← Back to all articles
Article 07Tips & tricks

Urgency: why it is unfortunately necessary to get anything done

… and why you need a deadline for everything and why execution requires training

… and why you need a deadline for everything and why execution requires training

I get a lot of hate from all kinds of co-workers, from leadership to individual contributors, when I set deadlines. They always want to know why. Some deadlines are obvious, when there is an important external event like a marketing event or big conference at which we’re going to launch a product, for others though I don’t have an explanation.
The explanation is that without a deadline… nothing happens. And without a tight deadline… nothing happens in reasonable timeframes.

The only takeaway you need to learn from this article is simply that.
Nothing happens without a tight deadline

No deadlines => no or bad products

I have so many stories of why the learning of today is just simply true. I actually can’t think of a single example of when an amazing product was launched without a deadline. People argue that tight deadlines stress the teams and result in poor products. I unfortunately have to agree with the stressed team part; we are indeed in the century of stress. I do have to push back on the second. Without tight deadlines one of two things happen:

  1. No products get shipped because you miss the boat
  2. An excellent product gets shipped… too late. Late products are always poor products. And they end up not being successful

You combine these, and one of these happen to the team:

  1. If the product didn’t ship, they get frustrated since they need to throw away their work
  2. If the product did ship, they get frustrated because it fails and they won’t have anything to add to their resume or promo package
  3. In startups, the startup fails and the team ends up without a job
  4. In high bar big companies, the product fails and the team gets fired

All of these result in a lot more stress compared to the stress of having to hit a tight deadline. So tight deadlines are actually good for the team too, they just don’t know it.

Stories of lackluster execution

Let’s start with some iconic stories that I’ve lived through and that have shaped the thinking in this article.

1 - Infinite time for the best product

This is the only story that I didn't live in person though but I got acquainted with as soon as I joined NVIDIA. I saw the tail end of it, witnessing the reaction of the team.
NVIDIA at the time was working on tablets. Not many people will remember that, though at that time (2016) it was still a tablet hype type of market.
After a few versions, it was time to iterate and deliver the next version. The direction of the VP was clear: this has to be the best tablet ever for NVIDIA, that learns from all the previous versions and absolutely aces it.
In order to achieve that, the team had as much time as they wanted. No restraints. All that mattered was that the product had to be amazing and, when released, absolutely win the market.

After 2 years (give or take, I don’t remember exactly), the product was cancelled and never shipped. The team was devastated, though leadership did a good job to spin the story as yet another learning opportunity.

What happened?

Time happened. As time goes by, products which were great ideas become stale. The market evolves. The hype passes. This is the common thread you'll see in the other stories.
In this particular case, what happened was that tablets were not as popular after two years from the project start. The market which had been nascent was now completely saturated. More and more options were available and making something that was a standout became a lot harder. Prices went down, so suddenly, even a higher end product would have to fight for its price tag.
The wave had passed. It didn't matter if the tablet was going to be the best tablet ever. Nobody cared about having the best tablet ever at that time. In particular, nobody wanted the best tablet ever that was also expensive.

It was hard for the team to see it before the two years because the wave doesn't pass all at once. It moves slowly every day. Every competitor release, every feature that a competitor invents, moves the goal post a little further. What was a good idea at the start of a project became a less good idea in time and required the team to throw away that part of the work. Things that had already been done had to be redone to keep up with the technology advances. Every moment it looks like you are nearly there, but the target continues moving, especially if you want to make the best product. It slowly drains all the energies and resources assigned to the project. Even when the lease on life of the project is extended, usually by an executive decree or by assigning new people to the project, the goal post continues moving.

Especially in industries such as semiconductors, consumer hardware and software, the speed at which the whole industry and customer expectation move is very fast. Every day there is something new that could render the product that you’re working on useless. A lot of AI startups learnt that in 2025. OpenAI wrappers became irrelevant as soon as the models got better and were able to do the same tasks that the wrappers were previously needed to do.

Infinite timelines always end up in team delusion and a lot of wasted investments.

2 - you miss the launch, you miss the wave

I distinctly remember this scene. I was walking with a product director in the hallways of the new NVIDIA building. We were talking about a developer kit we were working on that was a new foray of NVIDIA into the healthcare space. The director was a veteran that had done many things at NVIDIA, though he was new to the team (the same as the one of the tablet) which had pivoted from consumer electronics to B2B developer kits. I specifically remember saying “if we don’t have the dev kit ready for the XYZ event, we will never ship it”. It was not received well. The director reminded me that we were doing something new and that we had to get it right.

