2026-03-12 08:00:00
AI tools hit a true inflection point in late 2025. Building things got cheaper. AI tools got expensive. And the gap between good management and bad management got a whole lot wider.
Here’s how to think about management in 2026.
Managers must be builders in 2026 for two reasons.
First, you have to learn AI tools. Without a deep understanding of how people build and execute with AI, you will be completely clueless about what to expect from your team and how to guide them.
This cannot be overstated.
A manager’s job is to do things like help people get better at their job and set expectations. Without hands-on fluency with AI tooling, you will be fundamentally unable to do these things well. It would be like trying to manage a software team in 2015 without knowing how to use the internet.
Second, building is often the most efficient path for the team. In the old world, a manager might spend five hours in meetings defending the team’s time against a drive-by request from another team. In 2026, the manager should just spend an hour and build the damn thing.
Managers must build because not building is now the bigger waste of time.
People have incredibly powerful tools at their disposal and companies are spending real money on them. Managers have to raise expectations of output. This will require willful effort, but that effort must be taken.
If you need a cheat sheet for raising expectations:
AI tooling is also putting immense pressure on underperformers. Engineers who can’t review code effectively. PMs and Designers who are bottlenecks. Leaders who can’t adapt to change. 2026 is going to put everyone below the bar underwater, and you need to be ready to step in.
AI tools are moving to consumption-based pricing, which means managers are going to have to think about how much money to invest in each individual. This is a massive paradigm shift. It’s like if you had to decide every month how good of a laptop each person on your team gets, and sometimes people run out of laptop halfway through the month.
Start thinking through these challenges now:
None of this has established best practice yet. The managers who figure it out first will have a meaningful edge.
With more tooling and raw power than ever, your teams need precise goals. The fuzziness that used to get figured out later during slow build cycles will now kill you in the short run.
When your team can build fast, building the wrong thing is the primary risk.
Make sure your teams and individuals have precise goals, or you’ll spend $10k per person per month on AI tools and find out it all added up to a pile of features that didn’t move the business.
Documents are getting longer. Code is getting more verbose. Toolchains are exploding in complexity. Everyone is heads-down with their personal fleet of agents, cranking out work in parallel.
This is great for raw throughput. It is terrible for coherence.
Collaboration in 2026 requires intense, deliberate focus from managers. Week by week, you will need to pull people out of their individual sprints long enough to make sure they’re all running in the same direction. If you don’t force this, your team will ship fast and end up with a product that feels like it was built by five different companies.
A mis-hire in 2026 is catastrophic. The delta between a great engineer with AI tools and a mediocre one with the same tools is not 2x — it’s 100x. One ships compounding value. The other ships compounding slop.
Raise your hiring bar now or spend the rest of the year cleaning up after it.
Management is not dead. In fact, 2026 is the year where management becomes a key differentiator between teams that win and teams that drown in their own output.
This is the year you need to adapt faster, expect more, and start building again.
2026-03-06 08:00:00
There is enormous variability in the frequency with which teams have emergencies.
Some teams have emergencies regularly. We need a new report; someone has to put together a presentation; we need to change plans to incorporate new feedback. Other teams essentially only have emergencies due to exogenous or hard-to-predict factors: us-east-1 went down; we’re getting sued; our CEO got in a car crash. This variability is universal across industries and company sizes; I’ve even heard of non-profits with a culture of constant emergencies. Madness.
As you would expect, better managers have fewer emergencies, and worse managers have more. And at the extremes, it’s common to find that the best managers basically never have preventable emergencies, and the worst managers have teams which are constantly in a state of emergency, to the point that they’re 100% reactive.
There are a few factors that contribute to a very significantly lower rate of emergencies on well-managed teams, and happily they’re easy to copy:
One of the dumbest form of emergencies is simple underestimation of the amount of effort required to get a team’s projects done. Good management can prevent this in a few ways.
First and most importantly, managers need to be deep experts in what their teams actually do, and put in real effort to stay informed. If they’re an engineering manager, they need to be a solid engineer themselves and also stay up-to-date on the state of the technology that their team owns, their biggest challenges, and their capabilities. Bad managers love to black box their teams in the name of delegation: “I shouldn’t need to inspect what’s going on with my team, they should just handle it.” Good managers trust but verify by staying informed before delegating. They know if that report they’re suddenly asking for is easy or hard to produce.
Next, good managers just ask questions. Contrast two identical situations where an ask came from your CEO:
Simply asking the question resolves a surprising amount of emergencies before they get started. If it’s truly impractical to get the report this afternoon, it’s better to just find out upfront rather than pulling all of the fire alarms immediately.
