How to Manage Multiple Projects: A Capacity Check First
To manage multiple projects, keep active projects to two or three per person, then ask your sponsor which project yields when hours run short, because tools and priority labels cannot create capacity.
By Sanjeev Kumar, Founder · Updated 15 Nov 2026

Key takeaways
- Two to three active projects per person is the range most of the studies reviewed in a 2015 paper support; for complex projects we recommend two.
- Every project beyond the first costs its own share of your week plus switching time. In our model, a 45-hour week gives 18 hours per project at two projects, 9 at three and 4.5 at four.
- The decision that matters most is who sets priority between projects. Take that person your hours and the project that will slip, not a complaint about workload.
In our working terms (see What Is Program Management?), projects that share an intended benefit are a programme. Projects that share only people, budget or a sponsor are managed as a portfolio, and unrelated projects are separate jobs.
In this article
- 1How to manage multiple projects: how many can one person carry?
- 2What does each extra project cost in weekly hours?
- 3Why does a tracker not fix an overloaded week?
- 4How do you run a capacity check on your own projects?
- 5Step 1: Can you name your real weekly hours and the priority-setter for each project?
- 6Step 2: Is your project count within the range the research supports?
- 7Step 3: Do your projects share a benefit, or only people and money?
- 8Step 4: Do you have a named decision on which project yields?
- 9Who decides which project slips when the hours do not add up?
- 10What changes when your projects belong to a programme or a portfolio?
- 11Which of our courses helps when your projects have become a programme?
- 12Frequently asked questions
Three projects at 15 hours a week each is not 45 hours of progress. On our model it is closer to 27, because the other 18 hours go on context switching, the time lost each time your attention moves from one project to another. Add a fourth project and the model leaves 18 usable hours.
The usual response is to organise harder. The better response is to treat your hours as a fixed budget and decide which project gets which share of it.
The check ends with a decision that belongs to someone else, usually your sponsor, the person who funds a project and sets its priority. Knowing that early saves weeks of working harder on the wrong problem.
Some readers will find that their projects share a goal. That is a different job from running several unrelated projects, and the last sections explain how to tell the two apart.
How to manage multiple projects: how many can one person carry?
Two to three active projects per person is the range most of the studies reviewed in a 2015 paper support. One study of 64 high-technology firms put two as optimal. For complex projects we recommend two.
Studies of engineering and technology organisations, collected in a 2015 paper in the South African Journal of Industrial Engineering, point the same way. Two projects "seemed to be optimal" and one to three were not problematic. A survey of 64 high-technology firms agreed. Managers across five manufacturing engineering departments put the effective load at two to three major projects. Where people work on several projects at once, engineers can spend as little as 25 to 30% of their time on value-adding tasks (Wheelwright and Clark, Harvard Business School, 1992, quoted in the 2015 paper). Why the loss grows is our inference: coordination and switching take time.
The paper's own survey had 117 usable responses, a 5% response rate, mostly from engineering and construction, mining and government. Its authors recommend fewer projects per person and staggered starts.
Industrial data agrees. A study presented at CHASE 2020 (the Cooperative and Human Aspects of Software Engineering workshop) covered 1,213 developers across 282 projects at one Brazilian technology company. They worked on 2.41 projects a week on average and multitasked in 22.37% of developer-weeks. Focusing on fewer projects, and working repetitively from one day to the next, was associated with higher productivity.
Both studies cover engineers and developers, not project managers. Applying them to your role is our inference. The mechanism is the same: attention split across separate contexts.
What does each extra project cost in weekly hours?
Each extra project costs more than its own share of the week. In our model, one project gets all 45 hours, three projects get 9 hours each and five get 1.8 hours each, because every project beyond the first removes 20% of the week to switching between them.
The anchor is a rule of thumb from Gerald Weinberg's 1992 book Quality Software Management, reported by Scrum.org: with two projects, each gets 40% of your effort and 20% is lost. We have not checked that figure against the original book. It is a heuristic, not a measurement. For three to five projects we assume the same 20-point loss for each extra project, purely as an illustration; real losses are unlikely to be linear and the model reaches zero at six projects.
Our model. Usable share of the week = 100% − 20 points × (projects − 1). Hours per project = 45 × usable share ÷ projects. We assume 45 hours of project work and an equal split.

| Active projects | Usable share of week | Usable hours | Hours per project | Hours lost to switching |
|---|---|---|---|---|
| 1 | 100% | 45 | 45 | 0 |
| 2 | 80% | 36 | 18 | 9 |
| 3 | 60% | 27 | 9 | 18 |
| 4 | 40% | 18 | 4.5 | 27 |
| 5 | 20% | 9 | 1.8 | 36 |
Real weeks include meetings, so your real figures will be lower. The loss per project also varies with how different the projects are. Treat the table as a way to see the shape of the problem, not as an official figure.
Why does a tracker not fix an overloaded week?
A tracker shows the work but adds no hours. If five projects share nine usable hours in our model, a dashboard shows five red rows more clearly and changes nothing about the week.
Priority labels have the same limit. High, medium and low give each project the same weekly hours until someone decides what yields. A tracker or spreadsheet is useful once the count is settled, because it can hold the decisions you make below. Run the check first.
How do you run a capacity check on your own projects?
Run the capacity check by listing each active project with your real weekly hours and its named priority-setter, comparing your count with the table, classifying what the projects share, then sending your sponsor the trade-off. Each step has a pass or fail test.
Step 1: Can you name your real weekly hours and the priority-setter for each project?
Take the last four weeks of your calendar and time records. For each project, write the average hours you spent and the name of the person who sets its priority.
- Pass: every project has an hours figure and a named person.
- Fail: any project has no named person. That is your first conversation.
Step 2: Is your project count within the range the research supports?
Count the projects that need your attention every week, and look up the hours per project in the table above. The thresholds below are our recommendation, not a research finding.
- Pass: three or fewer.
- Borderline: four, acceptable only if the projects are simple and well staffed (our judgement).
- Fail: five or more. Your hours per project will be below 2 in our model.
Step 3: Do your projects share a benefit, or only people and money?
For each pair of projects, ask whether a delay in one changes what the other delivers. Then use the table.
| What the projects share | Type of work | First move |
|---|---|---|
| Nothing | Separate jobs held by one person | Cap active projects at three and stagger starts |
| People, budget or a sponsor only | Portfolio-type work | Rank by value and limit what is in progress |
| An intended benefit and dependencies | Programme-type work | Manage dependencies, risks and benefits across projects together |
- Pass: you can place every project in a row and say why.
- Fail: you cannot say what two of your projects share. Ask the sponsor what each is meant to deliver.
Step 4: Do you have a named decision on which project yields?
Send this to the person who sets priority, with your hours table attached:
I am running [number] projects at about [hours] hours a week each. If nothing changes, [project] slips by [weeks] weeks. Which project should take the slot, or can one start later?
- Pass: a named person answers in writing which project yields.
- Fail: no answer. Your own plan has to name the project that slips, with a date, and you tell the sponsor.
Who decides which project slips when the hours do not add up?
The person who sets business priority for each project decides, normally your sponsor, not you. Your task is to bring the trade-off in hours and weeks, so the choice is made on purpose and not by whichever project is loudest.
Offer three options, in this order:
- Stagger starts. The 2015 survey's authors recommend this to cut concurrent load.
- Hand one project to someone else.
- Let one project slip, with a new date.
Our recommendation is to leave slack between deliverables across your projects, so one unexpected problem does not land in the same week as another project's deadline.
For your own days, our recommendation is to give each block of the day to one project and switch at natural stopping points, where nothing is half finished. In a study of 1,213 developers, working on fewer projects and following a similar pattern from one day to the next was associated with higher productivity; switching more often within a day showed no significant effect.

What changes when your projects belong to a programme or a portfolio?
When projects share an intended benefit, you are doing programme management: coordinating dependencies, risks and benefits across them, not running more projects. When they share only people or budget, you are doing portfolio management: choosing and balancing projects against the capacity available.
Those are our working definitions. Programme work and portfolio work call for different tools from a personal task list, because the unit you manage is the benefit or the mix, not the task. Program vs Portfolio Management: Which Seat Are You In? helps you decide which one you hold, and What Is Program Management? The Benefit Test explains the test. If your main constraint is shared people, see Portfolio Resource Management.
If step 3 put most of your projects in the last row, the answer is a different skill set, not more effort.
Which of our courses helps when your projects have become a programme?
Program Management Essentials suits readers whose step 3 result is programme-type work. It is a 12-week foundation course with 10 modules, live Q&A, templates and case studies, led by our founder Sanjeev Kumar. Modules include Program Risk Management, Program Stakeholder Engagement and Benefits Planning & Value Realization. It is listed for professionals managing multiple projects. As of November 2026 it costs US$99. It is a foundation course, not preparation for the Program Management Professional (PgMP®) exam from the Project Management Institute (PMI).
Program Leadership Blueprint is a starter course for moving from project to programme manager, with five modules, live sessions and case studies. As of November 2026 it costs US$10. Prices vary by market; check each course page for your price.
We have mentored or trained more than 6,000 professionals. We specialise in programme and portfolio management, and we prepare candidates for the PgMP® and Portfolio Management Professional (PfMP®) credentials. PMI awards those credentials, not us.
If you are unsure which course fits, we offer a free discovery consultation before enrolment. Write to support@augmentconsultancy.com or call or WhatsApp +91 9816 360 116.
Frequently asked questions
Rank first the tasks that block other people's work, because their delay spreads. These sit on the critical path, the chain of tasks that sets the earliest finish date. Then take the tasks on the project whose priority your sponsor has confirmed.
Sources
- 1.Steyn, H. and Schnetler, R., "Concurrent projects: How many can you handle?", South African Journal of Industrial Engineering 26(3), 2015
- 2.Kohl, K., Vasilescu, B. and Prikladnicki, R., "Multitasking Across Industry Projects: A Replication Study", CHASE 2020 (Cooperative and Human Aspects of Software Engineering)
- 3.Trapps, S., "The Financial Cost of Task Switching", Scrum.org, 7 March 2021 (reporting Weinberg, Quality Software Management, 1992)
- 4.PMI, Program Management Professional (PgMP)
- 5.Augment Consultancy, Program Management Essentials
- 6.Augment Consultancy, Program Leadership Blueprint



