Better Decisions Come From Better Conversations, Not Bigger Spreadsheets

The owner of a growing printing business is deciding whether to buy a large new press. It’s the biggest investment the company has ever made. Her operations manager builds a detailed spreadsheet: forty tabs of costs, finance options, projected volumes and payback scenarios. It shows the press paying for itself in under three years. The owner, who has wanted the machine for a while, signs the deal.

Eighteen months later, the press is running at half capacity. Demand for the large print runs it was built for has fallen, as several customers had moved to digital. A couple of staff had suspected this might happen, but nobody had asked them. The spreadsheet was thorough. It just modelled the future the owner hoped for.

Most of us assume that better decisions come from better analysis: more data, more detail, more scenarios. Research on more than a thousand business decisions suggests something different. How a decision is discussed and challenged matters even more than how much analysis goes into it.

One Like insight: Analysis tells you what the numbers say. A good process tells you whether anyone has questioned them.

Process beats analysis

In 2010, Dan Lovallo of the University of Sydney and Olivier Sibony of McKinsey & Company published an article in McKinsey Quarterly called “The case for behavioral strategy”. They had studied 1,048 major business decisions made over the previous five years, including new-product investments, acquisitions and large capital spending.

For each decision, managers reported how much analysis had gone into it, such as detailed financial models and sensitivity analyses, and how the decision had been made. The process questions included whether major uncertainties had been explicitly discussed and whether viewpoints contradicting the senior leader’s had been raised. The researchers then compared these practices with the outcomes of each decision.

Their conclusion was striking: process mattered more than analysis, by a factor of six. For the decisions with return-on-investment data, moving from the bottom to the top quarter on decision process was associated with an ROI improvement of 6.9 percentage points.

The authors were careful to say this didn’t make analysis unimportant. Almost none of the decisions made through a very strong process were backed by very poor analysis, because a good process tends to expose weak analysis. The reverse wasn’t true: excellent analysis was of little use if the process never gave it, or the doubts around it, a fair hearing.

The data relied on managers’ own reports and came from large companies, and Lovallo was an adviser to McKinsey. But the finding fits a large body of research showing how biases such as overconfidence and the urge to confirm what we already believe shape even careful decisions.

What a good process looks like

Sibony has compared the typical business decision to a courtroom where only the prosecution speaks: one team presents a single recommendation, supported by slides, and the boss decides. Nobody is asked to argue the other side. A better process doesn’t need to be bureaucratic. For a small business, it can come down to a few habits used for big decisions.

Consider more than one option. Before deciding whether to buy the new press, ask what else could solve the same problem: outsourcing large runs, leasing, upgrading the existing machine, or doing nothing for a year.

Name the uncertainties. List the two or three assumptions the decision depends on most, such as future demand, and ask how confident you really are about each.

Invite disagreement deliberately. Ask someone to make the strongest case against the decision, or run the premortem we explored in the previous article. Make it clear that challenging the plan is part of the job, not disloyalty.

Hear from people closest to the work. Staff who deal with customers every day often notice changes long before they show up in the numbers.

Agree what would change your mind. Before you decide, write down what evidence would make you choose differently, and look for it.

Making it routine

You don’t need a formal decision process for choosing a new supplier of printer paper. Save it for decisions that are large, hard to reverse or likely to shape the business for years. For those, a simple one-page decision note can help: the options considered, the key uncertainties, the strongest argument against, what people closest to the work think, and what would change your mind. Writing it takes an hour. Skipping it can cost far more.

It also helps to look back. After a big decision has played out, spend twenty minutes asking what you got right, what you missed and what you’d do differently in the process, not just the outcome. Over time, that builds a habit of deciding well.

A few questions can help:

• For our last big decision, how many genuine alternatives did we consider?
• Did anyone argue the opposite case, and was it taken seriously?
• Which assumptions did the decision depend on most, and how confident were we?
• Who closest to the work did we ask, and who did we forget?

Which brings us back to the printing business and its forty-tab spreadsheet. The analysis was careful. What the decision lacked was a moment where someone asked whether demand for large print runs might fall, and a chance for the staff who suspected it to say so. One honest conversation might have led to leasing a smaller machine, or waiting a year, and saved the business a very expensive lesson.

Sources

1. Dan Lovallo and Olivier Sibony, “The case for behavioral strategy”, McKinsey Quarterly, March 2010.
2. Farnam Street, “What Matters More in Decisions: Analysis or Process?”, for Sibony’s courtroom analogy.