How to Make Your Matrix Organization Really Work

Shifting the reporting structure to a matrix model has its challenges, but leaders can find success by focusing on four key elements.

Reading Time: 11 min 


Permissions and PDF

Much has been written about why matrix organizations are needed and what they look like at a surface level. Far less advice is available about what it takes to make them work. This information gap sets many teams up for disappointment because matrix organizations flourish or fail based on attention to their design and dynamics.

Consider the case of Juan, a regional supply chain leader in a large health care system who was caught between competing agendas from multiple bosses in his organization. Juan reports to Brenda, his enterprise-level boss in the supply chain organization, but he also has a reporting relationship to Steve, a regional operations executive. (Note: All names have been changed for anonymity.)

Brenda’s goals for Juan included implementing a new supplier network model with ambitious timelines. Meeting her goals would require a substantial time investment for Juan and his small team. Meanwhile, Steve was grappling with critical materials and staffing shortages and had asked Juan to optimize workforce and supply costs. Steve expected Juan to meet a tight schedule for opening a new clinical facility to help reach regional volume targets.

Brenda wanted Juan to avoid changes to the workforce or supply commitments until the new network model was approved by senior leadership. However, this would negatively impact Steve’s regional budget and goals.

This situation left Juan frustrated and uncertain about how to prioritize the trade-offs and frame solutions mutually beneficial to Brenda and Steve. Additionally, this made him worry about how his leaders would perceive his performance and ability to deliver on goals.

The challenges Juan, Brenda, and Steve faced are typical of those faced by leaders moving to a matrix organization at Atrium Health. Atrium Health is a nonprofit health care system based in Charlotte, North Carolina, with more than 75,000 employees and over $12 billion in annual revenues. Founded in 1940 as a single charitable hospital serving the local community in Charlotte, the system had grown mainly in North Carolina until 2018. At that time, Atrium Health’s leadership, recognizing that greater resources and scale were necessary to meet competitive challenges, initiated a series of mergers with other health care systems to create a larger system with operations in North and South Carolina, Georgia, and Alabama.

Atrium Health’s leadership understood early on that the organization’s legacy functional structure could not cope with the increased scale and geographical dispersion.


Reprint #:


More Like This

Add a comment

You must to post a comment.

First time here? Sign up for a free account: Comment on articles and get access to many more articles.

Comments (3)
Alexander Goettling
In theory, that all sounds convincing. But in real life, a matrix organization lets it up to petty “Juan” to negotiate with the “Brendas” and “Steves”, just as the board and executive leadership avoid a clear prioritization. If “Juan” is lucky, “Brenda” and “Steve” are sensible and act in the best interest of Atrium Health. But real-life experience tells us that this is not always the case - or, to be realistic, it is rarely the case. Even Brenda and Steve are in a challenging spot as their performance is measured by getting things done - hence there is not much reward to being overly collaborative. The Key is clear strategic priorities from the top, which requires alignment of shareholders, boards, and executive leadership teams. If you have that, you don’t need a matrix organization, which usually creates more harm than good. Academically, it's a good idea. Practically it doesn't work.
Procyon Mukherjee
The tensions are unavoidable in a matrix structure and no matter how well we design the processes that lead us to a better result, we must acknowledge that leaders are agents with motives; this leads us back to the agency levers that deliver results, the most powerful of them is still hinging on finding the right dynamic balance that will optimise several objective functions. But as the puzzles of the future could be very different from the current, allocating resources for the present and the future is never an easy task or doing exceedingly well on the goals set out for the short period need not be sufficient for winning in the longer term. 

This article is more focused on the easier puzzles of local and global optima across increased nodal tensions, the moment you introduce the time dimension and opacity into the future factors and limiting constraints, the leaders are faced with the cognitive dissonance hazard; how much into the future should the current actions be discounted for the externalities, constraints, et al. Becomes the dominant theme then. 

The solutions that emerge seem to point to a collective entropy, an emerging postponement  of the future problems; the collective consensus acts among the competition, till the future leaders emerge to dispel and they demonstrate a new paradigm. For all you know they could be the ones who are being hammered by the markets for their current behavior of bucking the trend lines on the reverse. 

Nodal structures, platform design, non-hierarchical communication and feedback are good building blocks for organizational therapy; the outliers are those that are simply built on trust engines of purpose and value based work, where means becomes as important as objectives.
Alan Gageler
Missing or underemphasised for me in this model is ensuring that the Where, What and How are driven by a clear and practical understanding of how decisions are made in the organisation, ranging from strategic through to dial operations. In my experience advising organisations around strategy execution & organisation design, too many organisations are following fashion or leadership whims in how they structure and don't have a clear conception of the how the operating model needs to be designed to support the strategic & operational execution. The failings of most matrix structures are not only the contest of priorities & decision rights they often give rise to but just as critically the efficiency and timeliness with which decisions are made.  Network analysis tells us that the more nodes in a network, the higher the transactional load.  In a world which has drifted ever further towards narrower and narrower scopes for organisational units through the proliferation of specialist functions, this sees what should be simple, in-the-moment decisions turned into tortuous endeavours for those involved. In my design approach, we place decision-making efficiency as just as important as efficacy.