IIIOrganization & Project ManagementGuide
Too Many Projects at Once: Running a Project Portfolio
New website, new software, moving the stockroom, launching a maintenance contract: in many businesses, more projects are running than the team can handle. None of them is wrong. Together they slow each other down.
Key takeaways
- A project portfolio is the sum of all projects a business pursues alongside day-to-day operations. Running it means deciding which ones run, which wait, and which are stopped.
- According to Little's law, lead time equals work in progress divided by throughput. Whoever starts more projects at once without more capacity extends the duration of each one.
- In a survey reported by Cooper (2009), 76 percent of companies said they had too many development projects for their resources. Projects were rarely stopped.
- A longitudinal study with 580 employees shows an inverted U-shape: some parallel work helps on-time delivery, too much harms it (Colicev, Hakkarainen, and Pedersen 2023).
Why does nothing get finished even though everyone works hard?
In many small businesses, the project list does not arise from planning but from occasions. A customer would like a customer portal, the tax advisor recommends new accounting software, the website is outdated, and the stockroom is bursting at the seams. Each project is started because it makes sense. Hardly any is finished, because the same three people work on all of them, alongside customers, quotes, and complaints.
The result is a state owners know well: everyone is busy, but at the monthly meeting every project reports that it is on track and only slightly delayed. This leads to the guiding question of this guide: how does a small business run its projects as a portfolio so that fewer run at the same time and more get finished? To answer it, we clarify the terms, present a thinking model from queuing theory and portfolio research, review the findings, weigh two perspectives, and describe a one-page portfolio.
A business that says yes to everything is really saying later to everything.
What is a project portfolio?
A project is a one-off undertaking with a fixed result, a beginning, and an end, as the guide on planning projects in small teams states. The project portfolio is the sum of all projects a business is pursuing or wants to pursue at the same time. Running a portfolio means regularly deciding which projects are actively worked on, which wait, and which are stopped. This differs from planning a single project. There it is about milestones and owners; here it is about whether and when a project gets time at all.
| Feature | Project planning | Portfolio management |
|---|---|---|
| Question | How do we reach this goal? | Which goals do we pursue now, which later? |
| Object | one project | all projects together |
| Rhythm | weekly within the project | monthly or quarterly |
| Who decides | project owner | owner or management meeting |
| Typical decision | move a milestone | start, pause, or stop a project |
The thinking model: Little's law and the funnel that must not become a tunnel
In 1961, the mathematician John Little proved a relationship that holds in every queue. Little's law states: the average lead time equals the average work in progress divided by throughput. The guide on measuring on-time delivery applies it to orders. It holds equally for projects. If a team has capacity to complete one project a month and six projects are running, each takes six months on average. If three are running, it is three months. The work does not shrink, but each project is finished earlier and starts delivering benefits earlier.
The product development researcher Robert Cooper captured the consequence for managing projects in an image. A selection process for projects should be a funnel: many ideas go in at the top, decisions are made at several checkpoints, and only the best come out at the bottom. In many companies it is instead a tunnel: whatever goes in also comes out, good and bad projects alike, because nobody says no at the checkpoints (Cooper 2009). According to Cooper, Edgett, and Kleinschmidt (2002), good portfolio management therefore explicitly pursues the right number of projects in addition to their value, their mix, and their link to strategy.
- Ideas and occasions
- Briefly assessed58% of previous stage
- Approved57% of previous stage
- Completed75% of previous stage
Worked example with freely chosen figures. From 24 occasions come six completed projects. A tunnel would have started all 24 and finished hardly any.
What studies show about too many parallel projects
How widespread the problem is can be seen in a survey reported by Cooper (2009). It concerns companies with their own product development. 76 percent said they had too many projects for the available resources, three quarters admitted to prioritizing their projects poorly, and almost 79 percent had no systematic method for selecting projects. Only a third had, by their own assessment, tough checkpoints at which weak projects are actually stopped. One manager summed this up in a sentence Cooper quotes: 'We never kill projects, we just wound them.'
Source: Cooper (2009)
What this means for employees was studied by Colicev, Hakkarainen, and Pedersen (2023). They analyzed 20 months of project data from a global manufacturer of hydraulic pumps: 42 development projects, 580 employees, 9,649 observations per person, project, and month. Working on several projects at once initially helped on-time delivery, because idle time in one project was filled with work in another. Beyond a certain point, this reversed. The result was an inverted U-shape. Switching between similar projects and with familiar colleagues was less harmful.
Zika-Viktorsson, Sundström, and Engwall (2006) coined a term for this: project overload. It refers to the experience of fragmentation, disruption, and inefficiency that arises when someone switches between tasks from different, simultaneous projects. The authors describe multi-project settings as characterized by tight schedules, multitasking, growing coordination effort, and a lot of set-up time with every switch.
Start everything or finish a few? Two perspectives
The first perspective says: many projects at once are efficient. Whoever is waiting for a customer's approval can work on the next project. Idle time is avoided, and the team stays busy. The first part of the findings by Colicev and colleagues supports this: a certain amount of parallel work improved on-time delivery.
The second says: finishing a few projects is more efficient. Every switch costs set-up time, every additional construction site extends all the others according to Little's law, and a half-finished project delivers no benefit. The second part of the inverted U-shape and the findings on project overload support this.
Everything in parallel
- every project makes progress at some point
- waiting times are filled with other work
- every project takes longer
- benefits arrive late and all at once
Limited number of active projects
- two or three projects active, the rest visibly waiting
- a second project fills the waiting times of the first
- every project is finished earlier
- benefits arrive earlier and one after another
I consider both perspectives justified, and the inverted U-shape shows how they fit together. A second project alongside the first makes sense because it fills waiting times. A fifth no longer does. The conclusion for a small business is therefore not to run only one project, but to set a deliberate limit, for example two active projects per person, and to let everything else wait visibly.
The portfolio on one page: how to run it
Inventory of all projects
Write down every running and every promised project, including those only someone has in their head. One line per project: goal, responsible person, estimated hours to completion.
Determine project time per week
How much plannable time really remains for projects after day-to-day business? The guide on capacity planning shows that in many teams it is less than expected.
Set a limit for active projects
For example two active projects per person and a total number for the business. Everything above that is not deleted but set to waiting.
Set the sequence
Which project costs the most if it is finished a week later? This question about the cost of delay decides the order.
Allow stopping
Every project has a review date. A project whose occasion has disappeared or that has missed its milestone twice is deliberately stopped or rescoped.
Review monthly in 30 minutes
The management meeting goes through the list: what is finished, what moves up, what is stopped? New projects start only when a slot becomes free.
| Project | Owner | Remaining hours | Cost of delay | Status |
|---|---|---|---|---|
| Renew quote template | Sales lead | 30 | high | active |
| New accounting software | Accounting | 80 | medium, deadline at year-end | active |
| Reorganize stockroom | Foreman | 60 | medium | active |
| New website | Owner | 120 | low | waiting |
| Customer portal | Owner | 200 | unclear | waiting, check occasion |
| Fall trade show | Sales | 50 | fixed date in September | starts in June |
Worked example
For the sequence, a rough classification is usually enough in a small business. Whoever wants to calculate more precisely will find the method of cost of delay per week of work in the comparison prioritizing in a team. There, a short project with medium cost of delay beats a long one with high cost, because it is finished earlier and then stops costing money.
A worked example shows the effect of the limit. A team has 30 hours per week for projects and six projects of 60 hours each. If all run in parallel, each gets five hours per week and is finished after twelve weeks, more like 14 with switching costs. If two at a time run one after another, the first two are finished after four weeks, the next after eight, and the last after twelve. The total duration stays the same, but on average each project is usable four to six weeks earlier.
Worked example. The switching cost of around 15 percent is freely assumed. The calculation follows Little's law: same throughput, less work in progress, shorter lead time.
What follows beyond the individual case, and where the limits lie
The hardest element is not the table but the no. Whoever sets a project to waiting disappoints someone, often an employee who championed it, sometimes the owner. It therefore helps if the rule is fixed in advance: a new project starts when a slot becomes free. Then it is not the loudness of a wish that decides, but the sequence.
One limit lies in day-to-day business. Projects compete not only with each other but with customers and complaints. When day-to-day business grows, project time shrinks, and the limit has to go down. Whether a running project is still on track is shown by the project sheet from the guide on project controlling for small businesses. And how the limit on parallel work can be made visible on a board is described in the guide on Kanban in sales.
Finally, the portfolio links strategy with everyday work. A business that has chosen a direction can see from the portfolio whether the running projects fit it. Why good plans otherwise peter out in everyday work is described in the guide on implementing strategy.
Conclusion: how does a small business handle too many projects?
The question was a small business runs its projects as a portfolio so that fewer run at the same time and more get finished. The answer follows from Little's law and is supported by the findings: with the same capacity, more parallel projects extend the duration of each one, and beyond a certain point parallel work harms on-time delivery.
In practice this means: write all projects on one page, set a limit for active projects, determine the sequence by cost of delay, and treat stopping as a normal decision. A portfolio is only truly managed when it also contains projects that are deliberately not running.
How to get started
- Write all running and promised projects on one page, with owner and estimated remaining hours.
- Estimate how many hours per week your team actually has for projects.
- Set a limit for active projects and put the rest on waiting.
- Check for each project whether its original occasion still exists.
- Schedule a monthly 30-minute portfolio meeting.
Frequently asked questions
What is a project portfolio?
A project portfolio is the sum of all projects a business pursues or wants to pursue alongside day-to-day operations. Running it means regularly deciding which projects are active, which wait, and which are stopped.
How many projects should a small team work on at once?
Research does not prescribe a fixed number. A longitudinal study shows, however, that some parallel work helps and too much harms. A useful starting point is about two active projects per person, with an overall limit for the business.
Why do projects take longer when many run at once?
According to Little's law, lead time equals work in progress divided by throughput. If more projects run with the same capacity, each one takes longer. On top of that comes the effort of getting back into the work with every switch.
How do I decide which project comes first?
Ask what it costs if a project is finished a week later: lost contribution margin, ongoing extra work, risks, or fixed dates. Projects with high cost of delay and short duration come first.
When should a project be stopped?
When its occasion has disappeared, when it repeatedly misses its milestone, or when a more important project needs the capacity. A stopped project is not a failure but a decision that frees capacity for something more important.
Sources
- Cooper, R. G. (2009): Effective Gating: Make product innovation more productive by using gates with teeth. Marketing Management, 18(2), 12-17.
- Colicev, A.; Hakkarainen, T.; Pedersen, T. (2023): Multi-project work and project performance: Friends or foes? Strategic Management Journal, 44(2), 610-636.
- Zika-Viktorsson, A.; Sundström, P.; Engwall, M. (2006): Project overload: An exploratory study of work and management in multi-project settings. International Journal of Project Management, 24(5), 385-394.
- Little, J. D. C. (1961): A Proof for the Queuing Formula: L = λW. Operations Research, 9(3), 383-387.
- Cooper, R. G.; Edgett, S. J.; Kleinschmidt, E. J. (2002): Optimizing the Stage-Gate Process: What Best-Practice Companies Do, Part II. Research-Technology Management, 45(6), 43-49.
