Category: Insight

Part five of five in The Modern Technology Organization, a series from Paragon by Riviera Partners, the talent partner that helps top and growing technology companies scale, structure, and transform their teams. This series draws on our ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models.

What we learned that works

• Deploy your strongest leaders where complexity is highest—the cross-team seams with the greatest cost of misalignment.
• In autonomous orgs alignment is created, not enforced; it’s ongoing, invisible work that standard reviews under-reward.
• Distributed teams need explicit operating discipline: defined decision rights, information flow, and rhythms that don’t rely on proximity.
• Start where the strain is greatest—diagnose hiring throughput, decision-making, team interplay, and leadership deployment.

The first four pieces in this series described one problem from several angles. Modern tech companies flattened, distributed, and empowered themselves into a state where they feel fast and struggle to turn that speed into aligned results, which is the execution gap. Flattening redistributed complexity onto fewer leaders and into the seams between autonomous teams. Meeting that demand takes hiring for judgment and adaptability and treating hiring as a scaling lever rather than a support function.

This piece is about what you build once you accept all of that, because getting the diagnosis right only helps if you know what to put in its place. Modern org design raises the bar on how you lead, staff, and operate a company, and a leadership team has to get four things right for these models to work. The first is hiring, covered in the two previous pieces. The other three are the subject here.

Put your best leaders where it hurts most

Flattening cuts the number of leadership layers while raising the need for leadership. That sounds like a contradiction until you look at where the work went. It did not disappear with the layers. It spread across a structure with fewer people left to do it. So the answer is not to add layers back. It is to get far more deliberate about where you place the leaders you have.

The instinct most companies follow is to stack their strongest leaders at the top, or to spread them evenly so every team gets a share. Both are wrong for a flat, networked organization. The right move is to deploy your best people where the complexity is highest: the areas with the most cross-team dependencies and the steepest cost of misalignment. That is where a strong leader prevents the most damage and creates the most leverage, and where a weak one does the most harm.

Leadership density means leadership exists throughout the company rather than clustered at the top. Without it, a flat structure turns fragile, with too much responsibility resting on too few people, one departure or one overloaded quarter away from coming apart.

Hiring and evaluating for density looks different from traditional leadership assessment. Team size and scope of ownership, the old proxies, predict little here. What predicts success is a leader’s track record of moving decisions across functions, managing competing priorities, and holding alignment in a shifting environment. You may also need to rethink how you measure and reward these leaders, because their impact happens across teams rather than inside one domain. Judge a leader on their team’s output and on their ability to drive alignment, unblock others, and contribute to outcomes beyond their own patch. If your incentives only reward local results, you will get local optimization, and that is the exact failure a networked organization cannot afford.

Alignment is a job, not a byproduct

In a flat, autonomous organization, alignment becomes ongoing work that leaders perform rather than a state the structure guarantees. In a hierarchy the org chart itself decides who defers to whom and whose priorities win. A network-of-teams organization gives that up. Once teams are autonomous, no one can impose alignment from above. Leaders have to create and maintain it continuously from within.

That turns alignment from a state the structure guarantees into an activity leaders perform. Effective leaders in these environments keep reinforcing priorities, clarifying tradeoffs, and making sure a decision in one corner of the company does not quietly create a problem in another. They connect roadmaps across teams, surface conflicts before they escalate, and help others move faster because the shared picture is genuinely shared. This is ongoing work, not a quarterly planning ritual, and it is most of what keeps autonomous teams pointed the same way.

It also changes who you promote into these roles. Favor people comfortable operating beyond their own scope, leaders who can move decisions without formal authority over them, because in a network that is the only kind of influence on offer. The best of them understand not just what their team owns but how that work fits the whole system, and they adjust accordingly.

One thing makes this hard to manage: in autonomous environments, effective alignment work is nearly invisible. When a leader defuses a conflict before it becomes a fire, nothing appears to happen. That invisibility keeps it out of standard performance reviews, so you have to go looking for it. When you evaluate leaders, look past their team’s execution and ask how well they reduce friction across teams, resolve tradeoffs early, and contribute to shared outcomes. Miss this and you will systematically underpay the exact behavior your organization most depends on.

