
Pulse
Story
Your real competitor is a spreadsheet
Every builder keeps a mental folder of competitors: the other SaaS in the niche, their pricing, their changelogs. My day job is demand research, and every run we do maps what a segment currently hires for the job. A competing SaaS leading that list is the exception. What actually comes back: a Google Sheet named tracking_v3_FINAL, a VA who "handles it", an old habit, and the most undefeated rival in software history — doing nothing and living with it.
The mechanic behind this is worth knowing precisely, because two funnel problems hide inside it.
A person with a job doesn't scan the market. They pick from the handful of solutions they already know can do the job — their consideration set. It's small, it's lazy, and it's dominated by whatever they used last time.
Problem 1: you can't get hired if you're not in the set. Most "lost deals" of an unknown product were never comparisons you lost — you weren't in the room where it happened. For an early product, awareness isn't a marketing nicety, it is the funnel: before anyone weighs "is this better?", a stranger has to learn your thing can do their job at all. This is why content that names the job outperforms content that names features — it's how you enter the set of people who never searched for your category.
Problem 2: the bar isn't "better than the competing SaaS". It's "better than the current mess, by more than the switch costs". The spreadsheet is free, already set up, and asks for zero new habits. That's a genuinely strong deal. Beating it means your whole offer — outcome, price, migration, learning curve — has to beat good enough. A longer feature table doesn't even enter that equation.
A corollary that saves real money: poaching a competitor's customer only works per job. Nobody leaves product X "in general" — they re-hire for one specific job where X underdelivers, and they weigh the switch against their history with X, not against your feature grid. If you can't name the job and X's exact failure at it, your comparison page is aimed at nobody.
The exercise, doable this week: ask five recent signups one question — "what were you doing about this before us?" Past tense; imagined behavior lies. The answers are your real consideration set. Now reread your landing page against that list. Most pages argue with a SaaS the user has never opened — and say nothing to the spreadsheet, the VA, and the do-nothing, which is who you're actually up against.
Nobody wants your product
Not yours specifically. Anybody's.
A marketer wraps a 12-hour day, brain fried, and opens Netflix. He doesn't want Netflix. He wants to get from point A — fried, wired, can't sleep — to point B: switched off, asleep in time to function tomorrow. Netflix is just the cheapest known route from A to B. The day something walks that route better, his "loyalty" evaporates, and he won't even notice it leaving.
That A→B transition is what Jobs to Be Done calls a job. Products get hired for jobs — and fired the day a better hire shows up. Nobody wants your app, your onboarding, your settings page: whoever signed up had a job and didn't know a better way than pushing through your registration form. That's the entire relationship.
A job isn't a use case, and it isn't a feature request. It has a fixed anatomy:
— Context. Point A: where the person is, what they know, how they feel. Fried after a 12-hour day.
— Trigger. The event that starts the movement. Closed the laptop on the last must-do task.
— Expected outcome, with criteria. Point B as they imagine it before acting. Switched off, zero effort required, asleep by midnight.
— The higher outcome. What B is for. Be sharp tomorrow — he's mid-project with a deadline.
Written out as one sentence — this is the template, steal it:
When [context + trigger], I want [outcome + criteria], so that [the bigger outcome + how I'll feel].
"When I've closed the day's real work and my brain is fried, I want to switch off with zero effort, so that I fall asleep and I'm sharp tomorrow."
Two rules for using it:
1. Match detail to decision size. The minimal job is just want + verb: "want to close the cookie popup", "want to see the discount". Micro-feature — micro form is enough. Positioning, pricing, new product — write the full sentence.
2. Every word is load-bearing. "Switch off with zero effort" and "unwind properly" are different jobs — the first is won by autoplay, the second maybe by a yoga app. Half the roadmap fights I've refereed were two people holding different jobs written with different words, neither of them out loud.
If "jobs" smells like consultant vocabulary, here's the grounding. The real drivers of behavior are needs — rest, safety, status, belonging. But needs are unconscious: ask a user about theirs and you'll get a confident story that's mostly fiction. The brain converts needs into goals, and goals people can actually tell you about — especially in past tense. Goals are jobs. It's the deepest layer of "why people buy" you can observe and build on.
So, the practical move. "How do we make the product better?" has no answer — better at what? "What job do people hire us for, and what did they fire to hire us?" has one, and it's sitting in what your users did last Tuesday. Write your product's job in the template above. If you can't fill the blanks from real conversations, that's not a writing problem — that's the first thing worth finding out.
All the money gets made in session one
I read onboarding and paywall teardowns the way other people read fiction. Everything new in consumer product ships there first. That's where the tricks live.
Here's the user's POV:
- You're on the couch, scrolling
- An ad stops your thumb. You tap
- You install the app, open the bot, drop your email on the landing. Doesn't matter which
Then one of two things happens.
