Restructures, however, are inherently costly and disruptive. As the intended outcome of a restructure can be described but not guaranteed, it is basically an experiment. Too often, if you listen to post-restructure conversations among lower level (and sometimes senior) staff, the common theme will be something like: “all that happened was that they just ‘moved the deck chairs on the ‘Titanic’”.

Boards must really be on the ball when their chief executives start talking ‘restructure’, so this piece by Roger Martin on strategy and [re]organisation caught my eye.[i] In this quick read, Martin explains why shifting the walls that define an organisation’s structure does not address the genuine issues that might justify a reorganisation.

Use structure to improve focus

The reason for imposing structure on an organisation is to improve focus. It acknowledges the need to have some people focus on, for example, designing a product; other people on producing it; others on selling it; and still others on keeping track of finances, etc. Or the focus might be on products: some people focus on product A, some on product B, and some on product C, etc.

The ability to focus on just one domain of organisational activity elevates the skill levels applied to it. Higher skill means higher productivity in the task in question. Focus is also important to concentrate resources on priority tasks/goals.

Now add walls

As well as increasing focus, imposing structure on an organisation also sets up walls or boundaries between, for example, R&D, manufacturing, marketing, sales, finance. Heads are appointed to each function who will inevitably look to optimise for their own requirements. Martin observes that the nature of these structure-related walls may vary, but their existence does not.

Evolving ideas about ideal organisational structure

Martin traces the evolution of thinking about the best kind of organisational structure. Through the 1940s, focus was achieved through functional specialisation. During the 1950s and ’60s, organisations began to organise predominantly by product line. Top executives shifted from being function heads to product heads. As well as walls between functions, companies now had walls between products and functions, and between each product and every other product.

Martin wryly notes that “this proliferation of walls was not without consequences”. To manage the complex dynamics of running a product line structure plus their functions at the same time, the next popular structural solution was the matrix organisation, which drew attention to the importance of the intersection between the two axes of product and function. For a matrix structure to work, a functional manager and a product line manager need to make decisions cooperatively at each node.

But, surprise, surprise—that kind of cooperative decision-making proved much harder to master than anticipated. It became even more complicated in the 1980s and ’90s as the largest companies globalised and added a third structural dimension: geography.

Martin pours scorn on one response to the resultant complexity, which was to get rid of the matrix structure and replace it with ‘single point’ accountability: “The minute a company decides to sell multiple product lines in multiple geographies it has a matrix whether it admits it or not.”

Structure creates focus by erecting walls

Following the history lesson, Martin sets out a better way to think about reorganisation. First, he acknowledges that it is conceivable that a company should reorganise to better align with its strategy. However, it should never forget that what structure does is create focus by erecting walls.

Both are important—as he demonstrates with a hypothetical example. Say the company strategy is to compete based on its global scale in R&D, which reports to four regional general managers, resulting in uncoordinated R&D across the four regions. You decide to eliminate the offending walls by having all the R&D people reporting to a global head of R&D. But the problem, says Martin, is that you don’t eliminate walls by reorganising, you simply change their location—in this case from between regional R&D and global R&D, to between regional general management and regional R&D. Martin suggests there needs to be a variation on Newton’s conservation of energy—the conservation of walls!

Think about the impact of the new wall

The consistent mistake made in reorganising is to focus on eliminating the problematic wall and not think at all about the new wall that will automatically replace it. The key, therefore, is to figure out how to deal productively with the inevitable new walls.

The problem, Martin explains, is the difficulty of having productive conversations across walls. There are too many incentives to throw things over walls—to, in effect, shift responsibility (“I’ve done what I am supposed to do on my side of the wall. Now I’m throwing it over to your side of the wall for you to do what you are supposed to do.”)

In Martin’s own words: “the key is for the managers at intersection points to share some jointness of responsibility to talk productively with one another in order to come up with the best joint decision for the company – not the best narrow decision for their side of the wall”. He calls this Integrative Thinking, which he has described at length elsewhere.

So, is reorganisation really needed?

Martin contends that reorganisation enthusiasts concentrate on the problematic walls they want to remove and ignore the potentially harmful places to which those walls shift. He says you can shift a wall to increase focus in one area, but the cost is dividing focus in another. As much time should be spent understanding where the wall is going and the impact that will have, as appreciating how its removal from a given position will benefit the strategy.

So much of effective governance is ensuring that things occur in the right sequence. A board should be wary of talk about restructuring until it is satisfied that top management has made tangible efforts to improve the quality of conversations at the organisational intersections critical to the strategy. Get those in the right order, and reorganisation may not be needed.

[i] Roger Martin. ‘Strategy and [Re]Organisation: Walls Don’t Disappear, They Just Move.’ Medium, 9 November 2021