Most engineering teams are not underperforming. They are being drained.
The drag is invisible in sprint velocity, headcount reports, and delivery timelines — but it shows up every quarter in margin, morale, and missed output. The engineers are working. The hours are being logged. The standups are happening. And yet the output per engineer, measured in shipped value, is a fraction of what the payroll justifies.
This is not a talent problem. It is a systems problem. And it is costing you more than you think.
Key Stat
Companies with 50–200 engineers typically lose 30–40% of their total engineering capacity to invisible drag — before a single line of code is written.
The Six Hidden Drains on Engineering Output
Engineering productivity does not leak through one large failure. It leaks through six compounding drains, each of which is individually manageable and collectively devastating.
1. Unclear Ownership and Blurred Accountability
When an engineer does not know who owns a decision, they stop making it. They wait. They escalate. They schedule a meeting to discuss whether to schedule another meeting.
In most engineering organisations, ownership is assumed rather than assigned. The product manager thinks the tech lead owns the architecture decision. The tech lead thinks the product manager owns the priority call. The engineer in the middle waits for clarity that never arrives.
The most expensive thing in engineering is not a bad decision. It is the absence of a decision — because the cost of waiting compounds every day.
The result is not laziness. It is rational behaviour in a system that has not defined who is responsible for what. Engineers learn quickly that making a decision without clear authority creates risk for them personally. Waiting is safer.
What this costs: In a team of 50 engineers, if each engineer loses an average of 3 hours per week to unclear ownership — waiting for decisions, attending clarification meetings, reworking after a decision is finally made — that is 150 engineer-hours per week. At a fully-loaded cost of $80 per hour, that is $12,000 per week, or $624,000 per year. From one drain.
2. Onboarding That Does Not Transfer Knowledge
The average time for a new engineer to reach full productivity in a mid-sized technology company is 6–9 months. In companies with poor onboarding, it is 12–18 months.
Most companies treat onboarding as an HR process: paperwork, system access, a welcome lunch. The technical onboarding — the transfer of context, architecture decisions, codebase conventions, and institutional knowledge — is left to whoever has time, which means it is left to whoever can least afford to give it.
The senior engineer who spends 30% of their time onboarding a new hire is not delivering. The new hire who spends 6 months finding their feet is not delivering. Both costs are invisible in the headcount report.
What this costs: If a senior engineer earning $150,000 per year spends 30% of their time onboarding new hires, that is $45,000 in senior capacity consumed annually per new hire cycle. Multiply by the number of hires per year and the cost becomes significant before the new hire has shipped a single feature.
3. Technical Debt That Is Never Priced
Technical debt is not a technical problem. It is a financial problem that is being managed as a technical one.
Every shortcut taken to meet a deadline, every workaround built to avoid a refactor, every legacy system kept running because migration is expensive — these are liabilities on your engineering balance sheet. They slow every future feature. They increase the cost of every bug fix. They make your codebase harder to understand, harder to test, and harder to change.
The problem is not that technical debt exists. It always will. The problem is that it is never priced. No one has calculated what the existing debt costs in engineering hours per quarter. No one has compared that cost to the cost of addressing it. The debt grows because the decision to address it is never made on financial terms.
Key Stat
In mature codebases, technical debt typically consumes 20–30% of total engineering capacity — not in dedicated debt-reduction sprints, but in the friction it adds to every other task.
4. Meetings That Consume Without Producing
The average software engineer attends 15–20 hours of meetings per week. In engineering leadership roles, that number is often 25–35 hours.
Not all of those meetings are unnecessary. Some are essential. But a significant proportion are meetings that exist because a process requires them, not because they produce a decision or advance work. The daily standup that has become a status report. The sprint planning session that takes four hours to produce a backlog that could have been written in one. The all-hands that communicates information that could have been an email.
Every hour an engineer spends in a meeting that does not require their specific input is an hour of engineering capacity converted into attendance.
What this costs: If 50 engineers each lose 5 hours per week to meetings that do not require their input, that is 250 engineer-hours per week. At $80 per hour fully loaded, that is $20,000 per week, or $1,040,000 per year. From meetings alone.