Distributed work needs an operating system

Distributed teams stay coordinated only through explicit operating discipline: defined decision rights, clear information flow, and rhythms that do not depend on being in the same room. Hybrid and remote work make this the operating system underneath everything else. In an office, a lot of coordination happens by accident: decisions made in hallway conversations, ambiguities cleared up in a quick face-to-face. Spread the company across locations and time zones and that informal layer evaporates. Everything turns into an email, a message, an async thread.

When that is the medium, the failures are predictable. Decisions stall while teams wait on input across time zones. Communication fragments across tools and channels. Teams work from different context and assumptions, so alignment erodes without anyone choosing to let it. Culture gets harder to sustain without the proximity that used to reinforce it.

The answer is deliberate operating discipline. A distributed company has to define, out loud, how decisions get made, how information moves, and how teams stay aligned, all the things a co-located company could leave unspoken. Leaders then have to reinforce those behaviors consistently, so teams know how to operate without someone in the next room. As organizations spread out, they lean harder on explicit systems for decision-making, information flow, and alignment. The most effective leaders in these settings make expectations explicit, communicate decisions in a form other people can act on, and set rhythms that keep teams aligned without relying on being in the same place.

Where to start

Shifting toward a flatter, more autonomous, more distributed organization is not something you do all at once, and you are almost certainly operating this way to some degree already. The most effective approach is to find where your organization is under the most strain and start there. Four diagnostics point the way.

Look at your hiring throughput and quality. How fast can you find and onboard high-judgment people into critical roles, and where are hiring gaps choking execution right now?

Look at how decisions actually get made. Where do they slow down, get escalated when they should not, or get reopened again and again? Those are the places alignment is failing.

Look at how teams operate relative to one another. Where do dependencies create friction, and where are teams moving fast while producing inconsistent outcomes? That inconsistency is the fingerprint of autonomy without shared context.

Look at how your leadership is deployed. Are your most experienced leaders sitting in the areas of highest complexity and impact, or somewhere more comfortable and less consequential?

The answers tell you where to begin. You do not have to fix everything at once, and trying to will spread you too thin to fix anything. Start where the strain is worst.

The through-line

Everything in this series comes back to one idea. As a company moves away from centralized control toward distributed decision-making, performance stops being a product of structure and becomes a product of the quality of leadership and talent across the whole system. The org chart cannot save you. The people making decisions inside it, and the systems that keep those decisions aligned, are what determine whether ambition turns into execution.

The organizations that get it right do four things well. They put leadership density where it matters most. They treat alignment as active work rather than a structural given. They treat hiring as a core scaling lever. And they build the operating discipline that lets distributed teams execute with consistency. None of it is free, and none of it is automatic. Together, it is what turns a fast-feeling company into one that is actually going somewhere.

Frequently asked questions

What is leadership density?

Leadership density means capable leadership exists throughout the organization rather than concentrated at the top. In flat structures with fewer layers, it is the deliberate placement of strong leaders where complexity is highest, so that responsibility isn’t dangerously concentrated in too few people.

How do distributed teams stay aligned without hierarchy?

Alignment becomes an active, ongoing leadership responsibility rather than something the structure enforces. Leaders continuously reinforce priorities, clarify tradeoffs, and connect roadmaps across teams to resolve conflicts before they escalate. Because this work is largely invisible, organizations have to evaluate and reward it deliberately.

Where should an organization start modernizing its structure?

Start where the strain is greatest. Diagnose four areas: hiring throughput and quality, how decisions actually get made, how teams affect one another, and where your strongest leaders are deployed. Address the highest-strain area first rather than attempting to change everything at once.

Previous: Part 4

About this series: This article and the accompanying ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models, were developed by Paragon by Riviera Partners , a leader in helping top and growing technology companies hire, scale, and structure their teams. Drawing on our work with hundreds of technology organizations, Paragon partners with CEOs and HR leaders to identify, evaluate, and place the leaders and talent who can scale decision-making, maintain alignment across teams, and turn ambition into execution. Download the full ebook or learn more about Paragon.

Part four of five in The Modern Technology Organization, a series from Paragon by Riviera Partners, the talent partner that helps top and growing technology companies scale, structure, and transform their teams. This series draws on our ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models.

What we learned that works

• In an empowered org a new hire adds a decision-maker, not just capacity—so every hire carries leadership weight.
• Flat orgs need to hire fast and well at once; that combination overwhelms recruiting teams built for a steadier cadence.
• Leading companies treat hiring as strategy: they define high-judgment talent precisely and augment internal teams with external partners in critical areas.
• The question shifts from ‘can they do the job?’ to ‘can they be trusted to make the right call when no one is watching?’

For a long time, hiring stayed in its lane. A team earned a headcount, HR filled it, the team got bigger, and everyone went back to their actual jobs. Recruiting was a service function, respected in a distant sort of way and firmly downstream of the decisions that mattered. For a good while, the model held.

Then the ground moved. As technology companies flattened their structures, pushed decisions to the edges, and asked people to operate with more autonomy than ever, the stakes on every hire climbed fast. The shift outran most companies’ mental model of hiring, which is why so many still run a support-function playbook in a world that has quietly turned talent into the main constraint on growth.

In an empowered company, a new hire does not just add capacity. It adds a decision-maker. And that changes what hiring is for.

Every hire is a leadership hire now

In a flat, autonomous organization every hire adds a decision-maker, which makes it a leadership hire regardless of the title on the offer. Look at what these companies genuinely ask of their people. No manager signs off on every call. No hierarchy waits to catch things the moment they get murky. The company expects people to set their own priorities, keep them aligned with company goals, and make judgment calls on incomplete information, usually under time pressure.

Read that back and notice what it describes. It reads like a leadership profile, and it now applies to nearly everyone, because in a flat structure every hire carries leadership weight that used to belong only to people with “manager” in their title.

Most hiring processes have not caught up. They still optimize for years of experience, technical skills, and a resume that checks the right boxes. Those things have value. What they do not tell you is whether someone can operate on their own in a fast-moving environment where a single bad decision ripples across a whole team. That question, the one that actually predicts success, is precisely the one traditional hiring is worst at answering, which is the subject of hiring for judgment over experience.

Once every seat is a decision-making seat, hiring stops being a matter of filling capacity. It becomes the mechanism by which the company grows, or fails to grow, its ability to make good decisions at scale. That is not a support function. That is a growth lever.

The speed problem nobody plans for

Here is where it gets genuinely hard. Flat organizations are built to move fast, and fast organizations have to hire fast to keep pace. Hiring fast without hiring well is one of the most expensive mistakes a growing tech company can make, expensive because a weak hire in an empowered role does not just underperform quietly. It makes autonomous decisions that others build on before anyone notices the foundation is soft.

Once hiring becomes the primary scaling lever, the pressure on internal recruiting teams becomes enormous, and it comes from two directions at once. Filling roles is no longer the job. The job is finding and closing high-judgment people quickly enough to keep pace with a business that will not slow down to wait. Volume, quality, and speed, all at the same time, all pulling against each other, and all inside a market where the talent is scarce. ManpowerGroup’s 2025 survey found that roughly three in four employers worldwide struggle to fill roles, with AI skills now the single hardest capability to find.

Most internal recruiting teams were not built for that combination. They run on a steadier cadence of well-defined roles, not a continuous sprint to win scarce, high-judgment people against every other company chasing the same profile. When the pipeline slows, execution slows with it, and because hiring sits upstream of everything, the slowdown shows up everywhere at once. It is one of the clearest mechanical drivers of the execution gap, the widening distance between how fast a company wants to move and how fast it actually can.

What treating hiring as strategy looks like

The companies getting this right stopped treating hiring as a reactive function that springs to life when a req opens, and started treating it as a capability they invest in continuously. In practice that shows up in three places.

They define high-judgment talent for their own context, not in the abstract and not just on paper, but in terms of how a person has to operate inside their actual company. They know the best hires are not reliably the ones with the most impressive backgrounds. They are the ones who make sound calls under pressure, understand the downstream impact of their choices, and change course when conditions change. Naming that profile precisely is what makes it possible to hire for it on purpose rather than recognizing it by luck.

They are honest about capacity. This is the part most companies avoid, because it feels like an admission of weakness. Many leading technology organizations have started augmenting their internal recruiting teams with external partners in high-priority areas, and the reason is capacity, not competence. The volume and speed required to compete for top talent in critical roles exceed what any fixed internal team can sustain on its own. Treating that as a strategic choice rather than a last resort is a marker of the companies that take hiring seriously.

They connect hiring to the rest of the operating model. Great hiring in isolation does not hold. It has to feed into building leadership density and operating discipline, so you actually deploy the high-judgment people you worked so hard to land where they matter, and support them once they arrive. A brilliant hire dropped into an incoherent system underperforms the same as a mediocre one.

The question that actually matters

Once every hire is a decision-making hire, the question changes. It is no longer only whether this person can do the job. It is whether this person will make the right call when nobody is watching.

That is a harder thing to establish in an interview, and it is the right thing to be asking. Building a process that can answer it, consistently and at the speed the business demands, is what separates the companies that scale cleanly from the ones that scale straight into chaos.

The organizations that figure out how to hire for judgment, reliably and fast, are not only building better teams. They are building a compounding advantage. Every high-judgment person they add raises the ceiling on what the company can take on, and the gap between them and slower competitors widens with every quarter it goes unaddressed.

Hiring was never really just HR’s problem. It just took the rest of the organization a while to notice that talent had become the lever the whole business turns on.

Frequently asked questions

What does it mean to treat hiring as a scaling lever?

It means recognizing that in an empowered organization every hire adds a decision-maker, not just capacity. Hiring becomes the mechanism that grows the organization’s ability to make good decisions at scale, which makes it a strategic growth function rather than a reactive support task.

Why is hiring speed so critical for technology companies?

Flat organizations move fast and need to hire fast to keep pace, but hiring fast without hiring well is expensive—a poor hire in an empowered role makes autonomous decisions others build on. With roughly three in four employers worldwide struggling to fill roles, speed and quality compete constantly.

Should companies use external recruiting partners?

Many leading technology organizations augment internal recruiting with external partners in high-priority areas. The reason is capacity, not competence: the volume and speed required to win scarce, high-judgment talent often exceeds what a fixed internal team can sustain without sacrificing quality or slowing execution.

Previous: Part 3  |  Next: Part 5

About this series: This article and the accompanying ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models, were developed by Paragon by Riviera Partners , a leader in helping top and growing technology companies hire, scale, and structure their teams. Drawing on our work with hundreds of technology organizations, Paragon partners with CEOs and HR leaders to identify, evaluate, and place the leaders and talent who can scale decision-making, maintain alignment across teams, and turn ambition into execution. Download the full ebook or learn more about Paragon.

Part three of five in The Modern Technology Organization, a series from Paragon by Riviera Partners, the talent partner that helps top and growing technology companies scale, structure, and transform their teams. This series draws on our ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models.

What we learned that works

• Experience predicted success when hierarchy absorbed average judgment; flat, autonomous orgs removed that safety net.
• High-judgment people decide well with incomplete information, think in systems, recalibrate fast, and prioritize without being told.
• Traditional hiring optimizes for legible signals—years, titles, skills—and systematically misses judgment.
• Interview for how people navigate ambiguity, weight adaptability as a real criterion, and move fast enough to close them.

For a long time, hiring in technology ran on a familiar script. Count the years. Note the logos. Match the title to the role you were filling. Check the boxes, send the offer, move on. It was efficient, defensible, and easy to explain to everyone in the loop.

That script fit a particular kind of company, one built to absorb average judgment. When there were oversight layers, clear reporting lines, and managers positioned to catch mistakes before they cascaded, experience in a narrow role was a decent proxy for success. You did not need every hire to exercise independent judgment, because the structure supplied it for them.

Most technology companies do not run that way anymore. And the further a company drifts from that old model, the less you can trust its old hiring instincts.

The resume answers a question that matters less now

A resume tells you what someone has done and where they did it, and that mattered far more when a hierarchy stood ready to compensate for average judgment. Inside those structures past performance in a structured setting predicted future performance in a similar one, so the record was a fair proxy.

Modern tech companies have spent the last few years dismantling that setting on purpose. They flattened structures, pushed decisions to the edges, and built cultures of empowerment where people operate with real autonomy. The average manager now oversees nearly six direct reports, up from about three in 2019. Flat, autonomous structures removed the safety net that experience used to lean on.

What is left is a company where every person functions, in a real sense, as a leader. There is no middle-management cushion to catch a bad call, and no obvious hierarchy to escalate to when priorities get murky. People make decisions, often quickly, often with incomplete information, and they live with the consequences directly. In that environment a resume full of years and titles tells you remarkably little about whether someone will thrive.

