The coordination tax: the hidden cost your operation pays every day

In short: The coordination tax is the accumulated cost of the meetings, approvals and manual handoffs an operation needs because its systems depend on a person to pass information along. It grows with every sale, channel and new hire, and you measure it by counting the human touches in one process for a week. You cut it by designing that process so repeatable work runs on its own.

The coordination tax is the hidden cost of coordination inside an operation: the hours a team spends moving data between systems, waiting on approvals and asking where each case stands. It is what operational fragmentation, an operation split across systems that don't talk to each other, costs your team.

I saw it up close at Walmart Chile, where I worked with my co-founder, Octavio Flores, now CTO of Puente OS. I ran operations and he ran the technical side. Very talented people spent hours a day pasting data from one system into another. Today I see it in almost every client meeting: the email to confirm a price, the WhatsApp message to check whether a payment landed, the master spreadsheet someone updates so a colleague can keep going.

Human middleware. That is what the tax looks like.

I'm Cris Ugarte, CEO and co-founder of Puente OS, the platform that builds the systems that run your business, for retail, ecommerce and logistics companies in Chile, Peru and Mexico.

What is the coordination tax?

The coordination tax is the accumulated cost of the meetings, approvals and manual handoffs an operation needs because its systems depend on a person to pass information along. It is paid in team hours and in days of waiting. It is called a tax because it is charged on every transaction, every day, and no budget ever approved it.

Project managers call it coordination overhead. The term coordination tax shows up in a March 2026 study by Front, a customer operations software company, and in In Parallel's index, which defines it as "the cumulative cost of keeping people aligned." Our definition focuses on operations: orders, shipments, payments and returns.

What does the coordination tax look like in retail, ecommerce and logistics?

The coordination tax looks like normal work, which is why it is hard to spot. It shows up in four forms: swivel-chair data entry between systems, approvals waiting on a person, status checks and meetings to find out where a case stands, and month-end closes reconciled by hand. Together they fill much of an operations team's day.

• Manual handoffs (swivel-chair integration). An order lands in the storefront and an ops analyst re-keys it into the courier portal. A payment hits the bank and someone looks it up in the ERP to mark the invoice paid. Every handoff is a chance for error: a swapped digit, a row deleted from the master spreadsheet.
• Approvals. A late return, a discount above the limit. The case sits until the approver sees the message. The customer, the cash and the truck all wait.
• Status checks and meetings. To answer "where is my order?", a coordinator opens three systems and asks the team's WhatsApp group: "Hey, did anyone see if order 4812 shipped?" Whatever doesn't get solved there ends up in the 9 a.m. meeting.
• Closes and reconciliations. At month-end the team reconciles what the marketplaces, couriers, payment gateway and bank each say, with spreadsheets and one more meeting.

How much does the coordination tax cost?

Research puts coordination at 41% to 60% of the workweek, depending on how it is measured: Asana reports 60% on work about work, and In Parallel found 41% once managers itemized their hours. In your operation, the coordination tax is measured in team hours per process, and those hours can be priced in dollars.

• Asana. Its Anatomy of Work Index, which surveyed more than 10,000 knowledge workers, defines work about work as communicating about work, searching for information, switching between apps, managing shifting priorities and chasing the status of work. People spend 60% of their time on it.
• Front. In its survey of 665 B2B customer service, operations and account management leaders, the typical company spends nearly 3 hours coordinating for every hour spent solving customer problems. Only 14% spend more time solving than coordinating, and 42% don't track coordination time at all.
• Microsoft. Its 2023 Work Trend Index measured, across Microsoft 365 apps, that the average employee spends 57% of their time communicating in meetings, email and chat.
• In Parallel. In its June 2026 index of 247 managers across five countries, respondents estimated that coordination takes 21% of their week. When they itemized their hours, the total came to 41%.
• Harvard Business Review. Rob Cross, Reb Rebele and Adam Grant reported in Collaborative Overload that the time managers and employees spend on collaborative activities has ballooned by 50% or more over two decades.