At the dawn of event XYZ (which was an internal event, where teams got to show what they were betting on; the name is made up BTW, I don’t remember the real name), the developer kit wasn’t fully ready. It was nearly ready, but not totally. It was brought on stage and shown, though it wasn’t ready to be demoed in action. The VP directed us to perfect it, at the time it wasn’t a huge deal that we were not able to have it fully ready. As a result though, sales teams which could have gone out to pitch it to customers were not able to.
The whole point of a developer kit is to showcase the chips of the company and wrap them in IOs (input / output interfaces) that enable companies to try out the chip and understand how they will use its capabilities. That part was mostly done and we were solving form factor and cooling issues, which are important but can always be solved with functional but not pretty solutions. We were going for the cool solutions though.
A few months after the events though, the new chips came out. There was no point in trying to sell the old chips. All the focus was on the new ones.
We shelved the project and started to work on the developer kit for the new chip that required different form factor and capabilities. We learnt a lot from the project, at the same time though it was a missed chance. If we had made the internal XYZ event, we could have shipped it and put the dev kit in the hands of sales and in front of customers. We would have learnt a lot more from that.

3 - you can’t ride the hype wave if the wave is faster than you

For most of my initial time at Mythical, we were squarely riding on the huge NFT wave. As we were working on a blockchain platform for gaming (and many other companies were too) we already had a game out and we were doing sales of digital assets on it weekly (some that got to millions of dollars for single item releases), putting us ahead of competition.
However, we were missing a token, and an iconic collection of collectible NFTs. All big companies in NFT space had their banner NFT collection. Bored apes, Crypto Punks, Invisible friends… If you wanted to be known in blockchain space you needed to have your iconic limited edition NFT collection.
We also had something that others did not have. A partnership with the NFL. Not any random sports league. The NFL, the same organization that sets up the superbowl every year, one of the most important sporting events in the world. That was unique IP that no one else in the space had access to.
We were going to develop a mobile football game (which was going to become NFL Rivals) though it was not going to be ready for the superbowl of that year. So we were thinking of launching a collection of iconic NFT helmets, which was going to become our iconic collection of NFTs, with insane perks inside the game.
We wanted to make it absolutely perfect. We wanted the helmets to look amazing from a digital art perspective, we wanted the sale experience to go smoothly with an advanced queueing system, we wanted people to be able to buy with both USD and cryptocurrency, we wanted to nail the right perks inside the game. There was so much to do.
We started working on it more than 6 months before the superbowl, though we barely managed to launch them before it. We had time, so we used it to make them as perfect as we could.
There was only one problem. By the time we launched it, iconic NFT collectible collections were not trendy anymore. Sure, Board Ape was still popular, though the market didn’t have room for yet another collection. All the ones that had launched recently, including one from a competitor of ours focused on an arena style game, had gone terribly. The market simply had moved on from that model.

Ours also didn’t go that well. Rather, the first two teams went well, but we had a lot of helmets, since we wanted to cover all the teams. By the time we went through it all, it was clear that the demand simply was not there.

What was worse though, is that the focus on the helmets had delayed the release of the game, which was the main product of the company. Focusing on this collectible NFT release had consumed all of the company’s energy, resulting in a delay of a much more important product, the game.
I don’t know if focusing on the game more than the helmets would have made a big difference. By the time the game was released, the NFT hype had completely passed. We pivoted making the game essentially blockchain-transparent, such that users had no idea that the cards that they had in the game were on blockchain. If we had launched the game sooner, we might have still rode some of the wave, especially since there hadn’t been a “real” big hit out of all the blockchain based games, though the demand was there. Rivals had all the ingredients to be that one, if only it had shipped earlier. With the branding of the NFL it could have surpassed Axie Infinity, which was the “game to beat” in crypto, though still fairly niche in the broader gaming world.

If you are building in a hype market (AI in 2026?) then speed of delivery is a lot more important than “getting the launch right”. Usually getting to market first does get you a first mover advantage. If the wave moves on… then you’re better of scrapping the product entirely and moving on to the next thing.

So how do tight deadlines help with execution?

The universal truth is that tasks expand to take all the time that is allotted to them. Hence giving “buffers” or “optional time that the team may use only if necessary” always ends up in the time actually being used.

**Tasks expand to take up, at a minimum, all the time allotted to them**

The reason is that people usually care about their craft and want to do the best possible, so if there is more time, they will use it to try and perfect the product. There is always one more round of QA that can be done, one more attempt at optimizing the model, one extra try at perfecting the routing, one more pass on the designs etc. If time permits, the engineers will take that chance. This is not bad. It is the engineer’s (or whoever else’s) job to do the best possible. It is the product manager’s job to only give them enough time to do a good enough job.

**Perfection is the enemy of good**  
**Premature optimization is the root of all evil**

Both of those quotes are not mine, but are very true and are the reason why all tasks expand to take up their assigned time (at the minimum). When the task is completed in a “good enough” way, people start optimizing and perfecting the output, which usually doesn’t really move the needle for the success of the product.
Tight deadlines are the antidote to that.

Sandbagging

You probably haven’t realized it, but you have most likely been both a victim and a culprit of sandbagging.
Sandbagging is the practice that involves adding buffers to deadlines, in order to CYA (cover your backside) and look like you have hit a deadline.
There is no greater evil than sandbagging.
Sandbagging originates from the human aversion of being wrong. Rather than not hitting a deadline, humans prefer to add a buffer to then be able to hit it in the future. As a result, tasks take longer (and usually still go beyond the extended target) than what they should, just because people want to CYA.
This is terrible for company execution.

