# Eliminate Bullshit Management and Validate Assumptions

**Podcast:** Product Momentum Podcast
**Published:** 2026-07-15

## Transcript

Instead of talking about assumptions, I want to talk about something you've been writing about a little bit recently, which is bullshit management.
And so enlighten our crowd as to what bullshit management is and what's going on.
It's the enemy.
All right.
For those of you who don't know, bullshit management is an art.
The art of doing things that create no value, but drain your energy.
And there's an equation behind it.
The more bullshit you do, the last value you create.
And then people are wondering, am I doing bullshit or not?
How do you know that?
If you can't say how what you do create value, you have the answer.
We got some advice from Rich Mironoff, who's been on the podcast twice, just released his book Money Stories.
One of the things that Rich said that he started early on in his career is that if a meeting invite was sent to him with no agenda or no stated outcomes, he just said no.
That's what he would do also.
The very first thing, you need to have time for focus.
So for example, I block my morning.
Nobody can reach out to me in the morning because it's a time I'm going to do deep work.
And then meetings, they need to have a few things.
Like what is the objective of the meeting?
What is the context?
And then what is the agenda?
And the agenda for me is not saying like the bullet points, it's how much time I'm going to invest on each topic.
And then I need to know what is expected from me in that meeting.
If I can't answer that, I immediately reject.
Yeah, assumption checks, especially with AI, right?
It's really easy to come up with, you know, 50 assumptions.
How do we figure out what assumptions we really need to care about?
Because if we go chase all of them, we'll never get to the doing part.
We need to categorize them.
So first, we need to understand the following.
Ask the first question.
If this assumption is proven false, is my idea still valid?
And if the answer is yes, well, Then you can literally ignore this assumption because your idea will still survive.
But then you say, if this is proven false, my idea is worthless.
Start looking at this differently.
Then you say, how confident am I that this assumption is true?
That is your data.
We are here with David Pereira.
He is a coach, speaker, and author with 17 years of experience in helping companies achieve more by doing less.
I'm excited to hear a little bit about that, right?
But it sounds like you do that by helping them cut through the red tape and get to what matters the most, which is creating value.
So thank you so much for being with us today.
Thanks for getting me here.
Yeah, right.
So I got to be a part of your workshop yesterday, so I'm going to apply my training, right?
So I'm going to run an experiment, a real quick experiment, right?
I'm testing the assumption that you don't want to talk about assumptions, okay?
And my expected result is that you say yes, right?
So instead of talking about assumptions, I want to talk about something you've been writing about a little bit recently, which is bullshit management.
And so enlighten our crowd as to what bullshit management is and what's going on.
It's the enemy.
All right.
For those of you who don't know, bullshit management is an art.
The art of doing things that create no value, but drain your energy.
And there's an equation behind it.
The more bullshit you do, the less value you create.
And then people are wondering, am I doing bullshit or not?
How do you know that?
If you can't say how what you do create value, you have the answer.
Give us some examples of what you found when working with companies where they didn't realize that they were executing bullshit management.
Extensive backlog management of meeting marathons, opinion-driven decision-making, all of these kind of things, you realize you are doing them just because you are doing and don't understand why.
Sometimes I ask companies, like, how many hours a week each person spends on average in meetings?
Then I get crazy answers, like 30 plus hours.
And I say, who is working?
Because, like, what happens is that these companies, they start talking about the world instead of doing the work because they love talking.
I think, yeah, meetings is probably a sore subject for all of us.
What have you found in your time traveling from, you know, company to company as techniques to?
make meetings more effective, reduce the ones that don't need to be had.
So there's one thing that I like accepting.
Your real job is to simplify what others complicate.
So many times we have meanings because it's the default way of collaborating.
So we need to step back and say, what on earth am I trying to achieve?
Is it a decision?
Is it something that I want alignment?
And then I say, what is the best way of doing that?
Sometimes it's just a message on Slack or it's an email saying what you want to do.
And if you want to make a decision, instead of scheduling a meeting for one hour, just do the preparation.
And then you schedule a talk to make the decision for 15 minutes.
Because there's this law that the work will expand to the time you allow.
If you put one hour, it's going to last one hour.
If you put 15 minutes, people are forced to think about what makes sense and then make the decision.
Yeah.
I've also, we got some advice from Rich Mironoff, who's been on the podcast twice, just released his book, Money Stories.
