S&OP best practices: the four things that make it work, and the silent killer nobody plans for

About Eric Wilson

More than 25 years in S&OP planner, manager and director roles at manufacturers, then the educational side at the Institute of Business Forecasting and Planning (IBF), where he runs training, assessments and events. Author, speaker and consultant; his latest book is Beyond Borders, on S&OP across cultures. Listen to the episode.

Four factors, and the one he missed

Eric Wilson has written the maturity models, run the trainings and sat in more S&OP executive forums than almost anyone. Ask him what separates the processes that work from the ones that don't and he gives four factors. Then he adds the one he missed for years.

"When you don't address culture, it can be a silent killer for a lot of S&OP processes. Companies go in with the best of intentions, technology or processes, and even good people, and find out they're approaching it from different angles, different cultural upbringings that manifest in a business setting unknowingly."

Here are the four factors, the killer, and what to do about each.

1. Structure and cadence

"The more cadence and structure you put to it, the better I see the results are going to be," Eric says. "The first Thursday of every period is my demand review. The second Tuesday is the supply review. We know what you cover. We have defined roles and responsibilities."

Structure is not rigidity. The cadence stays fixed; what happens inside it adapts. But without a fixed rhythm, S&OP becomes a meeting that happens when someone remembers, and decisions wait for it.

2. Collaboration and one plan

S&OP is a decision-making forum: "It's going to align all functional areas on a unified set of assumptions for coordinated decision-making. That is what S&OP is about." The more functions in the room, the more insight; and the single plan everyone works to is what keeps them moving together.

Eric's research found something unexpected about that. "We found that is correlated to agility. When you have consensus, even if you're moving in the wrong direction, as you add in new insights, you can steer the ship together towards another direction." One plan is faster to change than five.

3. Transparency, including about what you don't know

"No one has all the answers. The companies I would talk to that said 'we really don't know' actually sometimes did better, because they were being honest internally."

This is also his answer to the most common S&OP derailer: data. "Two thirds of our time is still spent wrangling data. With that, you're going to have the 'I don't trust the data' mentality." His advice is not to wait for perfect data, because nobody has it, but to show where it comes from, how it was processed, and get agreement that this is the number the forum works from.

4. Trust, the golden thread

"You have to build trust. Trust in the process. Trust in the people. You're not going to have collaboration without trust. That's the golden thread between the other three."

And it explains the data problem better than the data does. "You're going to hear 'I didn't trust the data'. Underneath it is 'I didn't trust the person presenting the data', or 'I didn't trust the process', or 'I didn't trust the sources'. You have to find the root cause."

Eric's order of priorities follows from this: "People, process, data, analytics, technology. I would put them in that order."

The silent killer: the meeting was never designed for a pause

The story Eric tells is from an executive forum a couple of years ago. Cross-functional, cross-regional, CEO in the room, a good process. A marketing lead from Southeast Asia, "brilliant, had a lot of great insights", said nothing for the whole meeting while dominant voices from other regions talked over each other.

"Talking with her afterwards, I asked her why. Her response was: I was waiting for a pause. There was never going to be a pause in that meeting unless they designed that meeting intentionally for a pause, or asked her. They lost out on some valuable marketing insights."

That is culture in S&OP: how people make decisions, whether they challenge, whether a decision is final or pliable, whether they share mistakes. "There's no right or wrong answer here. They're different approaches. We need to blend those approaches together into a global S&OP process." The fix is design: build the pause in, direct the question, and understand that trust is built differently in different places.

Where AI helps S&OP today, and where it doesn't yet

Eric moderated an AI panel where a product manager from a large software company put the hype at 60-70%. What he hears from practitioners is simpler: "What is this, and what do we do about it? Boardrooms are saying 'go do something with AI', and it trickles down to the S&OP managers and demand planners."

The uses that work now are about communication and routine: turning data into something sales and finance understand, stress-testing scenarios, and the monthly grind of collecting data, building presentations and compiling action items. "Where I'm seeing less use cases right now is on the demand and analytics side. It's too dynamic still."

His best example is an executive forum where a question got an answer in the room instead of a day later. The executive took it, and told Eric why: "It spoke my language. And it had my attention. Typically all these smart people I pay a lot of money to would go out of the room and come back a day later. By that time, I'm three or four problems down the road." Eric's caveat: this is about cutting the latency of decisions, not real-time S&OP. "Nor would I want to turn S&OP into a firefighting exercise."

Your next steps

• Fix the calendar: demand review, supply review, pre-S&OP and executive meeting on the same days every cycle, with named owners.
• Agree one plan and one set of assumptions, and change it together when the facts change.
• Publish where the numbers come from and how they were built, and get the forum to sign off on working from them. Then stop debating the data.
• When someone says "I don't trust the data", ask what they actually don't trust: the source, the process or the person.
• Design the pause. Before the meeting, know who in the room will wait to be asked, and ask them.

Eric's order of priorities is the summary: "People, process, data, analytics, technology. I would put them in that order." Most S&OP fixes start at the wrong end of that list. For the process itself, see our explainer on integrated business planning.