SAN ANTONIO–Finding the best answer is not enough — it must also be sold to the client. After all, “the best answer will not make everyone in the room happy.”
That was the message delivered by Jerry Allyne in the keynote address that opened the INFORMS Conference on Business Analytics and Operations Research April 8 in San Antonio. Allyne, vice president for strategic planning and analysis at Boeing Commercial Airplanes, told the assembled analytics professionals that the acceptance of a solution is as important a factor as the analytical quality of that solution.
In his role, Allyne said he handles market forecasts, assessing what products will fit into the market in the future, the options for meeting that demand and evaluating those options. Using analytics, Allyne proposed an equation for evaluating the effectiveness of a given solution to a problem: the effectiveness (E) of a solution can be expressed as the quality (Q) of the solution multiplied by its acceptance (A), or Q x A = E, with both Q and A expressed on a scale of 1 to 10. The formula originated with General Electric.
“Many analysts work so hard on the quality of an answer that they get Q to a seven, but they don’t work on who would be using it, and only get a two for the A,” he said. “The result is a 14, which is not great. So they go back and do a greater job on quality and get it to a nine, but A is still two, so they only get an 18, which is still ineffective.
Taking the audience into account, Allyne said, changes the equation. If analysts “looked at the stakeholders and how those people would use the answer, they could initially get A to five and have 35 initially, rather than 14. It’s a helpful way to think about things,” he said.
Good answers do not sell themselves, he indicated, pointing out that there is plenty of data showing that smoking will shorten your life, but people continue to smoke.
“Something on the A side is broken,” he said.
Getting a decision accepted involves both art and science, he said. “When you have a decision to be made, ask who will make it, who has the authority, who can influence that decision, how will it be made, does it require a consensus, does anyone have veto rights? Understand the decision-makers, their motivations, their fears, and what they are trying to achieve. Do a lot of listening. You may have to meet with them individually, one-on-one.
“Some senior leaders will not make a big decision without talking to Ed, so find Ed, who might be five layers removed,” he added.
“A lot of people don’t like being challenged in a group, but you can handle that by settling things in advance — when Congress brings something to a vote the result is often already known since they worked it out ahead of time,” he explained.
When Decision Criteria Change
Allyne used his experiences in the design of the 787 Dreamliner and the 737 MAX to show the need for both art and science.
The configuration for the 787 was initially settled in March 2001, using a delta-wing canard layout (with the wing in back and the stabilizer in front). A conventional configuration would have been more fuel-efficient but the canard configuration was more “sexy,” he noted.
Then came the events of Sept. 11 and the price of fuel eventually rose by a factor of three, putting a premium on fuel efficiency, he recalled.
“It is hard to move away from an emotional attachment to something that is cool or feels good, but we could show that the [conventional] Dreamliner was a better choice,” he said.
With the 737 MAX, the question was whether to build an entirely new airplane or enhance the 737, which came out in 1965, and 10,500 have been sold, he said.
“With the 737, a lot of people were operating them, and commonality [with a new model] meant they could use the same pilots, crews, arrangements for spares, and schedules. An all-new airplane meant you had to assume it was going to represent disruptive technology, so it had to offer enough improvements to override the benefits of commonality,” he said.
“With so many stakeholders and lots of data we had to use a lot of art, and a lot of A,” for acceptance, he said. “We went with the 737 MAX, since we can get it to market faster than with a new plane.”
He explained that Boeing has been tracking the air travel market in terms of RPK (revenue passenger-kilometers) for decades, and seeing an upward trend that is remarkably smooth, increasing by factor of seven from 1991 to 2031. Over the next 20 years the global economy will need another 34,000 airplanes, representing a $4.5 trillion market, including associated services, Allyne said.
Many factors go into such predictions, including economic growth, airport construction trends, environmental sensitivity, regional political risks, and the growth of competition like high-speed rail, he acknowledged.
“But our macro-level predictions tend to hold fairly well,” he said. “The world tends to want to grow. RPK growth has been unbelievably steady. We tend to be off by only single-digit percentages,” he said.
Responding to a question, he said it is the flying public that decides the amount of space between airline seats by consistently choosing to buy tickets on the cheapest flights, so the airlines must put as many seats as possible on the planes in order to make money.
Lamont Wood is a freelance writer based in San Antonio. Reach him via email at email@example.com.