What the research says about AI

In Front's study, 93% of companies use AI in customer operations, and the coordination tax stayed where it was. In the In Parallel index, managers who use AI daily carry more coordination load than those who don't: 20.3 versus 9.1 hours a week. The tax comes down when the process design changes, and AI earns its keep by building the new process.

An honest caveat: these are measurements of office work, much of it self-reported, and several come from companies that sell software for this problem. None measures a retail, ecommerce or logistics operation in Latin America. The number that matters is the one from your own operation.

Why does the coordination tax grow as the business grows?

The coordination tax scales with volume and with the number of parties that need to stay in sync. Every sale brings its handoffs, every new channel adds a spreadsheet, every courier adds a portal and every new hire opens more conversations. That is why cost per order stops falling right when the business accelerates.

The arithmetic behind it is simple. Three people have three possible conversations; six have fifteen; ten have forty-five. Fred Brooks described it in 1975 for software projects, and it applies just as well to an operation.

The usual fixes move the tax somewhere else:

• Hiring more people adds capacity and also adds coordination channels: operations grow at the pace of headcount.
• Adding another tool solves one task and adds one more system to check.
• Point-to-point integrations move data between two systems. Exceptions keep arriving by email and WhatsApp, and when the integration breaks, the manual handoff comes back. The missing layer is a system of action that acts across all your systems at once.

The more channels, systems and people, the harder it hits.

How do you measure the coordination tax?

You measure the coordination tax one process at a time, with real cases. For one week, count how many times a person moves data, how many times a case waits for a reply and how many tools get opened. Multiply the touches by the time each one takes and by monthly volume to get the hours that process pays.

1. Pick one process with volume: order fulfillment, payment reconciliation, returns or price updates. Just one.
2. Map the path of a real case, from whatever triggers it to the moment it closes, naming every system it touches.
3. Mark every human touch: handoff (copy, download, upload), wait (an approval, a reply) or lookup (opening another tool, asking a colleague).
4. Time about twenty cases, including the odd ones. That is where much of the cost hides.
5. Multiply touches per case by minutes per touch by cases per month. That number is people's hours.
6. Add waiting time separately. Approvals cost little labor and a lot of calendar time: a customer waiting, cash stuck, a truck leaving late.
7. Note who makes each touch. It is usually the same two or three people: your human middleware.
8. Compare with what the team estimated. The measured number will likely come out higher: in the In Parallel index, itemized hours came to nearly double the estimate.
9. Put a price on it. Hours times the loaded hourly cost of the team, plus the cost of errors: reshipments, credit notes, chargebacks.

A worked example

With round numbers: the ecommerce order described below has four human touches of two minutes each. At 3,000 orders a month, that is 24,000 minutes, or 400 hours: more than two full-time people moving data between systems.

That figure is your baseline, and multiplied by the loaded hourly cost, it is the tax in dollars. From there, every improvement is measured in hours given back to the team.

Which metrics to track next

• Touches per case, the number the redesign has to bring down.
• Touchless rate, the share of cases closed with no human intervention. It is the metric behind the 97.9% figure below.
• Approval wait time, in hours or days.
• Exceptions per 100 cases, with the reason for each. When they rise, there is a new rule to write down.

What is the difference between coordination that adds judgment and the coordination tax?

Coordination that adds judgment sets a price, negotiates with a supplier or resolves a case no rule anticipated. The coordination tax is the coordination that exists because systems depend on a person to pass information to each other. The first kind is worth protecting. The second kind gets redesigned until it runs on its own.

Approving a 40% discount for a wholesale customer who is threatening to leave is judgment. Approving a return that meets policy is tax. Here is how they compare:

• Who does it. Judgment: the person who knows the customer, the margin and the context. Tax: whoever has both systems open.
• What it adds. Judgment: a decision that changes the outcome. Tax: a piece of data another system already had.
• How it scales. Judgment: with the complexity of the business. Tax: with volume, order by order.
• What happens after redesign. Judgment: it reaches the right person with full context. Tax: a workflow runs it under your rules.

How do you reduce the coordination tax?

You reduce the coordination tax by designing each process to meet three conditions: work starts on its own when data changes, systems talk to each other with no person in the middle, and the system executes the clear cases and routes whatever needs judgment to the right person. Those three conditions are the foundation of autonomous operations design.

