This is a cute categorization. The stories they tell are charming and reinforce the categorization. But, I've never found things to be so cut an dry. For example, a lot of the work we did in AWS often fell into the hair on fire category, and the products that showed future vision e.g. AWS Glue, Aurora, DynamoDB were the ones that rose above the noise.
It mostly boils down to this: delight customers and iterate as fast as you can.
> delight customers and iterate as fast as you can
One thing is the product. Another thing is your messaging. Yet another thing is what goes in people's minds when they interact with your product and your messaging. Fourth is how you reason about the complex interactions between the three prior factors.
This categorization helps entrepreneurs with #4. I found it very helpful.
> It mostly boils down to this: delight customers and iterate as fast as you can.
This is maybe the second phase after AWS found a fit and built a consumer base? Once I am in the housing market, I have a need for everything (mortgage, building contractor, construction materials, designer, hardware and accessories, upholstery, decor, etc).
My entry to AWS started with EC2 in the very early days because it of its commodity nature (any size and shape, for however long, with per-minute billing) and instant availability. The elastic nature solved scale. A lot of people didn't move to RDS until later but it was inevitable.
Everything else followed on from there, cross-sells and up-sells for reliability and convenience were always a click away for captive consumers who were already onboarded.
My perspective changed from thinking PMF was a lottery when in reality its just sales/marketing grinding until something works.
The reason I used to think this is like other technically orientated founders. If I build for a market that doesn't exist then you are playing lotto. It's not even good odds, you are better off playing 0dte spx coin flip.
What you fear you build for and if all you have are unknowns then you are going to simply end up building everything lose differentiation and your potential customers even if they find you will walk
Despite classics like the SBF article that was taken down, this framework is quite good but the best thing to do as bootstrappers is to focus on generating net positive cashflow as soon as possible not running a PMF search engine which requires lot of minds and capital injection.
My partner started a business and literally from her first ad she was inundated with customers. Now it's a $10m/year bootstrapped business. I've done exactly the same thing and had 0 customers.
If you are falling into the Hard Fact / Future Vision categories it can very hard to get customers because no one is ever looking for your product.
There's not enough details here to say what she did right and you did wrong, but it generally comes down to her idea probably solved a problem customers were willing to pay a lot for today and your idea didn't.
If her idea was "when you pay me x, you get 3x back in new revenue" and your idea was a tarpit idea like "help people find what to eat tonight", it's really obvious why you failed.
Wrong, even if both agents are identical, and they're operating in the same environment, each run of the simulation will be slightly different, i.e. the process is stochastic (non-deterministic).
In reality the agents are not identical (e.g. different skillsets, or attractiveness to clients), and the environments are different (e.g. different professional or social networks).
And, as another commenter pointed out, the time when agents are starting is also a factor, but my point is that even if they start at the same time the result might be different.
Lastly, there is a butterfly effect - that small changes in the initial conditions may result in vastly different outcomes.
I spoke with the ceo of a top ten AI company on Friday. They are definitely playing the PMF lottery and there is no alternative to this. They have lots of very smart people and tons of conversations with customers. But nobody has any idea what will end up working.
Plenty of businesses - probably especially AI companies that do original research - found PMF in a market that’s too small for their cost structure. Like I bet Anthropic and OpenAI have positive margins on their raw APIs if you exclude all R&D costs. Even Google’s AI research is subsidized by one of the largest profit generators in Silicon Valley.
Its not a lottery, most people just arent intelligent enough to anticipate new market needs. They have no business trying to innovate at that level, but are unaware of their inadequacies. Some people do have the raw intellectual and societal horsepower to bring new paradigms to life.
I wouldn't say its intelligence as there are different types of intelligence but more specifically emotional intelligence and self-awareness that is key.
The guys writing PMF is a lottery are jaded because they are afraid to admitting their own weak points. They failed because they ignored something. The ones with emotional intelligence and self-awareness are able to navigate and negotiate.