What high-judgment talent actually looks like

The people who do well in flat, autonomous organizations share a handful of traits that do not render cleanly on a CV. Naming them precisely is the first step to hiring for them.

They decide well under pressure, and, more tellingly, when they do not have the full picture. Anyone can make a good call with complete information. The rarer skill is making one with a partial view, which happens to be the actual condition of work in a fast-moving company.

They think in systems. They know a decision on their team does not stay on their team, that it ripples into three others, and they weigh those downstream effects before they act instead of apologizing for them afterward. This is the trait that separates a strong contributor from someone you can hand real autonomy.

They recalibrate. When the situation shifts, and in a fast-moving tech organization it always shifts, they do not wait to be told. They adjust, say so, and keep going.

They prioritize. In an empowered company nobody hands you a ranked to-do list every morning. The ability to work out what genuinely matters and aim your energy at it is one of the most undervalued skills in tech hiring right now, precisely because it is invisible on paper.

None of this rides on years of experience. Some of the most adaptable people come from backgrounds nobody would have predicted. Some of the most credentialed candidates fall apart the moment the structure around them disappears. Experience and adaptability are simply different variables, and modern org design has made the second one matter more than the first.

Why most hiring processes miss it

If adaptability is what matters, why do so many companies keep hiring for experience? Because their process is built to detect the wrong signal, and the process is hard to change.

Interviews test what someone has done, not how they think. Reference checks confirm past performance in a context that no longer resembles the one the person is walking into. Job descriptions list the skills you can name, because judgment does not fit in a bullet point. The whole apparatus optimizes for legibility, for signals that compare cleanly across candidates, and judgment stubbornly refuses to be legible.

The cost of that mismatch is easy to underrate. As companies hand out more autonomy, they need more people worthy of that trust. When the process cannot spot those people, the company fills seats without building capacity. In a flat organization that gap does not stay hidden for long. One mis-hire in an empowered role makes worse decisions faster, with less oversight to catch them, and the damage compounds. It is one of the most direct feeders of the execution gap between how busy a company looks and how much it accomplishes.

How to hire for judgment instead

Shifting the process is not complicated, but it takes intention and a tolerance for more ambiguity in your own evaluation.

In interviews, push past the record of what a candidate has done and into how they handled ambiguity. Ask about a decision they made with incomplete information and what they would change now. Ask about a time they aligned people with no authority to make them cooperate. Ask what they got wrong, and watch how honestly they answer, because candor about failure is itself a signal of the self-awareness these roles demand.

Weight adaptability as a real criterion, not the tiebreaker you reach for when two resumes look identical. The best person for where your company is going is frequently not the one with the most relevant experience for where it has been. That reframing is uncomfortable, because it asks a hiring manager to trust a read on how a person thinks, and that read is squishier and riskier than a list of credentials. It is also the read the work now requires.

Then move fast enough to actually land them. High-judgment people always have options. The companies that win are not only better at spotting the right person, they are quick enough to close before someone else does. That speed requirement is big enough to deserve its own treatment, which is where the next piece picks up, on hiring at the speed the business demands without giving up quality.

The uncomfortable part

There is a reason this shift is hard, and it is not ignorance. Hiring for experience feels safe. It survives a debrief, it satisfies a skeptical stakeholder, and it rests on credentials you can point to. Hiring for adaptability asks a team to trust its judgment about someone else’s judgment, a squishier bet that is harder to defend when it goes sideways.

It is also the bet the modern organization actually requires, and it pairs directly with how you deploy the leaders you already have. Deploying leaders where complexity is highest only works if those leaders have the judgment to handle complexity once they get there.

The best hire your company makes this year might not resemble any hire it has made before. Different background, different path, a resume that does not slot neatly into the old template. In a company built on autonomy and judgment, that is worth a closer look rather than a quick pass.

Frequently asked questions

Why is adaptability more important than experience in tech hiring?

Flat, autonomous organizations removed the oversight layers that once compensated for average judgment. When every person makes real decisions with incomplete information, the ability to adapt, prioritize, and reason about downstream impact predicts success better than years of experience in a narrowly defined role.

What does high-judgment talent actually look like?

High-judgment people decide well without complete information, think in systems and anticipate downstream effects, recalibrate quickly when conditions change, and prioritize without being handed a ranked list. These traits rarely show up on a resume, which is why traditional screening tends to miss them.