5. Rework Caused by Late Requirements
The cost of fixing a defect increases by a factor of 5–10 for every stage it moves through the development lifecycle. A requirement that is unclear at the design stage costs 10 times more to fix after it has been built than it would have cost to clarify before work began.
Most engineering teams accept rework as a normal cost of doing business. It is not. It is a symptom of a requirements process that is not producing clarity before work begins. The engineers are not failing. The process that hands them incomplete specifications is failing.
Key Stat
In teams without a structured requirements review process, rework typically accounts for 15–25% of total engineering output — meaning that 15–25% of what your engineers build is built twice.
6. Attrition of Senior Engineers
The most expensive form of engineering productivity loss is the one that appears on no report: the departure of a senior engineer who carries institutional knowledge, architectural context, and the informal authority to make decisions quickly.
When a senior engineer leaves, the visible cost is the recruitment fee and the salary of their replacement. The invisible cost is the 12–18 months it takes for their replacement to reach the same level of contextual understanding — and the drag that creates on every engineer who worked alongside them.
What this costs: The fully-loaded cost of replacing a senior engineer — recruitment, onboarding, productivity ramp, knowledge transfer, and the drag on surrounding team members — is typically 1.5–2x their annual salary. For a senior engineer earning $180,000, that is $270,000–$360,000 per departure.
The Aggregate Cost: What the Six Drains Add Up To
The following table shows a conservative estimate of the annual cost of each drain for a team of 100 engineers with an average fully-loaded cost of $100,000 per engineer per year (total payroll: $10M).
| Drain | Estimated Capacity Lost | Annual Cost (100-engineer team) |
|---|---|---|
| Unclear ownership | 8–12% | $800K–$1.2M |
| Poor onboarding | 5–8% | $500K–$800K |
| Technical debt friction | 20–30% | $2M–$3M |
| Unproductive meetings | 10–15% | $1M–$1.5M |
| Rework from late requirements | 15–25% | $1.5M–$2.5M |
| Senior attrition | 5–10% | $500K–$1M |
| Total | 63–100% | $6.3M–$10M |
Key Stat
In a 100-engineer team with a $10M payroll, the six drains together can consume the equivalent of the entire payroll — meaning you are paying for 100 engineers and receiving the output of 50 to 60.
What a Recovery Looks Like
The recovery does not require new tools, new processes, or new headcount. It requires clarity on three things: where the drag is, what it costs, and who owns fixing it.
Step 1: Diagnose the drag. Map each of the six drains against your current team. Quantify the cost of each one in engineer-hours per week. This is not a survey or a culture assessment — it is a financial analysis of where your engineering payroll is going.
Step 2: Price the recovery. For each drain, calculate the cost of addressing it versus the cost of leaving it in place. In most cases, the cost of addressing it is a fraction of the ongoing drain. Technical debt that costs $2M per year in friction can often be addressed with a 3-month focused sprint costing $300K.
Step 3: Assign ownership. Each drain needs a single owner with a clear metric and a deadline. Not a committee. Not a working group. One person who is accountable for the number moving.
Step 4: Measure output, not activity. Replace velocity metrics and story points with output metrics: features shipped per engineer per quarter, defect rate per release, time from requirement to production. These are the numbers that connect engineering performance to business value.
*M.K. Hasan is the Founder of EVOSYST and the author of The Hidden Leak. He works with CEOs and Boards to find the revenue their business is losing and lead the recovery.*
Key Takeaways
- Most engineering teams are not underperforming — they are being drained by six compounding, invisible forces.
- The six drains — unclear ownership, poor onboarding, technical debt, unproductive meetings, rework, and senior attrition — can consume 30–50% of total engineering capacity.
- In a 100-engineer team, the aggregate annual cost of these drains is typically $6M–$10M.
- The recovery does not require new tools or new headcount. It requires a financial diagnosis of where the drag is, followed by clear ownership and measurable targets.
- The value is already inside the business. The engineering team is already there. The question is how much of their capacity is reaching the product.
About the Author
M.K. Hasan
Founder, EVOSYST · Global CHRO & Strategic Advisor to CEOs
M.K. Hasan helps CEOs and Boards find the revenue their business is losing and get it back. He works across Strategy, People and Operations through focused diagnostic and recovery sprints.
Request an Executive Leak Assessment