Founders keep fixating on pick-me startup, hoping for a cushy exit, that happens less than 0.1% of the time, the odds are as bad as wanting to go to NFL, if not WORSE. VCs and investors also have to deliver returns they can't keep bank rolling something that they don't see a market in.
There’s also a lot of confusing finding / getting PMF with what PMF is.
Product market fit is pretty straightforward: I have a product that meets people’s needs in a way that the market demonstrates by paying money for (usually with the connotation of profitability).
Finding this is incredibly hard and rare, thus it seems to get convoluted with a lot of emotions.
You don't. And most of these OpenAI wrapper startups are failgets. The consensus from the buyers side is that they are disillusioned by the hype and dealing with the true reality of this hype cycle.
Can you recommend any practical resources on the topic? Not fluffy business-orientated self help books like all the product management books that are recommended.
But also books on business strategy and marketing are essential. I read several at uni. Maybe pick up an HBR guide or FT book if you don't want to go for a full textbook.
They should give you more ways to think about problems, let you be more confident in ideas you choose, and reinforce the need to iterate quickly and only gradually commit more resources.
Most entrepreneurial advice could probably be summarised as: make products with new tech that solve real problems.
The new tech part means you get in before competitors, and solving real problems keeps it laser focussed on customers with important enough problems they'll pay for it.
Obviously coming from Sequoia there won't be an bootstrapped approach, but I'd love to see one for that.
For bootstrappers, the approach I've heard of is simply just nicheing down to find a segment of the market that is big enough to support you, but small enough to not be worth it to bigger players.
This has little to do with whether you are VC backed or bootstrapped.
The amount of effort to sustain yourself is going to be almost the same regardless of which category you are in. Either you have a category that is popular but will attract dozens of competitors no matter how small your niche is or you have a product that requires effort to market.
These days with so many no/low-code tools getting a product into market has never been easier. And that's both a positive and negative for bootstrappers.
Bootstrapped and infinite runway doesn’t go hand in hand lol.
Runway always costs. By that definition it’s artificially subsidized.
That definition sounds more like a project than a product/startup.
Bootstrap is a business, cash flow is like blood flowing with oxygen in it, so the body can move and go do thing
Steve Blanks’s definition is a good or that a startup is a temporary organization dedicated to finding a repeatable and scalable business model.
Life costs money, if a project is coddled to never become a product. it’s artificially subsidized by an investor (you) and default dead.. and it’s not a business that makes profit to pay the Bills or pay for growth.
> For bootstrappers, the approach I've heard of is simply just nicheing down to find a segment of the market that is big enough to support you, but small enough to not be worth it to bigger players.
Thank you, as someone hoping to bootstrap myself this is makes sense. Any idea how to make the niche smaller? The broader market I'm thinking of is huge (likely billions of customers), I suppose I could charge more to narrow it down but that has other risks.
Pick one single customer and fanatically meet their individual needs. Then add a second. Repeat until you’re no longer needing to tweak your system to delight a new customer. At that point you need sales and you have a business! :)
We had a high touch service for B2B installation: we charged clients for consulting (business analysis) to configure and integrate our system.
The cost to clients for the consulting was approx 1 year of our SaaS fees. The up front consulting helped heaps with our cashflow, but it also significantly limited our growth rate.
But it worked for us at the time (mid 2000's).
Definitely hard to avoid over-customisation of software for one client: but you need to avoid that problem with any software product servicing multiple clients.
Previous company with a similar product depended too heavily upon ongoing consulting fees and that caused a variety of troubles for that business.
In the idealized solution that I would build/provide, I'd already be meeting the needs of many individuals with "just" one top-of-the-line product. The only issue, and the need for quotes, is because this no-compromise product that would likely be competitive with the best in the industry, would also be quite expensive.
This, fortunately or not, kind of circles back to the "make more expensive products" aspect.
While I'm not convinced it is VC vs. bootstrapped that should differentiates frameworks, you should read this book, it is simply the closest thing to a step by step user guide to finding PMF as I've seen, and it worked for me: https://www.amazon.com/Four-Steps-Epiphany-Steve-Blank/dp/09...
you just need to solve a problem. you need to go talk to people and help them.
if someone isn’t using your product that you’re building to solve their problem and they are complaining about said problem, it’s because nobody else has tried selling to them.
I feel like there is quite a bit of overlap in hard-fact and future vision, and I'm not quite sure where our company would fit.
We're in the neurotech/sleeptech space, and while the majority of the market is selling "fall asleep faster", "sleep more", we're improving the efficiency of deep sleep with the health benefits.
This is a hard-fact - people are resolved to poor sleep quality, or no "real" solutions to improving deep sleep quality - yes, sleep hygiene is important, just like brushing your teeth is important, but that isn't the answer to a poor diet.
So customers are resolved to the "I'm just tired, that's the way it is" mentality or "I'll track my sleep, and now I know why I'm tired".
At the same time, they don't have the context of "there is a way to do this", which is why they were resolved to accept the status quo in the "hard-fact"
Does anyone else feel they are falling between the hard-fact and future vision?
Is anyone falling between Hair on Fire and Hard Fact? I feel this is possible too.
> I feel like there is quite a bit of overlap in hard-fact and future vision
Not to mention revisionism.
> Customers must believe that your product represents a whole new paradigm—often with its own ecosystem. (The iPhone wasn’t just a device; its App Store was a new way of interfacing with the internet.
"Its App Store" didn't come until a few years later. It was not a part of the iPhone vision, certainly not at launch.
There was a kind of behind the scenes ecosystem building in terms of AT&T's giving Apple an unusual (for the time) level of direct and complete customer/device control.
The iPhone wouldn't have been the iPhone without the work to create that wider context.
The better user experience that enabled Apple to give customers counted for a lot. The easy to use, free of telecom interference, App Store is a prime example that the unusual Apple/AT&T relationship allowed, even if it was not v1.0.
Looked at your product. If your marketing copy reflects reality, your product is Sci-Fi.
"The next phase of research is measuring the effect of stimulation on removal of beta-amyloids, related to Alzheimer’s prevention, concussion & TBI, and more. We’re actively speaking to researchers interested in examining the impact on insulin response for improved outcomes in people with type 2 diabetes, as well as athletic recovery and this is just the beginning."
But if you believed that this would actually have a significant effect, I feel like that'd be way up top, not an aside in a blog post. Which means you're a "Hard Fact" ("sleep is sleep") that isn't marketable, and you're reaching beyond your grasp to try to become "Future Vision."
Good point. The reason we don't lead with future impact is that we don't know if any of those things will work out. The reason researchers (not us, independent researchers) are looking into these areas is because they believe, and there is some evidence, that the stimulation could have an impact. Where our customers are focused on their problem.
This will become more obvious as we get closer to market and we start more directly targeting specific markets. We are currently building out our separate website which is more targeted at supporting researchers and other collaborations.
> This is a hard-fact - people are resolved to poor sleep quality, or no "real" solutions to improving deep sleep quality
I glanced at your website, and there’s already a product on the market that does the same thing (headband using EEG techniques to provide auditory stimulation based on the same research your site points out): The Philips Smart Sleep https://www.usa.philips.com/c-e/smartsleep/deep-sleep-headba...
So in your case, I’m not sure it’s accurate to say there aren’t any “real” solutions when a major market player has already released a product in the exact space. I think it’s more likely that wearing a headband every night isn’t an attractive proposition for many. That’s an entirely different problem to solve.
> I think it’s more likely that wearing a headband every night isn’t an attractive proposition for many.
Case in point, I personally cannot use wrist bands or any such accessory for sleeping as I sleep with my forehead on my wrists. Neither the wrist band nor the hand band are good options for me to use.
I wonder if there is something that can help someone who sleeps on their stomach not have to choose between crushing their nose or cramping their neck when they sleep...
Philips has not been available for years, and though there are many other EEG headbands, they are not focused on improving deep sleep. Though I agree, getting people to wear a headband is a difficult problem to solve as well.
iPhone was considered a future vision, but we had smartphones and even smartphones with apps previously, but at the bottom of the page, Sequoia points to the iPhone and Vision Pro as Future Vision examples.
I'm curious to see how that actually compares in effectiveness to wearing headband-type speakers and listening to sleep brainwave stimulation music (which I tried).
Completely different. Not only from a comfort perspective, but show me a decent study which looks at the impact of "sleep brainwave stimulation music" on immune function, cardiovascular function, memory, reaction time, etc etc.
This isn't "gentle tones to help you sleep". Directly interacting with slow-waves can increase the spectral power by 30%.
> Deep sleep is not improved by consistent sleep/wake times, and adjustments to noise/temperature/light conditions?
Deep sleep definitely benefits from consistent sleep/wake times. Your body tries to play catch up if you break from your normal rhythm but it’s not entirely effective.
I think the parent comment might be exaggerating the futility of sleep hygiene because they have a product to sell. :) Proper sleep timing, duration, and sleep hygiene are inarguably valuable for proper sleep architecture.
I'm not saying sleep hygiene is futile, I specifically say it is important, and sleep hygiene improves the potential for deep sleep, it does not directly improve deep sleep.
Sleep hygiene impacts sleep duration, which can impact all phases of sleep. It impacts the sleep phase potential, which is different from directly impacting sleep.
Measuring only sleep duration is like measuring only how much time you spent at the gym, ignoring what exercises, weights and reps you did.
Yeah I’m not sure where mine fits either. I have a video editor that removes pauses, saving lots of editing time for (certain kinds of) videos/podcasts. I would say I hear from 3 kinds of customers:
- They know about tools like mine and have compared a few
- They have had this problem for a while, decided to Google it one day, and found my Recut app
- They stumbled on it randomly via social media and say things like “where has this been all my life”
I would guess that some people from all 3 groups would describe it as Hair on Fire, but they have different levels of awareness of solutions. Maybe they just live with it as a Hard Fact, maybe they live with it because they know of solutions but don’t like them (of the flavor “I want to automate this but I don’t want to lose all creative control to AI”).
It’s tough to find good ways to get in front of the people who aren’t looking! It feels like that way lies a lot of broad/expensive advertising.
> It’s tough to find good ways to get in front of the people who aren’t looking! It feels like that way lies a lot of broad/expensive advertising.
What unrelated (to your solution) place are your potential customers looking for solutions to other problems they have, for whatever reason? Or at least a segment of them? Then figure out how to be of interest based on that motivation, in a way that lets you also naturally introduce your product - directly, or maybe better, seemingly incidentally.
There is currently a generation of founders who think that talking about PMF somehow gets you closer to PMF. Talking about it, having a great following online, being a decent marketer - does not mean zilch.
Knowing everything about PMF probably moves you single digit percentage-wise toward it.
It's like being a vampire, avoiding garlic (hell, even drinking blood) doesn't make you one.
However, OpenAI is a bad example of the future vision paradigm. In fact it fits 1/6 of the characteristics they identify in this paradigm. A better example would be something like Tesla.
I would say OpenAIs backing of the GPT model family in itsself is a better example of the future paradigm than anything to do with OpenAIs founding or non profit status that sequoia discuss. Going all in on GPT gave them a big headstart (although google should have been fighting much earlier if they had listened to internal researchers).
iPhone and Vision Pro are also not future vision. Both enter a market at the high end, with established competitors. BlackBerry and Nokia had smart phones that were functionally equivalent, but they were marketed towards business users and worked best with corporate integration. Apple delivered a phone that focused on personal use. Vision Pro is entering market where the competitors are focused solely on gaming or augmented work (hololens) with a product that doesn't do either, yet... but everyone is starry-eyed about what it might enable, some day.
I'm curious how education or entertainment companies fit into this framework, or how they think about PMF in general.
Was Duolingo targeting customers who accepted the "hard fact" that passive audio was the only way to learn a language? Was MasterClass a "future vision" because people didn't believe celebrities would spend their time teaching? Or is it that we NEED education and entertainment, so these two providers just differentiated from a crowded market.
Alternatively Duolingo sells a casual memory game that differentiate by giving the illusion of productivity. And Masterclass sells attention and status to a class of people who are already attention-oriented by stroking their egos. I’m not saying this hot take is necessarily true, but the market isn’t always categorized by what we intuitively think. And realizing that can sometimes give you a massive benefit, simply by applying existing and effective techniques (in these cases gamification and people-oriented success storytelling, respectively) from a domain your competitors don't understand or care about.
Yeah, to me, Duolingo seems to be Hard Fact. But MasterClass might be also Hard Fact. MasterClass is not a totally different education service compared to the existing incumbents. Online university could be MasterClass.
For our company we are in 2 phases.
In the cloud, it's 'hair on fire'. MLOps is a hot space.
For on-premises, MLOps is basically download MLFlow and Kubeflow and you think you're sorted. So it's the 'hard fact' segment.
Not on premises who are regulated and might have enterprise requirements. If you need eg role based access control you can’t use on prem MLFlow etc. So you could niche down to customers who don’t know this yet. Smaller banks, insurers etc.
Edit: we went through this. Started putzing around with open source things on prem on Openshift. Ended up on Databricks for a variety of reasons. The impact on the business to be wiring up lots of new widgets, maintaining them, training, ability to switch to new components when required. Lean team, so turn Databricks on and secure it properly, exclude PI where possible, sort out ETL and CICD etc, profit. Delegate MLOps stack innovation to the vendor and focus on our own job.
Yep, this journey is exactly the kind of thing I run into when I get prospects who want our product (observability) to be on-prem. What they actually want is assurances around data storage, retention, and controls; and a "throat to choke" to cover their asses in case something goes wrong. At least in this domain, very few people who need Observability actually need things to be on-prem as well.
A nice framework that's constructed retrospectively. Will it hold true in the future, maybe, maybe not. The thing that makes the most difference are not these broad ontological things, but short time horizon tradeoff equations around talent/skills available in the market, access to early customers and who/what they want, your personal network, your innate temperament/mindset in making day to day decisions.
RudderStack entered a crowded Customer Data Platform market in 2019. It clearly differentiated from other platforms by choosing to not host the customer data on its platform rather choosing warehouse-native and Open Source approach where you store data on your own central warehouse only. It made the solution best-in-class solving obvious data quality and governance issues. The early success in such market is seen clearly in the fact that customers replace their existing product (Customer Data Platform providers, the leading ones) with your product (RudderStack).
Is this correct categorization (I feel lines are quite blurred between the categories but nevertheless useful)?
in a way, this framework represents how far off the reservation you're going. You're either building in the known, challenging existing assumptions, or creating new landscape.
Other wise, life on planet earth is a lucky accident, you being born was an accident, you now getting sever autism is mistake, you reading my comment is a lucky accident, You being able to understand what I am reading is just luck.
Life/reality does have some randomness but it's really naive to call all Randomness simply luck.
What is more accurate to do is to define things in terms of odds (probabilities).
And call something as having higher probability or a lower probability. Otherwise, everything does have an element luck/randomness.
no malice intended, but I will take my cloud provider's managed database service.. for which we get support, SOC2 compliance, years of proven stability, no major approval needed by my organization, and budget already approved - rather than jump on the latest 'kubernetes is eating the world' fad.
I wish they discussed startups that didn’t require a lot of capital or even bootstrapped startups. But of course they’re a VC firm so of course they can’t
This sounds a little nutty but if you want to be a bootstrapped Rippling, it's possible.
A trick I've found repeatable is to find an enterprise that doesn't shy away from big budgets (where six and preferably seven figure budgets are fine even for unproven vendors, because what's big/existential for you is negligible/experimental for them), and is willing to buy a product from you that requires "integration" (they love to spend money on integration, plus that means it doesn't work on day one!), where the "integration" cost is enough to bootstrap the product AND integration (and your ramen!) at your internal rates.
To avoid the bootstrap money going to overhead, to keep the money for product build (engineering instead of paperwork), you need to avoid getting "risk managed" like a mature vendor. Its best for you to be a pilot, or proof of concept, or pet project for a sponsor to show a concept they want the enterprise to do.
Pick your customers and product integrations carefully where the work will pay for the enterprise capabilities most likely to win your next sale. Use these big contracts like lilypads to cross the pond to the other side of Geoffry Moore's "chasm" between early adopters and early majoriy.
Impossible, unless you already have an in with someone high up in the management chain to approve your $1m/3months proposal.
Enterprises that don't shy away from big budgets have a constant stream of multiple suitors all thinking the way you are, many of them more experienced with doing this as well.
In case anyone was wondering how to make such connections: attend a top 5 MBA program or have a long history in management consulting, preferably both.
Contrary to sibling replies, it's possible to find.
For example, when these mega enterprises realize they're behind, they undertake "Transformations" or strategic initiatives to "Digitize" this or that.
Look at enterprises that could use your product, see if any are making noises in the press about landing a Transformation Officer or Digital Officer or similar roles. Or work the other way, search for hiring announcements around such titles, and see if the firms hiring them are ones that need you to pull off that change.
Either way, when they land that new leader, then look for hiring announcements for their directs or in their product or technology areas related to what you could help do. If you see they hired from "outside", particularly if they hired from Silicon Valley, the Magnificent 7*, or other "Fast Company" cultures, you may be able to cold contact those new hires (LinkedIn InMail, standard email address formats, etc., there are hustle guides to this step).
Do not form mail at this step. You get one tight clean shot to hook them that you might have a genuine way to help them, it's worth a conversation. Make having that conversation cheap, meaning, offer to make time convenient for them, 30 minutes is fine to just talk to see if your idea fits, etc. They don't want a dog and pony show. Be painless for them to check if you can help. (Note: Be completely unlike Oracle, Salesforce, Snowflake, or any other vendor where if your desired sponsor takes a call or email they'll be hounded by dozens of sales people for months, and that's just to start a 9 month procurement cycle. In contrast, your sponsor guesses a small firm could ship by then, and they wouldn't have to waste time ...)
If what you have could legitimately help them deliver on the Transformation or Digital initiatives' goals, they may get back to you**, because the last thing they need is to deal with even more enterprise vendors, they need some small fast vendors to help them actually put wins on the board.
** I have been that guy, and I have gotten back to small firms. (I also occasionally reach out to startups here.) Peers at similar roles have gotten back to people too. It is much more common to get introduced through VC-hosted field days, senior exec or board room referrals, and the like, but this path can work with enough hustle.
yes, i was part of doing something like this at a small elearning startup for one of the largest publishers in the world circa 2010.
Exactly as you say, brought in to provide the tech platform for a strategic initiative involving a transformative shift to digital-first courseware and content.
You have to also treat write ups like this as mostly marketing for the firm. There was a discussion recently about A16Z (https://news.ycombinator.com/item?id=39901289) that made me think about that angle here as well. All these frameworks may have useful information but they are also simplistic. I think this is marketing for their Arc program, which is a selling point for early stage startups to consider Sequoia. In other words, this is about improving their deal flow. It also probably helps filter the flow, since now anyone seeking early stage funding from Sequoia will do the work to try to fit to this framework before wasting anyone’s time, and their pitch will also be easier for Sequoia to understand and evaluate.
It mostly boils down to this: delight customers and iterate as fast as you can.