1. You get dumped into someone's pet project. Stuff scattered everywhere, arrows pointing at nothing, and now you're supposed to figure it out — something you had exactly zero intention of doing sixty seconds ago.
2. You get walked down a hallway. Here's what it does. Here, try it yourself. Here's everyone who walked down this hallway before you and what they got out of it.
Option 1, you close it. Gone forever.
Option 2, every minute in you're a little more invested.
So what are people actually doing in there?
- Founder's personal story, dropped right into the flow
- Before/after images
- 50% off the instant you dismiss the paywall
- Video testimonials
- Some teams testing 99% discounts and weekly plans
None of this is decoration. All of it is stealable. Steal it.
Traffic costs the same for everyone. Conversion to paid — and to renewal — does not.
Which is exactly why one founder posts "paid acquisition doesn't work," while another one keeps working the product, keeps testing, and runs 3-4x on the same traffic at the same CPM.
Until your first 10 paying customers there is one job. Polish the first session. That's it. That's the whole list.
The catch: you can polish it blind for a year.
Since GPT-4, AI predictions keep missing in the same direction
There are 4 camps on AI:
- It's a scam
- AI today — already does most digital work
- AGI tomorrow — does any work a human can
- ASI the day after — does anything cognition can theoretically do
People from different camps can't really argue; they're describing different worlds. But one thing is checkable: since GPT-4, serious predictions about model capabilities have missed in one direction only. Low. Progress runs on an exponential; we think in straight lines. And linear is comfortable — "Fable is the plateau, it slows down from here" is being said right now, the way it was said at every point of the curve so far.
The other comfortable belief is commoditization: the AI layer flattens out, roughly equal and roughly cheap for everyone, and your product just sits on top of it.
Look closer at the gap between the frontier and the second tier, though. Benchmarks understate it, because benchmarks measure puzzles — HLE questions professors spent a month composing. The economic effect lives somewhere else: how long an agent can run real work — code, support, sales, analytics — before it falls over. That's where the billions are, and there the gap is embarrassing.
And it won't close. Recursive self-improvement is no longer a thought experiment — labs use today's models to build tomorrow's. The frontier takes the whole Pareto curve: best at intelligence first, then at speed, then — competition handles this part — at price.
Which makes today's capabilities mostly irrelevant to your roadmap. The next 2 years will move further than the last 5 did.
Now map that back onto the camps:
- Fable can't feel or dream, useless toy
- Fable is neat — maybe a bit better than the rest
- What can a team of a thousand Sol agents, working the same hard problem, actually do?
- Sol holds a business problem for 10+ hours; GPT-3 lost the plot by paragraph 3. That distance took 5 years. Draw it again, forward.
If you're building, here's the practical read. The how keeps collapsing toward free, and every moat made of how collapses with it. Two things don't ride that exponential: knowing what's worth building, and proving that what you built is real. The first is why BHAG exists. The second is why the platform you're reading this on does.
We ran BHAG on the two most uncomfortable subjects available
Itself, and my other show.
The self-run won an argument with me that took a week. The Indie Valley run left fingerprints you can still see on that platform's positioning.
Both reports, receipts included: E04 · Patient Zero and E05 · Crossover.
The first crossover in the Valley: show #2 runs discovery on show #1. What a simulated market said about Indie Valley before it launched — and the fingerprints you can still find on the product.
E5: Crossover
We pointed BHAG at the most uncomfortable idea available: itself. What a panel that doesn't care about our feelings said about our market — including the finding I argued with for a week.
E4: Patient Zero
Nobody calls a wind tunnel fake air
It's a model of the sky, labeled as a model, and you put the aircraft through it before the sky — because discovering your design flaws at altitude is an expensive way to learn.
That's the line between simulation and lie, and it's the whole argument of E03 · Fieldwork.
Ask a chatbot about your market and you get a beautifully written literature review. Fieldwork is a different thing — how we cast a market instead of summarizing it.
E3: Fieldwork
"Cool idea" is what people say instead of no
Compliments end conversations. Customers start them. If everyone loves your idea and nobody signs up, you don't have validation — you have manners.
E02 · Politeness is about the difference, and the 5 questions that cut through it.
"Cool idea" is what people say instead of no. The 2am doubt is right to ask — here's how to actually answer it, and the list of questions that only ever return compliments.
E2: Politeness
Show #2 is live in the Valley
BHAG — the product I co-founded before Indie Valley existed. It answers the question the AI building tools made urgent: fine, you can build it in a weekend. Should you?
Its revenue line here starts where every honest line starts. The pilot is up: E01 · Direction.
Anyone can build the thing now. Nobody got better at knowing which thing. The pilot of my second show here — the product I co-founded before Indie Valley existed, and why it goes first.