One of the things that Rich said that he started early on in his career is that if a meeting invite was sent to him with no agenda or no stated outcomes, he just said no.
That's what people do also.
So the very first thing, you need to have time for focus.
So for example, I block my morning.
Nobody can reach out to me in the morning because it's a time I'm going to do deep work.
And then meetings, they need to have a few things.
Like what is the objective of the meeting?
What is the context?
And then what is the agenda?
And the agenda for me is not saying like the bullet points.
It's how much time I'm going to invest on each topic.
And then I need to know what is expected from me in that meeting.
If I can't answer that, I immediately reject.
And when I was a CEO of a company at Hamburg, I was a little bit annoyed with that.
I started poking around and seeing meetings.
Meetings without this, I would just delete.
I delete from the calendar and people say, I had an exchange there, catch up.
I said, no.
You had a time waster.
First, you structure, then you get people to talk about it.
But don't build the conversation without the topic.
There's no need for that.
All right.
So everybody, if you get a meeting invite with no agenda or outcome stated, you get to immediately cancel it.
And then if somebody asks why, you say, David, David told you that you had to.
Or Sean, but not me or Dan.
Right.
But there's one thing I learned.
You need to set boundaries because if you don't set boundaries, nobody will.
You own that to yourself.
Yeah.
Yeah, I think the boundary piece is interesting, right?
Like, yo, as product managers, as design leaders, as dev managers, right?
There's 8 jillion inputs on you at any given time, yet you're also in a position of being tasked to solve complex problems, right?
How do you find time to think, right?
First, you need to decide what to think about.
So once I read the book called The One Thing, so I started applying this.
Like every Monday morning, I would say, what is the one thing I have to achieve by the end of this week?
This is a must have.
And then I would look at my agenda and say, what is contributing to that?
And sometimes there was nothing contributing to that and many things distracting me from particular that.
So I said, I need to adopt my agenda for what I need to achieve this week.
And then once I achieve that, I allow myself to do other things.
Yeah.
I had a really good time in your workshop yesterday.
I am now going to ask you about assumptions.
I thought one of the most valuable exercises that I did in the two workshops yesterday was the exercise where you asked us to actually just list the assumptions about the eight ideas that we had created around the scenario that you had us run through.
Why are assumptions so important?
And why is like cataloging them before you actually start to build anything so important?
Because of our lab of confidence.
It's very fast for us to fall in love with ideas.
And we start asking questions to confirm how right we are.
And then the more we get positive signs, the more we are willing to invest.
So I want to help people to understand that they don't have all the answers.
And the way of getting people to understand that is by...
helping them see what they don't know.
So if you just slow down for 10 minutes, asking like, what am I assuming to have?
And it's very interesting, the results when I do that.
Sometimes I do the exercise and I ask like, how confident you are in your idea?
Zero to 10.
A lot of people say something between seven and eight.
And then they do the exercise.
I ask, what about now?
Two or three.
The execution after that will be very different.
You will build to learn.
And then you will focus like, I need to learn, I need to uncover this and so on and so on.
And in the end, you are going to create something that makes sense because you learn from reality.
If you don't do that, you're going to make something that only works in fantasy.
Yeah.
So assumption checks, especially with AI, right?
It's really easy to come up with, you know, 50 assumptions.
How do we figure out what assumptions we really need to care about?
Because if we go chase all of them, we'll never get to the doing part.
We need to categorize them.
So first we need to understand the following.
Ask the first question.
If this assumption is proven false, is my idea still valid?
And if the answer is yes, well, then you can literally ignore this assumption because your idea will still survive.
But then you say, hmm, if this is proven false, my idea is worthless, you start looking at this differently.
Then you say, How confident am I that this assumption is true?
That is your data.
And if you talk to a few people, your confidence should be very weak.
But if you have a payment commitment or something like this, then it's strong.
So for this strong evidence, you don't have to test.
But for the ones like weak evidence and business critical, yes.
Then you need to figure out how you test.
And then it goes to the thing.
Most people will default to...
the same experiment methods most probably it will be a b testing prototype or something like this that's the moment ai can be your ally you can ask ai to help you create experiments you can run in four hours eight hours two days or a week and then you look at that and say oh let me start with the ones i can do in four hours so i learned something fast instead of just building something yeah you had a real simple text on me for like what is in like what makes up an experiment.
So like what are like the, what are like the key data elements of a great experiment?
So an experiment will have the assumption you were testing with whom you were testing.
What do you expect as a result?
Defining success.
There is one mistake many people do, which is very clear.
Like they say, let's run an experiment.
And what is the experiment?
We're going to interview a lot of people and then interview a lot of people.
And then they spend two hours discussing the results.
And then if you ask like, Was it a pass or fail?
So you need to define success criteria before.
A clear example like I gave yesterday is when I wanted to launch my book, I wanted to get to interview a few people.
But I said, I don't want to send an interview invitation.
I want to say I'm writing a book.
And then based on the book, table of contents, I wanted to find people who were interested to talk to me.
So at the end of the LinkedIn post, I said, If this lands well with you, I want to talk to you.
Success for me meant I will get seven people who want to talk to me in four hours.
I needed to get seven people DMing me and say, I want to talk to you.
And I got and I said, OK, this is a pass, so it works.
But that was clear to me.
If I get five, I would say, OK, the copy was not good enough, so I need to do something.
Yeah.
When we did this exercise, you said this earlier, one of the things that stuck out to the group of four that I was working with, is the first assumption that we actually had to wrestle with is that like we all thought the idea was a great idea.
Right.
And that if we liked it, then obviously everyone else would like it.
That's the best idea that's ever been had.
Yeah.
Right.
Yeah.
So I just I love the way that that exercise helped us kind of uncover some of the bias that we have about ideas that we create and how you can get too attached to them.
So throughout the talks today and it came up yesterday, we can talk about AI and the need for human thought.
Yesterday, you mentioned a technique and how you prompt where maybe you're asked for assumptions, but one of the skills you've put in is, hey, ask me to put my stuff in first.
I would love to hear you expand upon how you came to that as a working model.
So it's an interesting thing.
There are some people, I respect them a lot.
And one of these people, one person told me the following, David, I'm looking for a new job and I am failing.
every single interview and I said what has happened tell me more about it she said I don't know I understand now everything about skills evals agents orchestration and I explain all of that in the interview and then they fail me and then I started looking at that and said walk me through and what I realized that this person who is super authentic she was not her anymore she was just a toolbox I said what I am missing is who are you and Once I heard that, I started observing what people are doing.
They are figuring out how to automate things so agents can run things flawlessly, end to end.
And then what happens is throughout the way, agents will make some assumptions and then you will not check and you will get to some results in the end you don't understand.
So what I wanted is like decision points.
I wanted you to, you know, you can't get muscles without exercising.
So I said, you need to warm up your brain, do some exercise.
So once you see the result, you are already warmed up so you can criticize.
So for example, instead of somehow asking AI to come up with assumptions, you first need to, I force, you first come up with a few assumptions and then AI will amplify and then you are going to decide based on what you see, which are the ones that make sense, then you advance.
So decision point.
And this gets through another thing.
It's not always faster.
But it can get better.
Okay.
How do people, when they're looking at implementing an agentic system, how can they maybe understand where AI is going to have license to make those assumptions?
Like, where do they find, like, actually, this is a weak point, and, like, it's going to make a bunch of assumptions here if I don't actually articulate what I'm looking for at this step?
AI is going to make assumptions all over the place.
Because...
You give a task and AI will complete the task.
So many people go from idea to prototype.
But from idea to prototype, there are some things in between.
There is like, what are the assumptions?
What are the ones we should test?
What do we know?
What don't we know?
So we need to slow down.
So what I want is to break whatever you're trying to achieve into smaller steps with outcomes.
And then you have the checks.
You can use agents for that, make it transparent.
For example, you can integrate with Miro and say, first I want to see this, then I will give the thumbs up, then you move to the next part and to the next part.
But you have the chance of making transground because one of the things that AI will do if you don't intervene is we'll amplify misalignment because it will be silo work.
Not everyone will see what is going on.
But if you make it transparent, then everyone has a chance of seeing what is going on.
step by step, and then you can just find why are you taking this path instead of another.
You hit upon something really interesting there, right?
And the idea of siloization and AI amplifying, right?
And if you're prone to working by yourself, now you have AI and now why ever speak to anybody?
I mean, I'm curious, you know, how you kind of break that pattern and kind of force yourself to go and collaborate when it's like, hey, I can just talk to Claude and here I'm done.
You need to first step.
So if you want to do this, for example, I understand that the teams are going to change the size because I imagine in the future, every person in the team will have a series of agents executing things.
So if execution becomes cheap, where do you put your energy?
Your energy goes to thinking, decision-making.
But thinking and decision-making should be a collaborative exercise, not a silo one.
So for example, with the team, you agree where you are going.
And he didn't say, we have decision points, so let's align on this.
And how you align can be different, can be asynchronous, or can be meeting, as I mentioned, 15 minutes to decide, say, we have this output, we have three different paths to take, which is the one, and why?
Or maybe we are going to come up with a fourth one.
So you force this decision parts collaboratively.
All the times I feel like I've had a brilliant idea, I'll go talk to somebody else, and they're like, yeah, no, that was obvious to me.
It's like, ah, great.
You told a great story that I am hoping that you'll recount for everybody, but we talked about kind of like that bias of like, our idea is good, we've made the decision, and we don't want to hear anything else.
And you took the opportunity when you were working for a famous car manufacturer to actually challenge that assumption.
Can you regale the crowd with that story?
Sure.
Not sure if the crowd know this car manufacturing.
Let's check.
Do you know British motor wagon?
BMW.
So what happened is I was working with them and they asked, how do we build the best navigation system in the world?
And they had a lot of money to do that.
And I was trying to understand what they tried to optimize and why they had to build that.
And I asked them, how do I talk to customers?
I want to talk to a customer to understand what is going on in their mind so we can optimize for their needs.
And they told me, I cannot talk to customers.
No, compliance reasons, you cannot talk to them.
Oh, but I live in Munich, Germany, so, you know, I said, every second person here has a BMW, I'm going to walk on the street, and the first person that parks the car, I approach.
And then I would ask, like, if you're going to go from A to B, how do you choose to go there?
Nine out of ten answers were, like, either Apple car or Android Auto.
And then I got one person saying, I use...
the BMW navigation.
And I said, okay, how do you like it?
I hate everything about it.
And the only reason I use that is because BMW is charging a subscription for Apple car.
I refuse to pay.
And then I look at this.
I talked to a few more people, but the pattern was clear.
I went back to BMW.
I said, it didn't allow me to talk to your customers.
I talked to BMW car owners who live in Munich.
And here's what I learned.
And BMW was looking at me and said, What are you saying?
I said, I am saying that people use their phones for navigation and they like that.
What they don't like is plugging the cable.
Because plugging the cable adds a lot of friction.
Sometimes they break the cable or something like this.
I said, if there would be a different way to remove this friction, it would be better than building the best navigation system in the world.
And BMW said, all right, let's explore this.
That's when they started building a frictionless way that you just enter in the car and it's connected automatically.
You don't have to do anything.
They still invested in the navigation though, but they prioritized that one first.
Yeah.
I love that story.
I love just taking it upon yourself to just say, well, I have to go talk to customers, right?
Right.
I feel like you got through a little bit of bullshit management there.
All right.
A lot of talk.
So there's one thing I learned earlier in my career is about one of the best friends of a product people I mentioned yesterday, Nihito.
Some of you probably know Nihito by now.
If you don't know Nihito, it's nothing important happens in the office.
So I learned about it and I was like, I always look at companies.
How are they making decisions about what they are doing?
And many times decisions are inside out.
So inside an office room and so on and so on.
And when I lived in Brazil, I remember once I was developing an app for food delivery and so on, and I was checking, I said, if you really want to understand how these things happen, you'll need to be on the move.
So, Brazilians always do things on the move, thinking, when I arrive at home, I want my food to arrive with me.
So, they are going to order that when they are in the subway or something like this.
So, I started testing the app when I was inside the subway.
Like if the metro would fit like 100 people, there were 300.
And I was trying to do this and so on and it wouldn't work.
And I was telling people, this app needs to work very well with one hand.
And I need to see what is important when I have a screen like this.
So you need to go and check reality.
Because if you do in the office, you have a lot of space.
You are sitting there very cozy, very calm and so on.
But that's not how people order.
You need to understand reality.
And that's what I thought about being a doctor.
It's the same thing.
How do people choose?
What is reality?
What is the real friction?
Not what is happening inside the office.
David, you are an amazing storyteller.
Thank you for the opportunity to use a little casual profanity on the podcast.
They normally don't let me do that.
This one's getting an E.
Yeah, right?
Thank you so much for being with us today.
Thanks for coming and hanging out with us.
My pleasure.
Thank you.