How do you interview for judgment and adaptability?

Push past what candidates have done into how they navigated ambiguity. Ask about decisions made with incomplete information, times they aligned people without authority, and what they would do differently. Candor about failure signals self-awareness, and adaptability should be weighted as a real criterion rather than a tiebreaker.

Previous: Part 2  |  Next: Part 4

About this series: This article and the accompanying ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models, were developed by Paragon by Riviera Partners , a leader in helping top and growing technology companies hire, scale, and structure their teams. Drawing on our work with hundreds of technology organizations, Paragon partners with CEOs and HR leaders to identify, evaluate, and place the leaders and talent who can scale decision-making, maintain alignment across teams, and turn ambition into execution. Download the full ebook or learn more about Paragon.

Part two of five in The Modern Technology Organization, a series from Paragon by Riviera Partners, the talent partner that helps top and growing technology companies scale, structure, and transform their teams. This series draws on our ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models.

What we learned that works

• Flattening relocates complexity rather than removing it; the work of the missing layers lands on fewer, broader-scoped leaders.
• Stretched leaders quietly stop coaching, carrying context, and developing the next generation—weakening the internal pipeline.
• Autonomous teams accelerate work but scatter coordination, producing duplicate work, diverging systems, and unexpected dependencies.
• Flattening is a bet on the capability of the people who remain; win it deliberately by placing leadership where complexity is highest.

Somewhere in the last few years, “flat” turned into a compliment. Cutting management layers signaled a company that trusted its people, moved fast, and refused to drown in bureaucracy. Widening spans of control became a mark of operational maturity. The average manager now oversees nearly six direct reports, roughly double the number in 2019, and most leadership teams read that as a win.

For a lot of what it promises, flattening delivers. Fewer layers mean fewer handoffs. Communication runs more directly. Decisions that once crawled up and down a chain of approvals now happen closer to the work. Teams pick up more ownership, and managers see more of what their people are actually doing.

Then there is the part that never makes the slide. Flattening does not reduce the complexity of running an organization. It moves it. Knowing where that complexity goes, and who ends up carrying it, is what separates a flat company that scales from one that turns brittle without noticing.

The work in those layers did not disappear

Every management layer was doing real work, even when the work was invisible. Middle managers absorbed context and passed it between teams. They settled competing priorities before those fights reached the people building the product. They caught weak decisions early, coached people through the hard ones, and held the connective tissue that kept separate teams working from the same understanding.

Strip the layers out and that work does not leave with the boxes on the chart. It redistributes to whoever is left, which usually means fewer leaders, each now responsible for a wider scope, each carrying more context across more teams, each with less time to do any of it well.

The company gets structurally simpler and operationally more dependent on individual leadership. That is the real trade inside flattening, and it is a genuine trade: you are swapping layers for concentration. When the remaining leaders have the range and judgment to hold it together, the swap pays off. When they do not, you have moved the bottleneck off the org chart and into a handful of overloaded people, which is one of the core dynamics behind the execution gap that shows up across fast-moving tech companies.

What a stretched leader quietly stops doing

A stretched leader stops coaching, carrying context between teams, and setting clear priorities, and the whole thing looks like capable people operating in triage. Once a leader’s span stretches past what they can actually cover, predictable things start to slip.

Decisions turn reactive, because a leader in triage handles whatever is loudest instead of whatever matters most. The bandwidth math is unforgiving: Gallup found that 97% of managers now carry individual-contributor work on top of leading, and they spend a median of 40% of their time on it. Context stops circulating between teams, because the person who used to carry it has run out of hours. Priorities blur, decision rights get fuzzy, and the small ambiguities a present manager would have cleared up in a hallway now sit unresolved for weeks.

The quieter losses are coaching and development. Managers spread across too many reports stop developing the people under them, right when the company needs more capable operators. The scale of the thinning is easy to underestimate. Gusto found that 14% of managerial roles were cut across small and mid-sized businesses, and the share of workers in a people-manager role fell 34% over five years. The people who would have coached the next generation are the ones getting removed. Promotion paths narrow because there are fewer leadership roles to grow into. Over time a structure built to move faster starts producing fewer of the leaders it depends on to keep moving, and the erosion stays invisible until you go looking for your next tier of leaders and find the bench empty.