Execution speed before and after the deadline

**Execution speed increases as the deadline approaches**  
**Execution speed increases dramatically after a deadline has been missed (for internal deadlines)**

This is where the sense of urgency comes in. As a team approaches the deadline, and especially after a deadline has passed and the task is late, execution speed increases. This is due to the same aversion to missing deadlines that causes sandbagging. Because people dread being late or wrong, they will put in more effort to complete the task as the deadline approaches, and will put in a lot more effort when the deadline has passed, since they don’t want to be indicated as the people that are causing delays to the product. The caveat to this last case is only if the deadline is internal. For a marketing event for example, if the team misses it they are aware that the next event will be a lot out, so you will lose any speed increase advantage.

Let’s see now why you really have to fight sandbagging with tight deadlines.

How to fight sandbagging and drive execution with tight deadlines

Consider a task that would take a good team 6 weeks to complete. Let’s see these two examples:

1 - tight deadline
The PM requests the task to be done in 4 weeks. The team pushes back and mentions that it is going to take 6 weeks. The PM doesn’t budge and still sets the deadline in 4 weeks. Based on what I’ve seen, teams would usually hit the 6 weeks, since they start to really focus on getting things done at the 4th week, descope any unnecessary features and bugs come up sooner, so then they get done in 6 weeks.

Total time needed: 6 weeks
Total lateness perceived by everyone else outside of the team: 2 weeks

2 - reasonable deadline
The PM requests the task to be done in 6 weeks. The team agrees. After 6 weeks the task is nearly complete, though a last minute bug comes up and requires changes and testing. The tasks gets done in 7 weeks.

Total time needed: 7 weeks
Total lateness perceived by everyone else outside of the team: 1 weeks

3 - sandbagged deadline
The PM wants to be sure to hit whatever date they promise, so they set the internal target for the team at 6 weeks and communicate an external target at 8 weeks. Every time I’ve seen this, the team ends up taking 8 weeks in the best of cases. Usually you get to the same point as case (2) where last minute bugs show up and delay you by 1 week. Let’s ignore that and take the best case where the team actually completes the work in 8 weeks.

Total time needed: 8 weeks
Total lateness perceived by everyone else outside of the team: 0 weeks

You might not believe me though I can attest that the above is how usually things go. Ask other PMs or project managers.

TAKE A LOOK AT THOSE TOTALS. What happens is that people focus on “Total lateness perceived by everyone else outside of the team” instead of focusing on “Total time needed”.

The more you sandbag, the less time “Total lateness perceived by everyone else outside of the team” will happen whereas the more “Total time needed” will happen.

When you can’t use aggressive deadlines that the team will most likely miss

The one non-negotiable on tight deadlines is when there are external deadlines that are not set by the company (or on which the team doesn’t have a say) and that require cross functional coordination and a “work back” schedule. Examples include conferences, big launch events or important deadlines for partners.
If for example you are at Google and you need to deliver a product for Google IO, you can't be late, not even one day. If your product is scheduled to be shown at CES, you have to have it ready.

What will happen in the case of cross functional products is that there is going to be some kind of program management initiative asking each team on when they will have ready intermediate milestones to make sure to hit the date. If you provide an aggressive schedule, for example accounting for six weeks even though the team is more likely to take seven, then you will deliver one week late and you will delay the rest of the company. That is absolutely not OK. In those cases. It does make sense to aim for a realistic estimate.

What I have found to be useful is to have separate deadlines, one to communicate outside of a team and one to communicate internally to a team. You will lose some trust with the team in case they find out that they had “more time”, though it is the only way to make sure you can provide the deliverable to the rest of the cross functional project without having to excessively pad the schedule.

Indirectly, this showcases how big cross functional projects take a lot longer than expected. Each team wants to make sure that they will hit their deadline and hence sandbag a little bit, resulting in an overly padded end to end schedule. The sad thing with cross functional projects is that, even if a team delivers early, usually the team next in the critical path can't actually start early because they haven't planned for it. And as a result, any padding becomes reality for the end schedule.

Conclusions and what to do if you’re behind on a deadline

Inevitably, sometime in your life, you will be getting up to a deadline and looking like your team is going to deliver late. What to do in those cases relies on the famous tradeoff triangle between cost, quality and time. There will be a follow up article describing that in more detail. In general, if you’re coming up to a deadline and are behind, you can’t use the cost lever, since it usually has a slow response, so you have to act on the quality lever. So you can either rush things and decrease quality overall, or, what is usually the best option, de-scope some features focusing on what your product really needs.

Too often product managers can't bring themselves to de-scope features once the product execution has started, however, that is often the best option and also the one the team feels better with.

Hopefully you’re now empowered and confident in setting tight deadlines for everything! Happy shipping to everyone and remember: ship now, optimize later. No point in optimizing if the ship is going to arrive late.