• Triggers: a paid order or stock below the minimum kicks off the next step.
• Connected systems: data flows from the storefront to the ERP and the courier. Your ERP remains your system of record.
• Exceptions with context: they arrive with the customer's history, the order's margin and the rule that wasn't met.

At Puente OS we call this discipline Diseño Autónomo. I cover it in depth in how to design operations that run themselves.

Where it runs: the system of action

That is what we build at Puente OS: in our clients' stacks, Puente is the system of action. Your ERP records what happened. Your BI shows you what happened. Puente makes it happen. I explain the concept in what a system of action is and where it fits in your stack.

With Puente OS you describe in plain language what your operation needs and get software running. Every redesigned process combines four building blocks:

• Workflows, with explicit, auditable rules, for repeatable work: validate, assign, generate, notify.
• Databases, holding the context that spans systems. Where operations live in spreadsheets, they also keep the record.
• Apps, where the team sees exceptions, decides and approves.
• Agents, for reading language or applying judgment, like answering a customer on WhatsApp.

Every action is logged, within your rules and with your approvals. A process goes live in production in 2 to 4 weeks.

What does an ecommerce order look like with and without the coordination tax?

With the coordination tax, the order depends on a person at every system boundary: four human touches before the customer gets a tracking number. In the redesigned flow, the order is validated, assigned to a courier and labeled automatically, and the customer is notified. The team receives only the exceptions.

Before: the order with the coordination tax

1. An ops analyst reviews the order in the marketplace dashboard and checks the payment in another tab.
2. The analyst copies the address into the courier portal and downloads the label.
3. They update the master spreadsheet with the tracking number.
4. When the customer asks, a support rep opens three systems to answer.

Four touches per order.

After: the redesigned order

1. A workflow validates the order against your payment, stock and address rules.
2. The workflow assigns the courier by zone, weight and delivery window, and generates the label.
3. The tracking number flows back to the marketplace and the customer gets notified.
4. If something doesn't match, the order lands in an app with the reason and the context, and a person decides.

From design goal to measured data

Picture a different Monday. You walk in and the weekend's operations are already closed out: orders shipped, payments reconciled, customers notified. Your team touches one order in a hundred, the ones that need judgment. That Monday is a design goal.

The closest measured data point we have: at one of our clients, a multichannel electronics ecommerce business in Mexico, 97.9% of orders get their shipping label with no human intervention. Generating labels used to take 5 to 7 hours per cutoff, with two cutoffs a day. Today it takes 15 minutes. Its monthly volume grew 40% with the same operations team.

At a last-mile courier in Chile, its 663-location agency network is monitored automatically every week: locations that stop printing labels enter a contact campaign. Over seven weeks, weekly label volume across the network rose 46.5% with the same team.

Which process should you redesign first to cut the coordination tax?

Start with the process that pays the most coordination tax and has rules you can write down: high volume, repeatable steps, accessible data and a pain the team already names. We start with a single process that goes live in production in 2 to 4 weeks and gets measured against its baseline.

• Volume. Every touch you remove gets multiplied by the month's cases.
• Rules you can write down. If the team can explain how it decides, a workflow can execute it.
• Accessible data through an API, a file export or email.
• An owner on the team who knows the process and decides the exceptions.

That is how you automate operational processes with results you can measure. For the processes we prioritize, we aim for payback in under 90 days, and as each process runs on its own, operations can grow 2 to 5 times without multiplying headcount.

Key takeaways

• The coordination tax is the cost of the meetings, approvals and manual handoffs that exist because systems depend on a person.
• It grows with volume, channels, systems and headcount.
• You measure it in one week, one process at a time: touches per case, times minutes, times monthly volume, plus waiting time and errors.
• Research puts coordination at 41% to 60% of the workweek.
• You reduce it by design: work triggered by data, systems that talk to each other and a team that decides what needs judgment.

---

Want to measure the coordination tax in your operation? We start with a one-month pilot: we map and prioritize your processes, and the first one goes live in production. If you decide to continue, the pilot fee is credited in full toward implementation. Book a demo and let's start with the process that costs you the most.