Watch this video on YouTube (opens in a new tab)

Here's a pattern I've seen so many times. An organization is struggling with delivery. Things feel slow. Stakeholders are frustrated. And when you look underneath, the priorities keep changing. Not because anybody's being malicious. Not because leadership is incompetent. It's just that there isn't enough agreed-upon runway — not enough stable, committed backlog — to let the teams focus.

And when that happens, something really ugly starts to fester.

Nobody Decided to Multitask

When priorities shift, the old work doesn't stop. It just gets deprioritized, which basically means it sits there half-finished while the team pivots to the new thing. And now the team is working on two things. Then three. Then four. Nobody made a conscious decision to multitask. It just happened because nothing ever fully stopped before the next thing started.

This is the part that makes it so sneaky. It doesn't feel like a decision. It feels like the natural consequence of doing your job. But it comes with a tax. Actually, three taxes. And the numbers are brutal.

Tax #1: The Cost Tax

Gerald Weinberg did a study back in 1992 on software developers juggling multiple tasks simultaneously. At one task, you're 100% productive. At two, you're at 40% on each with 20% waste. By the time you're juggling five things at once — which, let's be honest, is pretty common — you're at 5% productivity on each task and 75% of your time is pure waste from context switching.

Let me put that in real terms. Five simultaneous tasks means Monday, Tuesday, Wednesday, and most of Thursday are gone. You get some of Thursday and Friday to actually be productive. That's what your organization is paying for when it lets teams juggle.

Now here's the thing that should make every executive sit up straight. If you're in that five-task hole and you can manage to drop down to three tasks — which is still a lot of juggling — that's roughly a 35% improvement in productivity. Not from hiring anyone. Not from buying anything. Just from doing fewer things at once. That's your "do more with less" miracle, and it's sitting right there in front of you.

Tax #2: The Speed Tax

The cost tax is painful, but the speed tax might be even more damaging because it's the one your customers actually feel.

Say you've got four items to work on, and each one takes exactly one day of effort. If you work on them one at a time, each one takes one day from start to finish. Cycle time equals effort. Nice and clean.

But if you work on all four simultaneously, splitting your time across them, each one still takes one day of your effort — but now each one waits three days while you're working on the others. Your cycle time is four days for one day of work.

I know what you're going to say: "But the cycle time on each item is one day. I did one day of work on it." Technically true. One day of active work. But cycle time isn't active time. It's elapsed time. It includes the delay. And your customer doesn't care how much active time you spent. They care about when they get their thing.

Higher WIP doesn't get more done faster. It makes everything take longer. If you want to see this math worked out step by step on a kanban board, I did a whole video on it: The Math Behind the WIP Trap.

Tax #3: The Predictability Tax

This is the one that really sets up the doom loop. When your teams are juggling a bunch of stuff and cycle times are all over the place, you can't forecast when anything is going to land.

Think about predicting delivery on a backlog. Say you're 75% confident on each individual item. Not bad odds. But probabilities multiply. By five items, you're at 24%. By ten items, you're at 5.63%. Even though you were 75% confident on each one individually, the combined probability of hitting your date across the whole backlog is basically zero.

So when somebody asks "when will I get my stuff?" and you give them a date based on estimates, that date is almost certainly wrong. Not because your individual estimates are bad, but because chained probabilities are brutal.

The Doom Loop

Now here's where it all comes together. These three taxes don't just coexist. They feed each other. And they create a self-reinforcing cycle that I call the priority doom loop.

It starts with not enough agreed-upon runway. Priorities thrash. Teams end up juggling multiple things. The three taxes kick in: you're paying more, everything takes longer, and you can't predict when anything will land. Delivery slows down. Stakeholders lose trust because they don't know when they're going to get their thing.

And here's the doom loop: when stakeholders lose trust, they push harder to get their priority to the front of the line. And that's more thrashing. Which causes more multitasking. Which makes everything even slower.

The very behavior that feels like the rational response to slow delivery is actually the cause of slow delivery. And it gets worse every time around the loop.

I wrote about a version of this in Azure DevOps or GitHub? Make Sure You're Asking the Right Question — teams that are frustrated with delivery often blame their tooling, but the problems that actually slow them down are almost always upstream. Poor prioritization, too much work in progress, and the classic "everything is priority one." Those problems follow you to any platform.

So the next time somebody says "just bump my thing to the top of the list," know that what they're actually asking for — probably without realizing it — is to make the whole system slower. For everyone. Including themselves.

Breaking the Loop

So how do you actually break the doom loop? It's not complicated, but it does require some discipline.

You don't need a year of committed backlog. You don't even need a quarter. You just need enough agreed-upon runway that your teams can focus for a few weeks without the rug getting pulled out from under them. Two to four weeks of stable, agreed-upon priorities is usually enough to break the cycle.

This means treating your backlog as a negotiating tool, not a to-do list. When somebody wants to add something to the top, the conversation isn't yes or no. It's: "sure, help me figure out what we bump." That's a fundamentally different conversation, and it's the one that protects your teams from thrashing.

And the single most powerful thing you can do? Just do fewer things at once. Fewer things in progress means shorter cycle times, better predictability, less waste, and stakeholders who actually trust your delivery dates. Not because you got faster, but because you stopped making everything slower.

-Ben

If your organization is stuck in the doom loop — priorities thrashing, teams juggling, delivery grinding — that's the kind of problem I help teams diagnose and fix. Let's talk.