#no-stack startups

In my previous post, I combined several concepts into what I call the ”No Stack Workweek.” The name comes from an earlier concept called the ”No Stack Startup,” but I wanted to make a distinction between a startup (which is typically growth-oriented and often VC-funded) and a cash-cow business that enables a ”4 Hour Workweek” (a concept developed by Tim Ferriss in his book of the same name). Check out that previous post if you’re interested in how I came up with my combination of these concepts.

Building a business is kind of like juggling and driving a unicycle at the same time – an interplay of multiple stakeholders and the machinery of processes (which are sometimes invisible) that enables the simultaneous delivery and capture of value.

Building this machinery is usually expensive and time consuming. An entire movement called “lean startup” is trying to explore how to minimize up-front building. The goal is to ensure that if a business doesn’t take off as envisioned, the waste of resources is minimal. However, lean-startup thinking usually isn’t focused on the lower-level engineering-related steps one could take to minimize the build process. Rather, it's focused on avoiding building at all. It utilizes techniques such as using landing pages to test interest (smoke testing) or delivering value manually (concierge MVP) before building anything.

Additionally, discovering lucrative business opportunities and executing them is hard. It is not easy to deliver and capture value in real markets of real people, especially in a niche that's accessible to a new player within a given window in time. If it were easy, everyone would be an entrepreneur working for themselves.

The “No Stack Way” reduces the need to invest in basic infrastructure

What I like in the “No Stack Startup” approach is that you pay other companies to maintain APIs that create value for both you and your end-customer. That way, you don’t have to care about stuff other companies can do for you, stuff that is getting quickly commoditized, anyway. The easy examples include Amazon AWS or Heroku for handling server infrastructure, Twilio for SMS and voice infrastructure, Stripe for credit card processing online, or Square at the point-of-sale.

Even before anyone used the term “startup,” banks had always been securing financial transactions for new businesses, accountants and lawyers had been handling the paperwork, and so on. The difference today is that you can use programmatic interfaces (APIs) to automate all of those tasks so that no humans need to be involved in repetitive “infrastructuresque” chores. These tasks are necessary, yet they are commodities. That is, they are hygiene parts of the value creation/capture process that produce little or no competitive edge.

So how could we build an entire business almost entirely using other companies’ APIs to assemble a value-creation-and-capture machine that requires a minimal amount of coding effort?

How would you deliver a package via APIs?

Beyond using the obvious APIs from companies like Amazon AWS, Twilio and Stripe, at first I draw a blank when it comes to whole end-to-end architectures.

Let’s say we’d want to automate end-to-end delivery of postal packages. I'm not saying that would be a great business opportunity at this point, but let's assess whether or not we could do it the "no stack way." Do we have a company that offers an API without minimum volume limitations or a requirement for long-term contracts to ship packages? Does that API include tracking of packages to make the user experience smooth? How would I know?

When I search for “package delivery API,” the first results include a company called EasyPost, which seems to be exactly what I’m looking for. The search also returns multiple articles from Programmable Web (PW), which is a site specializing in listing APIs, and 92 of them are related to shipping. However, PW’s listings are very qualitative, and comparing the APIs seems to be somewhat “apples and oranges.” It would require quite some investigation effort to determine which of the 92 APIs actually are suitable for our hypothetical business. The same goes for EasyPost. How do I know if it is actually a good service for this use case?

A playbook for “no stack” businesses?

What if there were a playbook of tested-out “plays” of different stacks working together? More than just a list of APIs, such a playbook would also include some thoughts on how the APIs fit together, how pleasurable they are to use, and how their pricing works for very small businesses. PW could be a great source of discovery of new APIs, but the value of a playbook would be in more than just the technical aspects of an API and a technical proof of concept. As they say, the devil is in the details, and only by testing a service (not just the API) in the context of an actual business process can you learn enough about its properties.

This would obviously require lifting the kimono, as they say, regarding the business and technical architecture of a company, which is something entrepreneurs often aren’t willing to do. Often when we read about the “stacks” of companies such as Uber (https://eng.uber.com/tech-stack-part-one/), for example, the descriptions are so high-level that they don’t easily allow us to understand exactly how the different pieces have been glued together and how well they actually work.

Hardware DIY sites already provide playbooks for DIY enthusiasts

An analogue here, much more fitting than PW, could be hardware DIY project sites such as Hackster.io (https://www.hackster.io/arduino/projects) or DIYhacking.com (http://diyhacking.com/diy-projects/arduino-projects/).

Both of these DIY sites have a familiar format for project descriptions that start with an interesting use case (an analog for our concept would be a business case). The sites then go on to show what parts are needed (a list of APIs in our case) and demonstrate how the project was completed. Most importantly, the sites discuss how the different parts were combined, including what needs to be taken into account for the components to work well together. Such a discussion might also include the different alternatives that were tried in the process, and why the choice was made to use a specific component instead of another in the same category. This sort of discussion is what’s missing from PW.

Making it as easy as drag-and-drop

The playbook could even use a visualization technique similar to popular drag-and-drop programming environments such as MIT’s Scratch (https://scratch.mit.edu/), Google’s Blockly (https://developers.google.com/blockly/) or Berkeley Snap! (http://snap.berkeley.edu/). By creating connector nodes similar to the way that Zapier (https://zapier.com/) works, an integration of two APIs would need to be created only once, and afterwards people using those APIs together could avoid programming entirely.

This combination of a business case and a technical end-to-end implementation guide (or even prefabricated sample projects) would make assessing the suitability of APIs much easier, since instead of reading through the sales materials of the API provider, one could directly get a first-hand account of how the API actually worked in a real-life scenario.

Playbooks for reducing business “boilerplate” and increasing innovation

This way, when building a new business, one could more easily discover new APIs suitable for the situation, as well as get help in the actual implementation. Most of the startup's work could then go into the unique aspects of the business, instead of the setup of what is known in programming as “boilerplate."

When reading other peoples’ showcases, it’s much easier to be “creative” when you have stencils that you can combine, instead of facing a blank slate. Since transferring a business model that's working in one geographic market to another market is already the business of many companies, there isn’t really anything dramatic in this kind of copying and pasting, except that the actual implementation would be much cheaper. Many business ideas that would otherwise be left in the heads of up-and-coming entrepreneurs due to lack of build resources might now get built and delivered to market, and that would benefit the end-customer.

Do you have a budding idea for a combination of APIs that could form an end-to-end business already? Do you feel like sharing it?

** UPDATE: Based on initial feedback to this post, I’d like to clarify that when I’m talking about tech entrepreneurship and the “no stack startup” approach, I specifically am not talking about creating the next Google or Microsoft (heavy lifting kind of tech) but instead gluing together existing pieces of services via API’s and focusing on the last-mile UX (this is my understanding of what the “no stack startup” means). So think Amazon or Zappos (in their early days, with their little-to-no inventory approach to e-Commerce, by leveraging partners in fulfillment) type of tech-enabled service business rather than the Curious AI company or Google type of tech business. **

A relative of mine told me when having a conversation about work: ”I’m working so hard and getting paid only marginally more than when I was doing less demanding work – perhaps I should start working on a startup instead..”

As a person with some startup experience, I responded with my view on the expected value (average of different payoffs in different scenarios weighed by their relative probability) of a startup: most likely lower than that of a 9-to-5 job, due to the high failure rate, convention of paying lower-than-market rate salaries even when VC funded as well as long hours anyway.

But then we spoke about another relative who is making a decent living by buying a few used expensive items abroad per month, and importing them for a profit. Since this relative already has a committed buyer (the demand) before buying anything, the income is relatively secure.

“Startups try to search for a profitable but scalable business model while being unprofitable for a long time”

We agreed with the first relative that if the latter relative would attempt to scale this operation (with the help of, say, VC funding), most likely the profitability would quickly decline as the ”good deals” on the market are quite rare, and profiting only possible when finding rare gems (or pricing errors) on the market.

So here we have a profitable but not scalable business model. Startups try to search for a profitable but scalable business model while being unprofitable for a long time – in the hopes of becoming profitable with scale, helped by e.g. digital technology and/or leveraging crowds (such as AirBnB and Uber). For every Uber and AirBnB there are a hundred other very similar companies that we never hear of, but still burn through thousands, or millions of euros or dollars of their investors’ money.

the ”good deals” on the market are quite rare

So I suggested to my relative that he check out Timothy Ferriss’ book ”The 4 Hour Workweek” (sometimes shortened to 4HWW) to think about a business that requires very little risk or maintenance but would give some extra ”passive income”. However, my personal experience about the 4HWW model is that in practice finding a business that is both profitable and low-maintenance is much harder than how I feel Ferriss makes it sound. So I let the whole thing slide – until..

Today everything seemed to click. I’ve recently read many very interesting blog posts about aspects that today just came together when talking to a friend who is an experienced data scientist and enterprise consultant. He gave a talk about how ”software is eating the world” (a phrase attributed to tech entrepreneur and investor Marc Andreessen) and how one can build tech solutions with cloud software 10x faster than just 3-4 years ago thanks to both advances in open source software as well as cloud providers’ commercial efforts (Google Cloud, Amazon AWS, Microsoft Azure, etc.) to make using these software packages as easy as possible without configuration.

one central capability for a ”smart creative” is the ”art of the possible”

My data scientist friend also referred to Eric Schmidt’s ”How Google Works” presentation and ”smart creatives” working in a near-chaotic way on wicked problems with a large degree of freedom, as up-front planning doesn’t help in these scenarios. My friend added that one central capability for a ”smart creative” is the ”art of the possible” (as an actual concept it seems to refer to political pragmatism, but that’s a coincidence here), with which my friend meant that you need to have a rough understanding of what is possible in order to be able to come up with groundbreaking innovations leveraging existing technology in new ways as well as building small pieces of new technology to make these new combinations possible.

An example (I’m making this up since we didn’t discuss this in detail) would be using the accelerometer of a smartphone to do sleep tracking, and thus building on a technology platform provided by Apple and Google, instead of having to build expensive dedicated hardware, as well as leveraging the App Store and Google Play Stores for distribution. In this way a small team can make a decent amount of money with quite a niche product without needing much upfront investment. Still, for someone to be able to come up with such a business requires mastery of the ”art of the possible” in the domain of mobile technology and sleep monitoring science.

“..it feels like it is an amazing time to be a creative entrepreneur utilizing emerging no stack techniques..”

So how could one take this notion forward? Earlier, I bumped into a Slideshare presentation by VC Albert Wenger quoting Andy Weissman’s blog post called ”No Stack Startups” (http://blog.aweissman.com/2015/05/no-stack-startups.html) where Weissman concludes with, among other things that ”it feels like it is an amazing time to be a creative entrepreneur utilizing emerging no stack techniques…”. By ”no stack startups” Weissman means ”instead of being good at many things, companies can just focus on the last mile of value they provide, the one thing they can excel at better than anyone else. Maybe this should be called the No Stack Startup - services that can focus on doing only one thing - hopefully well - and utilize other services for everything else.”

Yesterday I read Jon Lax’s Medium post ”Great Products Don’t Happen By Accident – Using playbooks for designing and building products” (https://medium.com/great-products-dont-happen-by-accident/great-products-dont-happen-by-accident-f46323d8ad94). 

“let’s run Statue of Liberty Buck Sweep”

Lax suggests building a ”playbook”, or documenting ”plays” similar to how NFL coaches do so that they have a set of adaptable responses to a set of reoccurring circumstances, which have been rehearsed with the whole team so that the team has common language for a very specific maneuver so that when the coach says “let’s run Statue of Liberty Buck Sweep”, everyone knows immediately what to do. 

After all, a lot of time in startups and big companies alike is both wasted on re-inventing the wheel as well as following preset processes that don’t match the current situation. Why wouldn’t startups or corporate product development departments document a set of ”plays” which would create unambiguous shared language as well as be reused while retaining flexibility in a changing landscape that still share some commonalities? Kind of like ”process 2.0 for product development”.

The blog posts that have been marinating in the back of my head combined with what my friend said today, sparked the question that could we do ”No Stack Startups” (meaning providing only the last mile user experience with low technological risk by using other people’s platforms and products via API’s) in a systematic way (hence, the ”playbook” thinking) while not requiring unicorn scale and VC investment?

“..so cheap that it wouldn’t have to have any kind of scale to provide a very nice living for its solo founder or small founding team..”

Let’s say instead of trying to have everyone and their uncle become startup entrepreneurs, could technology entrepreneurship (if we remove the requirement of scaling the business) be something that is done as a side job?

A ”no stack startup” could, at least theoretically, built for so cheap that it wouldn’t have to have any kind of scale to provide a very nice living for its solo founder or small founding team. In this way it would end up being very similar to what Ferriss is proposing in ”the 4 Hour Workweek”.

In that way, a person could slowly and with low risk, build entrepreneurial capability that could then be either used in intrapreneurial activities inside of big corporations (not all people, after all, hate working for big companies), for serving lifestyle entrepreneurial aspirations (say, serving a market which traditionally isn’t seen at all attractive by tech entrepreneurs), or to eventually go on to be a founder of a tech startup unicorn, depending on the person’s risk aversion profile.

If the answer is ”yes” or ”maybe”, the next question is: what would be required for this kind of approach to work? That is where I believe ”the art of the possible” (as defined by my friend) comes in. One needs to know what relevant existing technologies there are, how to leverage them in the given domain, as well as some basic marketing/finance capabilities so as to get the product to market in a way that the market can accept (UX, pricing, support, etc.) without overspending so as to require massive injections of investment capital and thus bring in demands of five to twenty X exit multiples the venture capital investors require.

And again, to keep the costs down, this would need to be done by a very small team or a ”one (wo)man band” of ”smart creatives”, which places certain requirements on who these people can be. Since typically the people with the ”art of the possible” covered are rarely design, marketing and finance experts at once, the next question seems to be: are there some aspects to the process that are often time similar that they could be distilled into ”plays” to form a ”playbook”?

After reading the blog post about product development ”playbooks”, I think yes. And I think the same goes for many go-to-market activities. Especially if there is a community of these ”no stack startup” entrepreneurs working together. I think even prototypes or blueprints of different business model ”plays” could be distilled, which is something Alexander Osterwalder has tried to do by coming up with the Business Model Canvas to compare different established businesses’ business models, and Ash Maurya taking this closer to the problem areas of startups with the Lean Canvas.

Where could we get the first brave ”smart creatives” wielding the ”art of the possible” while simultaneously supporting them financially to have time to learn from initial mistakes, as well as distilling the learnings into a ”playbook” so that people coming after the pioneers could ”stand on the shoulders of giants” instead of starting from scratch every time?

After all, if cloud technology (Amazon AWS et al) has made it possible to cut down on software development costs up to 10X or even more, we might be possible to come up with social technology to cut down on the ”trial and error costs” in a similar way to make searching for these niche businesses much cheaper and thus attractive to rookie entrepreneurs with no previous successes under their belts giving them creditability and providing them either money in the bank and/or access to investor financing.

If the challenge is to encourage low risk, moderate reward part-time tech entrepreneurship, what would your playbook look like?