One of the items that I think is key for a startup in the synthetic biology space — a space that has so much _future_ potential — is ensuring there could be fit for customers today and in the near future. Could the product and market make sense together and the result work as a business. Specifically, is it commercially sustainable and commercially interesting. As an aside: if it is not commercially sustainable there are other valuable ways of taking the project as research, not profit, community development; but in my case I am focused on a business as the current objective. Most importantly you need to satisfy yourself with this, but it will also quickly be essential to get others engaged and give the project its own momentum. You will need this for teammates both cofounders and any early employee. And if the business model is venture fundable (very big potential of return but with material capital risk, significant or compounding advantages) and you choose to take this kind of investment it will typically be very important
High on the list of priorities for me at this stage it is building out the reasoning and evidence for customer interest and commercial sustainability. Because I mentioned priority though, I wanted to continue the “build in the open” and share the live priorities and a prioritization technique I use inspired from a system called Kanban. I think my prioritization technique is suitable for a situation with a lot of choices, dependencies that aren’t subtle, and a lot of volume of potential tasks. Instead of a full on task list I just keep a “Top Three” items that I am doing. Keeping that max work-in-progress low helps with focus. Shockingly (for some) I don’t normally keep the full backlog of tasks in this situation. I love comprehensive and well maintained lists of things like people I am going to reach out to, but “the next most important thing” is so affected by the work as its completed and the scope of a top 3 is not normally as forgettable as an item off a grocery list. Further the cost (time and emotion) of maintaining a 50 item task list in useful condition is not worth it in some situations. I update it when activities are completed or if something changes . Here is my “top 3 priorities” live document. Weakness at the time of sharing is that these tasks are larger than I would normally want them to be — and the momentum built from small tasks is awesome.
Does it matter to customers?
Indirect Customer Research
You can start this with just overall industry numbers especially for communication, but when reasoning and modelling you generally want a more specific market as your product won’t address everyone even projecting out many years. The dramatic attention-grabbing number for the Happy Eggs and Algae idea is that there are approximately 3.5 billion eggs consumed EVERY DAY (derived from International Egg Commission (IEC) on global production). This report also highlights the largest consumers and producers and combined with the solar efficiency factor of algae lets the focus get more specific: for example, Mexico has the highest per capita consumption of eggs at about 1 per day, which is about twice the global average. Additionally, Mexico has a high solar insolation (gets more sunshine) which algae can very effectively convert to protein and has sophisticated industrial transportation and infrastructure. So Mexico is clearly one candidate location — this kind of reasoning is one example of narrowing the market to the customers that might be best addressable.
Beyond industry scale numbers you can also use indirect research to determine if the problem you are solving is likely a problem. In this case 2013 data on relative input costs suggests that feed cost is around 70% of the cost of production. This is an example where one needs to get smarter about the denominator — for example what other inputs are included or excluded here, how does this vary by region. This nuance is why it is desirable to triangulate from multiple sources if possible and know details about how the data is gathered an processed.
And even more qualitatively you can pull things like conference agendas to understand if the industry is thinking and talking about the problem space you are working on. Agendas for international egg producer conferences highlight the concerns of egg producers. For example I was curious if the consumer desire for ethical production was making the top concerns for producers — to understand if the non-animal protein movement was building momentum. And consumer trends and cage free both made the topic list. I was also curious if feedstock was a factor and grain markets and mitigations also made the topic list. And of course the primary concern facing producers is clearly avian diseases and that is also reinforced by the program.
Keep thinking about your problem space and ways of getting at how to assess and frame the problem from different perspectives.
