[2024 project] educate_product team #201
Replies: 2 comments
-
problemAlthough entrepreneurial decisions— how to test business hypotheses, when to scale, whether to develop in-house or outsource, and which processes to automate — are what every startup should make to grow, transferring knowledge and practices between startups is challenging. The root cause may be high causal density in growth as a function of founder, need, solution, customer, technology, industry effect. Moreover, these 6 factors create at least 7 dynamics between corporate strategy, customer preference, technology innovation, regulatory policy, capital market, industry structure, and business cycle, namely the gear model by Charles Fine. This dynamic web casts a multi-level challenge in generalizing knowledge on “how to grow”. First, learning in one industry cannot be directly replicated in other industries. High success rate of blitzscaling in software industry doesn’t work well [4] in hardware industry where premature scaling likely leads to failure of scaling operations in sync with market demand. Second, within one industry, iterative choices of needs and solutions, embodied as target customer and technology, affects growth trajectory. Better Place and Tesla had the same initial need (innovative electric vehicles) and solution (lithium-ion batteries) but diverged as Better Place quickly developed a unique swappable battery system requiring a car redesign, while Tesla gradually built on traditional car designs for wider technology impact (Gans et al, 2020). Third, even conditioning on every environmental factor, successful growth cannot be reproduced as we have no guarantee that all major factors are measured. For instance, belief and goal of startup members are critical for its growth but hard to measure. A founder with the goal of making emergency data available to the public may emphasize the high return of competing with incumbents to secure startup’s autonomy. However, a chief marketing officer who wants to be acqui-hired with intellectual property may annihilate the founder's effort by introducing high risks of competing with incumbents. Industry, target customer and technology, unmeasured factors challenges generalizing one startup's successful (or failed) growth. However, just as genotypes and phenotypes are mapped with usable precision in biology, despite environmental factors, unknown path and goal of evolution, and molecules within a base, we can try to map startup growth with its action sequences. We start by defining the smallest observable action as the startup base. This means the underlying mechanism of unmeasured belief and goal affecting startup’s belief and how that belief affects the action sequence are abstracted. When four types of startup base (A, G, C, T) form a sequence and produce successful (or failed) growth, we name this sequence as growth gene. Based on this definition of base and gene, we aspire to construct Startup Database 🧬. This database can further facilitate the cross-industry learning process of growth (Industry Map 🗺) and enable early-stage startups to navigate their growth (Startup Compass 🧭). solutionRegarding
Abdullah commented: Angie worked on:
B (doing)
C (todo) |
Beta Was this translation helpful? Give feedback.
-
startup it's interface to |
Beta Was this translation helpful? Give feedback.
-
Bold from below collaboration plan, in parallel with #202
The MIT PI Charles Fine and coPI Abdullah Almaatouq have built two parallel teams to combine theory and practice for this project. The theory team (Abdullah, Angie) will have biweekly meetings to enhance technical feasibility of the GPS such as developing computational statistical models that decompose effects of team, process, task, industry from the log of past startups. These models can help test generalizability of existing theories. Further experiments can be designed using integrative experiment design which contributes to generate cumulative empirical and theoretical knowledge (Almaatouq et al (2024)). Concurrently, the practice team (Charlie, Doug, Angie, Anup) will have biweekly meetings to enhance operational feasibility. This involves translating statistical evidence from the theory team into actionable business insights and continuously integrating insights into custom tools, and discussing the architecture for better collaboration with Google to use its data and computational resources. MVP implemented on MIT digital entrepreneurship platform and its enthusiastic reception from our beta users (5,000) shows potential of our product. Based on this pipeline, GPS can be delivered to startups to help them craft business plans in a fraction of the traditional time.
Beta Was this translation helpful? Give feedback.
All reactions