Navigating the Idea Maze
Launch Countdown to 2:22 02.22.2022
What do a hiker, an explorer, a machine learning algorithm, and entrepreneurship have in common? They all have to take a step without knowing if they are going in the right direction. Today we will discuss the idea maze and load up on metaphors.
Make sure to follow AllSpark [in]/getAllspark, @founderYonz, and join W3B Discord (a hacker community for people building a better consumer internet). We will continue to drop Easter eggs across our social and community channels. This is Part III of the Launch Countdown series, the last one before the alpha launch tomorrow at 2:22 02.22.2022. Feel free to check out the previous articles: Part I: Launch Fast Launch often and Part II: Single Decisive Reason for CVP. If you haven’t already signup for the waitlist below.
The idea maze was coined by Balaji S. Srinivasan known for being the x-CTO of Coinbase and one of the first people on the Bitcoin rocket. The idea traces its origin to the paper Market Research, Wireframing, and Design sometime before 2013 as part of the Stanford Startup Engineering course.
The idea maze is the massive search undertaken by entrepreneurs to create and capture value in a nebulous space. If you search for the term you will discover a lot of content around it, sort of like suddenly seeing your new car everywhere. The idea maze tries to push back against the common misconception that great businesses are built based on a great idea or a lightbulb moment instead of diligent work over a long period of time. You may be skeptical but when Paul Graham, Marc Andresen, Chris Dixon, and Balaji S. Srinivasan all mention the same thing, it is probably best to listen.
It is called a maze because you can not know if you are going the right way until you actually get to the end of the maze. Sometimes you will need to backtrack (CS pun intended) and maybe even go to the beginning if you have to.
Taking the first step
Programs that try to solve a problem by making the optimal decision at each step are called ‘greedy algorithms’. Implementing these systems is the equivalent of training wheels for your undergraduate education (in terms of complexity, they are still the right tools for a lot of problems).
In the program below, the heuristic is always picking the larger value. You can quickly see that you need to make the sub-optimal decision to get to the actual largest maximum number (99) also known as the global maxima, instead of settling on the discovered max (6) also known as the local maxima.
The idea of local vs global maxima permeates far beyond algorithms and machine learning; it is a big part of decision-making. Humans have bounded rationality — we attempt to satisfice, rather than optimize. For example, when you are hiking and you see the peak it generally doesn’t make sense to walk in the opposite direction. In the real world, we usually have limited information and little time to process it (bounded rationality). To make matters worse, it is far more common for greedy algorithms to be right than wrong. This predilection for greedy decision-making has spawned the famous adage: “When you hear hooves think horses, not zebras.” So where am I going with this? You need to identify if you are in a maze or if you are hiking when building your company.
One thing all businesses have in common is a certain degree of uncertainty, but the degree and type vary and impact how you build your company. Startups in existing markets, B2B companies, and ones with strong leading indicators of PMF skew more towards the hiking analogy than the maze. In these startups, you should be able to make more optimal decisions. Consequently, most entrepreneurial advice can be applied directly. But what if your business is more in the maze category? There is still a lot of advice you can and should leverage: ship as early as possible, talk to customers A LOT, build cheap prototypes, and the list goes on. However, you need to be aware of the difference and apply it to your context.
Allspark is a B2C company focused on building a new category: “Products that help you streamline your internet experience”. It would be great if we could describe the company as ‘Uber for X’. We imagine that Uber and Airbnb had to spend a lot of time explaining their business whereas new startups would just say “Uber for medicine” or “Airbnb for couches.” Since entrepreneurship is like catching lightning, we have done our best to follow best in class advice while recognizing we are in a maze. Here are some examples:
Customer interviews are highly informative but get extremely noisy when you force the customer to imagine the product without a comparable.
Complex ideas and language that exceeds the rule of three prompt the interviewee to start disconnecting. Common symptoms are “I don't think I am your target audience” and “I imagine <some customer group> or <use case> would probably use this”. Once you are aware and overcome the communication challenge, you start seeing more excitement, interest, and actionable insights.
It is difficult to build a falsifiable hypothesis for an entry product. At AllSpark we considered a lot of options for our entry product: a health and fitness aggregator for an EHR play, an ML-based gym workout capture software, a budgeting application, a read-only portfolio viewer, an email organizer, and so on. We searched a large maze within the account sprawl pain point for a product that is blue sea and likely to capture consumer excitement.
The common advice of building an MVP allows you to know if your business has merit before you spend a lot of time and money. For us, the MVP was an incredible communication medium but did not go far to de-risk our business. A lot of the challenges around explaining the story gave way for conversations about the utility of the product.
Here are some tactical decisions we took:
Take an empirical approach (no idea is stupid until you test it) to key decisions. We have since heavily focused on product iteration that trying to spend a lot of time picking our entry product.
Create room for multiple shots on goal. At a tactical level: minimize Figma to ship time( Builder.io or Flutterflow.io for dart people) and avoid operating expenses (salary) by using flexible resources.
Capital efficiency to dampen the cost of failure.
The intent of posting this now, when the path is uncertain, is to provide our thinking while we are in the thick of it. We know we have a lot to learn, a very challenging path, and a lot to prove. We hope you stay with us on this long winding maze to put users back in control of their internet experience.
Great, tell me more about AllSpark?
As denizens of the internet ourselves, we are deeply passionate about AllSpark’s undertaking. But you have heard from us enough for one day, so let us keep the rest short and sweet. We are all dealing with our account sprawl problem and every product today has a bouncer at the door. The authentication part of our “Authentication, Automation, and Curation” value prop is to make logging in a lot less painful. In the future, we will completely remove the need for authentication through DID but for today here is the AllSpark login experience once you onboard.
Thank you for reading,
Thanks for reading AllSpark Newsletter! Subscribe for free to receive new posts and support my work.
In case you missed the previous posts: