🛠️ Ventures

While the final three stages have become standard practice for startups and accelerators, the initial two stages are experimental in nature and are believed to be novel in the context of venture building.

The Two Bit venture builder methodology revolves around the idea of early team building in the service of a common goal — winning prizes in the context of data science competitions — in order to first assess technical viability and group dynamics in solving real world problems as a prelude to new venture formation.

The Two Bit approach has the advantage of eliminating technology risks and people risks on the get go — at the same time reducing market risks — prior to feeding each project into a conventional deal pipeline that connects with the capital market. For a highly technical field such as machine learning that depends heavily upon the actual skills of its practitioners, such a mechanism for risk mitigation makes perfect sense in ensuring deal quality downstream feeding into traditional venture capital.

A key idea of the venture builder model is that collective talent from its community of experts will significantly increase the chances of new venture success. A systematic way of doing things, plus a learning and building methodology, lies at the heart of what a venture builder does.

Take Two Bit for example, a wide range of prize-winning 🏆 machine learning applications can be readily created and deployed by Two Bit community 🚀 members based upon ongoing curation of a data commons 🌞 and the harnessing of open source technologies 🛠 for D2O (i.e., DevOps + DataOps). This growing base of code and data is made accessible to the Two Bit community in the form of Common Ground 🌐 Web Services, upon which ongoing improvements and future services can be constructed.

Over a longer time horizon, new ventures in the 5-stage pipeline are expected to evolve and adapt to the surrounding economic environment by way of an iterative process featuring 3 components: (1) variation, (2) selection, and (3) retention — otherwise known as universal Darwinism.

The overall mechanism works by trial-and-error alongside generate-and-test. Therefore, by the logic of natural selection, the genetic pattern of new ventures given sufficient trials will evolve towards more adapted forms better suited for survival in the marketplace.

Think a school of fish — not an assembly line nor a production studio. The Two Bit venture model tweaks each of the VSR components in the evolutionary process in response to capital market feedback so as to grow bigger fish that also knows a great deal about how to survive in the open ocean.

In recent years, the lists of venture builders, also called tech studios, venture studios, startup studios or startup foundries or startup factories, have continued to grow (e.g., curated lists here, here, and here). Where different venture builders diverge most among one another is how they identify business ideas, e.g., gut feeling, cloning, or methodology. Two Bit solves real world data science problems competitively to win prizes and to identify opportunities for new ventures.

From a Two Bit perspective, the venture building exercise is simply real world problem solving meets design thinking. Therefore, it makes sense to apply a methodical approach to the process of building new ventures — systematically exploring a design space comprising of a grid of 4 factors of venture production x 5 venture stages upon which a characteristic set of venture building activities can be overlaid flexibly and adaptively for each venture in the pipeline.