One of the most important aspects for “does the product and market justify this being a business” question previously posed is to demonstrate what the solution would look like both to the customer and to the business. And this is where getting deep into spreadsheet modelling is key. If you can get a cofounder from finance or someone who has been an analyst they will rock this. One of the challenges here is making sure this is simple enough and nuanced enough at the same time. It is easy to get caught up in confirming a lot of little facts with low impact. And in that complexity you could fail to see the overall picture and introduce errors or miss key information or conclusions. I think a lot about accuracy and bias when working on these models and believe that as you add complexity it is increasingly possible to introduce bias of a type that is hard to see. When you do need to add the complexity I like to add ratios and checks to the sheet as a way to calibrate. For example, when doing a regional growth model that compounds month over month customer growth you can compare the result with some reasonable assumption like “10% of people who become aware of us will adopt” and then you can calculate if your growth model implicitly assumes that 200% of the regional population will become aware of you — a contradiction with reality.
Modelling is also where you make explicit how you expect the business to make money. At the earliest stage it may be desirable if there are few paths to revenue as there will be risk and uncertainty in any model. Unless your business _is_ a business model innovation it is typical to use an existing pattern. Otherwise you are diluting focus and compounding risk (multiplied uncertainty gets large quickly). My preferred approach is to demonstrate a lot of value being created, and have some reasonable options for capturing that value. If you are looking to have your project venture backed also be aware that there is a concept of “revenue quality” is also in play — this is why SaaS businesses with their recurring revenue are highly valued. In terms of build-in-the-open, I am also sharing the currently-still-a-skeleton doc to model the idea — this will get an initial round of updates transferring the rough “napkin” math and then subsequent improvements as it is refined. Feel free to follow along.
Direct Customer Research
Direct research is a huge topic that I will touch on very briefly at this time, mostly to acknowledge this will be key for me: This is because I have connections to the agriculture industry but not to the very specific customer I am proposing would be a good fit and not in the regions I am proposing. Software tech companies have a lot of tools here ranging from prototypes and focus groups and design testing, but also a real “build-measure-learn” cycle inspired by the Lean Startup and related concepts. The classic and inspiring example of Lean Startup is Zappos shoes who skipped the necessary but clear steps and focused on a quick and manual process to learn if customers would actually conduct the key uncertain step. Another example is a smoke and mirror test, most often just describing the value you expect to offer, putting that in a google ad, and then directing a user to a “coming soon, sign up here to be notified” style page. This test type captures in particular whether users believe they have the problem you are trying to solve. I am enthusiastic to try to transfer some of these lean startup, design testing, and business model testing techniques that can be done quickly to get primary data on this space.
Grappling with the business model, or “How to not be Monsanto circa 2000”
A last item that relates to using a validated type of business model. The type of distributed algae production I am thinking about so far makes some business models not viable. And I am not keen on the “aggressive licensing version” that in an extreme form led Monsanto to sue a number of customers a few decades ago. In part because I don’t think that adversarial posture is effective any more (if it ever was) and also because it is just not something I am interested in building. But yet I am keen on making this scalable, sustainable, and to grow quickly — and the best version I am aware of for that is to make the business self-sustaining and profitable. So if there are great Synbio business models out there that could apply in a distributed system with producers I would love to talk about them.
Intended takeaways about the project:
- 3.5 billion eggs a day
- 70% of operating cost is feedstock — it is important
- Producers do seem to be focused inputs and on ethical production
- Looking for input on validated business models (other than “overly aggressive licensing” of biotech in the early 2000s)
- Build-in-the-Open: You can follow the Top 3 priority tracker
- Build-in-the-Open: You can follow the modelling — this has started as a skeleton.
Intended takeaways about startups:
- Find a prioritization technique (like the Top 3 or others) to keep things moving
- Do the product and market make sense — could there be a fit.
- Closely related is “Does this justify a business” (or more generally “what is the best way to achieve this” )
- Google can take you a long way with indirect research if you try different framings and approaches to get at insights.
- Your model of customers and your business will be key in answering these questions. Complexity here is something to be careful with and of course variance and bias.
- Think about business models and generally aim to adopt not innovate here — you probably don’t want to multiply uncertainty.