One of the biggest hurdles for entrepreneurs is that they need to juggle short and long term planning at the same time. On the one hand, it makes no sense to worry about long-term if the short-term can’t be successful. On the other, if there is no long-term vision, we can better do something else.
In this continuous struggle between the vision and the day to day execution is when things can easily derail. Premature optimization can set a product in the path to failure. On the other hand, no long-term thinking can place the product roadmap in a dead end.
There’s a mantra I’ve seen repeated over and over that says “do things that don’t scale“, but very few people have understood what it actually means.
Paul Graham in of if his essays is perhaps the one who introduced the idea. Many people take his writings to heart, but few reflect on what thety actually mean.
There is a good example from Strava, a social network created around sports activities. Social networks are hard to scale, because the only reason for people to join is having their friends already in there. This situation creates a spiral hard to break. In its early days, Strava could work only with Garmin watches, so the founders made a bold move: they went out to buy Garmin watches and gave them as a present to friends and family. It was their approach to creating the initial user base.
Of course, giving watches for free does not seem like a good business model and, more importantly, it is not scalable.
However, the process that wouldn’t scale was user acquisition. The product they’ve built and wanted to commercialize had no scalability issues.
It is not surprising that many of the examples we find about things that don’t scale refer to the initial onboarding of users and customers. Knocking on the doors of every house to ask if they want to rent an air mattress does not scale. Having a person hand deliver shoes and ask each customer why they bought online will not scale.
Business models with a strong network effect, like Strava, or a double sided market, like airBnB, companies need to find ways to overcome the initial adoption friction.
For the companies discussed above, the things that wouldn’t scale were not core aspects of their business.
Things that really don’t scale
The history of Theranos is very relevant in this context, because it embodies the perfect example of embracing the Silicon Valley mantras while bringing them to the extreme.
Let’s tell the Theranos story from the perspective of a 20 year old who decided to follow the Silicon Valley pipe dream and dropped out of university.
Building hardware is hard. Building medical instruments is even harder. Therefore, before committing to developing new tech, you can focus on validating the market doing things that don’t scale. For Theranos, it meant performing tests using equipment from other companies.
As the market was being validated, and requirements of the users were honed in, the company could focus on building their instruments. They had a “guaranteed” product-market fit thanks to all the non-scalable things they did in the past.
That’s the story we would have been told if things would have scaled.
The fundamental flaw of Theranos was getting started from an idea that in itself wouldn’t have scaled. They had no evidence that what they claimed was possible at all. There’s a very large qualitative difference between software and hardware, and even more so between hardware and hard-tech.
Scaling the business model
Doing things that don’t scale has a different meaning when you focus on the core activities of your business. If you focus on things that can’t scale then you are in the path of failure.
Your user acquisition can scale if you just change the approach to an online form instead of knocking on doors. Shoe delivery can scale by using traditional mail instead of hand-delivering. However, there was no evidence that blood tests could scale in the way Theranos hoped for to justify their business model.
In software as a service (or SaaS) the business scaling factor seems trivial. There were little doubts regarding the ability of AirBnB to list 1 million houses on their website. I bet no one doubted that Strava could deliver their app to millions of users. Software is inherently scalable and distributable.
On the other hand, scaling product shipping businesses such as Zappos (or even Amazon) is not a given. Logistics, inventory management, and consumer demand prediction, are not necessarily scalable. They can easily be the source of issues that force companies to stop operating.
Handling scalability issues at the core of the business is where the skills of the founders to execute on strategy come at play.
For people working on the development of products instead of services, the idea of scaling is different. Scaling the production of hardware-based products is not guaranteed. The leap forward is what may become the show stopper for a small company.
Without falling too far from the Theranos example, we can use of microfluidics as a technology which is hard to scale. Microfluidic chips are broadly employed in fields such as lipid nanoparticle fabrication because they allow to use small amounts of reagents. They are great for small batch production and formulation iteration.
However, a pharma company does not care about small volumes when they go into production. Microfluidics is a technology which does not scale in itself. We could put 1000 machines in parallel, but that’s only linear scalability. Most processes need a different approach to achieve their goals.
3D printing is another example. Although it enables rapid prototyping and manufacturing, it is a process that scales linearly. Fabricating 100 pieces takes 100 times longer than making a single one.
It is important to separate whether the business proposition revolves around non-scalable processes, or whether we are using them as stepping stones to learn and validate the market.
The need to scale
All the discussion we had up to now was on the premise that business need to scale. And this is where it becomes important to distinguish between intrinsic and extrinsic needs.
Not all businesses need exponential growth to be successful.
If you look around, you’ll see many companies that generate employment, create value, and are not growing. A restaurant is a prime example. Their space is limited, the kitchen has a finite size. Once they reach a given number of servings a day, they can’t grow anymore. And there’s nothing wrong with it, lot’s of people derive a livelihood from such businesses.
On the other hand, if the business case revolves around the “winner-takes-all” mentality, scaling is the only option.
By growing our market share we can lower costs and maximize profits. If we are into manufacturing, scaling production normally means using techniques that are much cheaper but require larger volumes.
If we built our business plan around competing on price, for example, we have no way but to scale. This is true for many business-to-consumer type of products, but not for luxury items, where scarcity is a value generator.
Scaling becomes necessary when companies incur in debt, for example. Repaying with interests means that the output in the future needs to be larger than in the present. If the loan only enables a linear increase in productivity, it’ll be very hard to repay it.
Investment also comes with requirements for growth. Institutional investment (venture capital) will need a monetary return at least as large as what banks pay as interest rates. Increasing the valuation of your company comes with expanding business efficiency and market share in a non-linear way.
If your business model does not account for scalability, you’ll be in troubles sooner or later.
Bootstrapping avoids the scaling issues
If you start your business by chasing purely customer-driven growth, then you won’t need to focus on scaling exponentially.
If you grow only as much as your customers demand your products, and you built your business plan around it, then you can generate a completely different path. The overall idea is that if you can smartly leverage service contracts together with grants, you don’t need upfront investments.
You grow only if there is clear demand for what you are selling. And you don’t need an upfront injection of capital to achieve the first milestone.
Not all business ideas can afford it, though. If you are in the semiconductor industry, for example, there is no way around fabricating thousands of products at once, with the upfront costs they represent. If your business model requires competing on price, you will need to reach the economies of scale that only upfront capital allow.
Do things that can scale
If you are starting a new business, you should always be mindful of where do you want to go.
If you start with some unscalable approach, you need to develop a clear perspective on what it means in the long term. Keeping a long-term perspective is also important for your team. If the things that don’t scale become too ingrained in your products and processes, you may face show-stopper challenges later on.
Founders have to perform a balancing act between premature optimization and unscalable decisions. The trick is to find the sweet-spot somewhere in between where the path to success starts.