories. The allocation record showed the shift before the official strategy did.

The old identity held on. As Grove put it, "memories were us." The chips were products, but they were also how Intel understood itself.

Joseph Bower says strategy is what adds up from the ideas people bring forward and the ones that get funded, "not words on paper." At Intel, those choices had run ahead of the story.

On October 10, 1985, Intel announced its withdrawal from DRAM. It kept EPROMs, another kind of memory. Grove later called leaving DRAM "the best business decision we ever made."

The calendar: apart

Intel had a rule moving its wafers. An executive's day may have no rule connecting time to the stated job.

Henry Mintzberg found that managers described their work as planning, organizing, coordinating and controlling. Then he watched them.

With a stopwatch, he followed five chief executives for one intensive week each. They led a consulting firm, a technology company, a hospital, a consumer goods company and a school system.

Half their activities lasted less than nine minutes. Only 10% exceeded an hour. Their plans seemed to live "only in their heads," as flexible intentions rather than the reflective, systematic planning Mintzberg called folklore.

Peter Drucker found a similar gap in The Effective Executive. He asked executives to estimate where their time went, then compared those guesses with their logs. The two bore little resemblance.

Five executives are a small sample, and Mintzberg's account appeared in the 1970s. But Michael Mankins, Chris Brahm and Greg Caimi later examined time budgets at 17 large companies through Bain.

Their 2014 finding was that organizational time "goes largely unmanaged." They urged companies to give time budgets the discipline given to capital.

That is a different gap from Intel's. A fragmented day need not oppose the stated job. The plan can remain intact in someone's head while the calendar fills without reference to it.

The rewards: against

A calendar nobody governs can sit beside a ranking somebody enforces.

In Kurt Eichenwald's 2012 Vanity Fair reporting, a former Microsoft developer said you walked into a team of ten on your first day knowing the review split. Two would get great reviews, seven mediocre reviews and one a terrible review, no matter how good everyone was.

The pressure reached beyond one team. Executives told Eichenwald that strong developers avoided working beside other strong developers. A talented colleague could hurt your place in the ranking.

One engineer described "withholding just enough information" from colleagues to stay ahead. Helping someone else carried a possible cost to your own review.

Eichenwald reported that every current and former employee he interviewed named stack ranking Microsoft's most destructive process. It drove out "untold numbers of employees."

On November 12, 2013, human resources chief Lisa Brummel told staff Microsoft would end the predetermined distribution. She described the replacement as "designed to promote new levels of teamwork and agility."

The old ranking rewarded behavior that worked against that aim. Microsoft changed the rule, giving managers more freedom to allocate rewards within their compensation budgets.

At greater cost, Wells Fargo showed the same shape when a product count stood in for a customer relationship, including products customers had never requested.

What has already been decided

You could read Intel differently. Nobody at the top chose each wafer allocation. A rule governed those choices. Perhaps the record shows only what happened, not what anyone intended.

A record cannot tell me what someone privately meant. It can show which choices kept receiving resources, whether a person or a standing rule made them.

A rule decides too.

Those decisions pile up in the budget, the calendar and the rewards. They can run ahead, apart, or against what the organization says. None of those relations, by itself, proves anyone was lying.

The strategy those records show is what I call shadow strategy. Reading it doesn't tell you automatically which side to change.

Intel changed its account of the business to fit the direction its allocations had taken. Microsoft changed the ranking to support the teamwork it wanted.

Put your deck beside your last budget, last month's calendar and the last list of people rewarded. Which relation do you see? Which side needs to change?

Ask Grove's question of those records: what would a new CEO do? Something may already be doing it.

The free weekly letter brings one idea to check against the record.