Autonomous teams solve one problem and open another

Autonomous, cross-functional teams speed up the work inside each team while scattering coordination between teams. Flattening rarely travels alone, and companies that pull out layers usually reorganize what remains into small teams built around products, platforms, or customer outcomes. Engineering, product, design, and data sit together with shared ownership over delivery. The logic holds up: combine the capabilities into one team and you kill handoffs, speed up iteration, and push decisions toward the work.

Inside a single team, it works. The trouble lives in the space between teams. When every team is autonomous, coordination becomes everyone’s job and therefore no one’s, and the result is a familiar set of failures.

Teams optimize for what they can see and control, which is their own local outcome rather than the company’s. Technical decisions drift apart, and systems slide into inconsistency as each team makes reasonable choices in isolation. Duplicate work piles up as cross-team visibility fades. Dependencies surface late and painfully, discovered at integration rather than planned for. And when priorities across teams are unclear, teams stall, unsure whether to move or wait.

This is the paradox of the network-of-teams model. It is built for speed, and it delivers speed, right up to the point where fragmentation eats the gains. As the number of teams grows, so does the need for leaders who can set clear priorities, reinforce shared context, and keep local decisions serving global outcomes. The structure hands you autonomy for free. Coherence you have to pay for.

Complexity migrates to the seams

When you flatten and network an organization, complexity migrates to the seams between autonomous teams and onto the leaders spanning them. In a layered, siloed organization it concentrates in the hierarchy instead: in approvals, handoffs, and reporting lines. Flattening moves it rather than removing it.

That migration changes how you should deploy leadership, and it cuts against most companies’ instincts. The reflex is to spread experienced leaders evenly across teams, or to stack them at the top. In a flat, networked organization the right move is to put your strongest leaders where the complexity is highest: the areas with the most cross-team dependencies and the steepest cost of misalignment. That is the logic behind placing leadership where complexity is highest, and it only becomes visible once you accept that flattening moved the complexity in the first place.

It also changes what you hire for. In a hierarchy you could get away with specialists who executed cleanly inside a narrow lane, because the layers around them supplied context and caught mistakes. Take the layers away and every person needs enough range to operate without them. That profile is harder to find and harder to assess, which is exactly why hiring for judgment becomes central, and why, as these demands compound, hiring throughput becomes the constraint on how fast the whole organization can grow.

Flattening is a bet on your people

The most useful way to think about flattening is as a wager on the capability of the people you keep. Every layer you remove bets that the leaders and individuals left behind can carry the context, judgment, and coordination that layer used to provide. Win the bet and you get a fast, direct, ownership-driven company. Lose it and you get a tidy-looking org chart wrapped around an operation more fragile than the one it replaced.

Companies that flatten well are not the ones that remove the most layers. They are the ones that flatten on purpose, watch where the redistributed complexity lands, and build the leadership density and talent quality to absorb it. Flattening is the easy part. Everything the layers were quietly doing is the part you have to rebuild deliberately.

Frequently asked questions

Does flattening an org structure reduce complexity?

No. Flattening relocates complexity rather than removing it. The coordination, context-carrying, and priority-setting that middle layers handled don’t disappear; they shift onto fewer leaders with broader spans, making the organization structurally simpler but operationally more dependent on individual leadership capacity.

What is a healthy span of control for a manager?

There’s no universal number, but research links wider spans to weaker engagement and higher turnover. Gusto found the average manager now oversees nearly six direct reports, double the 2019 figure. Beyond roughly seven reports, coaching, context-sharing, and development are usually the first things to degrade.

What is a network-of-teams operating model?

It’s a design built around small, autonomous, cross-functional teams aligned to products or outcomes rather than functional silos. It accelerates work inside each team but scatters coordination across teams, so it requires deliberate leadership to keep local decisions aligned with global goals.

Previous: Part 1  |  Next: Part 3

About this series: This article and the accompanying ebook, The Modern Technology Organization: How to Hire and Structure Teams for Today’s Org Design Models, were developed by Paragon by Riviera Partners, a leader in helping top and growing technology companies hire, scale, and structure their teams. Drawing on our work with hundreds of technology organizations, Paragon partners with CEOs and HR leaders to identify, evaluate, and place the leaders and talent who can scale decision-making, maintain alignment across teams, and turn ambition into execution. Download the full ebook or learn more about Paragon.