Finally, good managers set expectations and communicate reasoning. “I need you to get me a report on project X but if it’s going to take more than 30min let me know before you do anything. I want to get a readout to a new sales prospect but it’s not essential.” Especially when seniority gaps are large, a little expectation-setting goes a long way.
A huge amount of artificial emergencies stem from managers who don’t understand what is actually important for their teams.
If you don’t know what matters for your team, the latest thing that just popped into your brain often feels critical. If you didn’t have strong conviction about whatever project was in-flight, any new good idea always seems like it could be worthwhile. Bad managers can’t stay on target, and because they can’t stay on target they never say no to new work, creating constant emergencies for their teams.
A simple technique that works here is to always make sure that you have very strong conviction that what your team is working on matters, and actively force yourself to make sure you always know why their roadmap is important. This is critical to give you the courage to push back on the worst sorts of emergencies: executive requests like “[CEO] needs a full proposal on how we’d tackle this idea today.” If your team is really working on business critical needs, responses like “we can get you that after [critical work] is finished” or “how about we get you 5 bullets and a summary, because we’re finishing [work we all agreed matters]” actually become available arrows in your quiver.
This is a more subtle concept, but one of the most important.
Good managers have a strong mental model for how their team operates and the role it plays within the company, and have a strong understanding of the state of the business and industry. This empowers them to more accurately forecast what will be needed of their teams in the future.
For example, let’s say that I’m running a Product Design team. I know that my team currently outputs pictures of what we’re going to build. But I also know that AI is shaking an already dynamic landscape: for example, there’s more AI prototyping including from non-designers, a rise in vibe coding, and all of this amid rising UX standards for enterprise software.
With this simple mental model, I can make a number of moves that will prevent emergencies down the line:
This concept always reminds me of college-level math. If you take college-level electromagnetism, differential equations, or probability and statistics, you’ll often find that there are two ways to pass your tests:
To really prevent emergencies, you need to be in the second state when it comes to your team.
And finally… good managers simply care more about their teams’ well-being. And if you care about your team, you’re much less likely to throw a hand grenade into the room and walk out the door while your team is still inside.
Despite the many bitter comments that you hear on internet forums and bars at 5pm, in my experience many managers really do care deeply about their teams. I’ve had so many conversations with managers who are working their butts off to prevent emergencies from impacting their teams.
The simple fact is that in many cases emergencies are a choice – specifically, they come from making the choice to satiate some desire by sacrificing your own team’s time. You can prevent this: you just need to care more about your team’s long-term productivity than a short-term boost to your career or mental health.
One of the best parts of following this advice to prevent emergencies is that it really will make your team happier and more productive, and it will do so almost immediately. Nobody wants to live in a world of constant emergencies, but it’s all too common across every industry. Great people want to do great work, and a culture of emergencies is anathema to the focus that great work requires. If you can find a way to keep your team operating from a level-headed playing field, you’ll have unlocked one of the best talent retention mechanisms that exists.
2026-01-15 20:59:22
The most important teams in a B2B software company are engineering and sales. Full stop, no exceptions, no further questions. You’re either building the product or selling the product, and everything else is secondary at most.
The intuition is simple:
This truth is so clear that it can be used as a litmus test for whether you have any idea about what the hell you’re doing in SaaS. If you find someone who doesn’t intuitively understand why these two teams make the world go ‘round, then you clearly haven’t been in the business for real.
The prominence of engineering and sales may seem like a trivial observation, but it actually comes with a number of very important operational takeaways.
A lot of people struggle with this, so I’ll just say it plainly – if you’re not in sales or engineering, you need to know your place. As a product manager myself, I’ll describe this in terms of how I want my teams to (ideally) work with our engineering counterparts:
This doesn’t mean that you should just do whatever engineering (or sales) wants. If sales or engineering are incompetent you need to deal with it.
But you need to assume that all else being equal, their immediate priorities – writing and launching code; closing deals; renewing customers – are at or among the very highest priorities for the business as a whole. For example:
If you’re actually leading a sales or engineering function, there’s a lot of pressure on you to perform.
As a company matures and grows your job will evolve, becoming more complex, with higher standards. You need to constantly improve at your job as well. This might be an academic exercise if you are, say, an HR team. But if you’re one of the two most important teams at a company you can literally be solely responsible for killing an otherwise thriving business if you don’t evolve fast enough.
The criticality of sales and engineering leads to a paradoxical situation:
Since sales and engineering are the pace-setters of an organization, there is no such thing as “good enough.” You will always be compared to the best team that the company believes it could theoretically build. Nobody wakes up in a cold sweat wondering how they could have a 1.5x better HR team; every CEO in the world would move heaven and earth for a 1.5x better sales or engineering organization.
As a result, it’s on you to constantly self-improve, because nobody is generally going to mess with their money-making teams… at least until they decide that you’re not cutting it, at which point you’ll be replaced.
If you are in an adjacent field to sales or engineering, you should strongly consider how you can build more experience in these critical fields.
In many situations, the priorities of your sales and engineering teams are going to take precedence over your own. The biggest deal of the quarter is more important than your marketing deck review; launching the rebuild of the product is more important than your PRD. This is fine and actually often the sign of a healthy business. As a result, learning about what makes both sales and engineering tick and why will help you do your job better. You cannot swim against this tide, so you need to learn how to swim with it.
The main way that I see leaders getting into bad situations is by assuming that sales and engineering are simpler than they really are. It’s fine for me to be a non-technical PM, I know the customers. It’s fine for me to do marketing without being a sales expert, I know how to build an audience. No. You need to understand sales and engineering as well as you possibly can – otherwise, you’re going to be like a quarterback who’s watched hours of Sportscenter but has never thrown a football.
The fact that sales and engineering are the most important teams means that sales and engineering failures are the largest risks to your business. If you’re at a company where these teams are weak, you should strongly consider leaving. You are highly unlikely to save the ship.
Once you see this truth, many other fact patterns start to make more sense.
2025-12-30 20:59:22
It was an interesting year in the Stay SaaSy universe. We grew our community significantly on Substack and X and via email. We met some amazing people through this blog and we feel increasingly plugged into a truly special group of builders, managers, and leaders.
AI was the story this year and it often left little air in the room for other topics. At different parts of this year: management was declared dead, SaaS was declared dead, blogging was declared dead.
And yet here we are. All of these things are not dead, just changing.
We look forward to another exciting year of writing, tweeting, and engaging with the great Stay SaaSy community. We’re confident that blogging, management, and SaaS will still be here this time next year. But we believe they’ll likely be meaningfully different, and that things are only speeding up from here.
We’re excited to go on the next part of this ride with you.
We did several episodes of a podcast this year!
Podcasting was a new format for us and certainly something we haven’t learned to be consistent at. We also used voice modulators to stay anonymous and we got feedback that for some people it was a bit too much.
If you enjoyed the podcast or have feedback - let us know! We’re going to try it a bit more this year, ideally with some software that gives us cooler voices.
2025-12-21 14:59:22
This is a guide on how individual contributors (ICs) can achieve outsized impact within a software organization.
Individual contributors have the special property that they can get real work done. Managers are often constrained because their leverage is through people, which often means slower and steadier change.
ICs can bypass this. You can work a weekend to deliver a working prototype. You can spend nights finding an insane performance optimization. You can Proof-of-Concept a feature scheduled for next year.
So much of a company is bound by risk assessment, expected timing, and prioritization minutiae. Finding a genuine breakthrough cuts through the bureaucracy and fundamentally shifts the timeline.
If this work is so important, why isn’t it prioritized? It’s not prioritized because if you make it part of the day-to-day, it gets bogged down and dragged out like everything else, often resulting in sunk time with nothing to show for it.
By taking a calculated risk on yourself—the risk that you might fail—you can deliver a breakthrough that dramatically increases your value and the company’s. Occasionally, ICs should go all-in, work intensely, and find a breakthrough.
If you want to be a leader, you have to act like one.
That begins with being optimistic, not starting or encouraging big commiseration sessions with more junior people, not trying to set up an us-vs-them dynamic against “management” or treat other company structures like the bad guy.
Many senior ICs think it’s their job to be the union rep for ICs. The only thing that actually accomplishes is undermining your own credibility. It doesn’t deliver the value that would make people’s bonuses go up, and it doesn’t change hearts and minds.
But you also shouldn’t be a shill for management.
Like it or not, your job is to be an unaffiliated, unbiased leader, and doing what’s right in individual situations.
Another critical part of being an IC leader is owning things.
One of the specific virtues of the manager position is that it provides radical clarity on ownership - you own the team. With that ownership comes risk, reward, and accountability. The team wins, you win. The team fails, you fail.
Too often ICs get slotted into being 1 of n engineers on a team, becoming a non-unique, non-owner of outcomes. In fact, agile development processes often specifically encourage interchangeable resourcing.
But ICs should avoid taking this too far. Ownership and accountability are critical forcing functions for growth. You should ask yourself - what specifically could fail that would cause me to get less rewards? What specifically could do well that I should uniquely get credit for?
If you don’t have a good answer, you don’t have enough specific, named ownership. While ownership can come in many forms (services, features..etc), the core unit of ownership is a goal. If you don’t uniquely own any specific goal, you are shortchanging yourself.
Send regular updates.
Bi-weekly is usually a good cadence.
Send your boss a doc that says what you’re doing, what you need help with, and cc their boss.
This serves multiple functions
The benefits of this ritual are truly massive.
You’ll be shocked at how much better you get at self-management when forced to do it.
You’ll be shocked at how often your boss/boss’s boss can solve your problems faster when presented this way, and how grateful they will be for the clear updates.
PS if you use AI to write these updates you’ve destroyed their purpose entirely and are subtracting value.
Talk to people more senior than you in the org chart.
Seriously - book time with senior people, or ask them for 30 minutes to discuss what they think is valuable and worth doing. Senior leaders absolutely love this, because it lets them get more important work done through their team, and it often doubles as a therapy session.
Talk to five people from different parts of the business.
Then go solve one of those problems.
This is an extremely underrated activity because:
2025-12-15 14:59:22
Compensation is difficult. Even more than that, it is sensitive, business critical, and often has almost nothing to do with the rest of your job as a leader. Here are some rules for making compensation decisions effectively and efficiently.
It might seem obvious… but your foremost goal when running a compensation cycle is to maximize the talent and enthusiasm of your team. Full stop.
Creating a talented team means that you can get great people to accept your hiring offers. Creating an enthusiastic team means that these great people are motivated to work for you, engaged in their work, and (in the upside case), will work much harder than otherwise due to their compensation structure. And you need to fit this into a budget that allows your business to function.
If you spend enough time as a manager, you start to realize that:
To hit that final metric, you roughly want to target paying at least the amount where team members can’t easily find another job that will pay them more. (And on top of that, of course you need to try to make your work environment as otherwise appealing as possible)
The intuition here is that many jobs basically kind of suck, so if you don’t mind your current job, switching jobs is risky – and most people know that. What you want in your team’s head is some version of “perhaps I could theoretically get paid more elsewhere, but it would probably take a lot of effort and maybe the new job would suck in other ways.” This will retain most otherwise happy employees, which is ultimately what you deserve and the best you can hope for in any case.
Of course, this is easier said than done! You need to track the market, both via surveys and seeing what new candidate offers are accepted / heavily negotiated / rejected. You need to watch for attrition on your team, observing where departing folks go and why. And when you get new signal you need to react – don’t allow your compensation to get out of line with your target market percentile. But all of that is tractable as long as you put in the work.
But while you can’t make an unhappy team member happy with compensation alone, the converse is not true: how you compensate can absolutely make an otherwise happy team member unhappy.
What makes people absolutely furious about compensation is finding out that they’re paid the same, or less, than someone they think is less deserving. There are some people who will literally quit a company on the same day if this happens.
The best way to prevent a sense of injustice is simply to have clear compensation bands, and essentially never allow employees’ compensation to drift outside of them. Compensation bands are excellent at preventing some of the most common sources of unfairness – managers using their discretion to play favorites, and new hires making more than existing team members at the same or higher seniority (salary compression). The latter occurs regularly when market compensation shifts, hiring managers are captivated by a shiny new hire, and they allow their team to fall behind the market. Rigorous compensation bands prevent this from occurring (if you move your bands to accommodate a new hire, you need to move your existing team’s comp as well).
The most common cause of unjust compensation is simply laziness. Keeping track of everyone’s compensation to prevent unfairness or setting up a system that enforces standards takes effort, and many teams fail to do it. This is a huge mistake, as perceptions of unjust compensation are one of the ugliest and most acute sources of employee dissatisfaction. Additionally…
Many people talk about their compensation with others. As a result, you should be willing to explain exactly why any team member’s compensation is set where it is, in front of the whole company. You won’t have to do this often, but it’s the standard to hold yourself to. The reasons that comp is uneven between employees can actually even be random or somewhat weak, but there needs to be some justification. Some common, defensible examples:
Some of these reasons aren’t necessarily emotionally satisfying, but they’re all completely intelligible without any whiff of injustice. And injustice is the critical factor that drives people crazy.
Since the goal of compensation is creating an enthusiastic team, there are additional ways that you need to think about pay in order to maximize return on your dollars.
The most immediate levers are performance-based compensation such as bonuses and sales commissions. The main heuristic to keep in mind for any performance-based comp is that:
But equity is a much more powerful lever – specifically, a lot of equity.
If someone has 10% of their annual compensation in company equity, they will act like an “owner” the way that you act like an owner of your rental car. They’ll take care of things, generally be responsible, and won’t badmouth your company (much).
But if someone feels like their equity will change their life, most will act like a completely different kind of owner. They’ll start to treat your company like their child – they’ll stay up all night caring for it if there are issues, they’ll think about its well-being all the time. For your top performers it’s really worth getting them to that point via equity compensation if you can. If you want a quick heuristic, the most common amount of money that most people consider to be life-changing is roughly the price of a good-condition 3 bedroom home in their nearest metro area.
In summary:
