Table 5.7 New ventures and startups New Ventures
Business model Management Customers Action plan
Landscape: Capabilities Segments: Refine the business model - Value network Finances - Characteristics Engage with customers
- Competition Current - Potentials Milestones:
- Success factors Portfolio: - Trends - Fund gathering - Barriers to entry - Potentials Innovations: - Budgeting Entry strategy - Differentiation - Technologies - Partnership Operations - Value chain - Product features - Control systems
Pricing Fit - Business models
P&L forecasts Market sizing
1. Business model
• What customer jobs to be done, do the client’s products address? What prob- lems are they trying to solve?
• How will the client compete in the market place? (differentiation, cost leader- ship, focus)
• How did market shares for key players evolve recently, what are the key suc- cess factors?
• What are the key barriers to entry? (e.g. resources, regulations, access to dis- tribution, IP)
• What are the advantages/disadvantages of starting from scratch vs. acquisi- tion vs. joint venture?
• How will the client operate its value chain (procurement, processing, delivery, marketing)
• How will the client price the product?
2. Management
• What are the accreditations and competencies of the management team?
• How is the client doing financially? Are there cash reserves?
• What capital structure (debt vs. equity) and allocation of funds may be considered?
71
• What is the client’s current product portfolio? Which products have the most potential?
• How does the client currently differentiate from competition?
• Will the current operational workflow benefit in/from the new business?
• How will the new business fit with the rest of the product portfolio? (synergy vs. cannibalization)
• Forecast costs and revenues to compute the expected ROI for different time horizons
3. Customers
• Define customer segments (demo-/psycho-graphics, jobs); Which will likely be the most profitable?
• Can we find innovative ideas in-/out-side the organization that have potential for the business?
• What is the total market size? Can we identify potential non-consumption or currently over-served market opportunities? (e.g. disruptive innovations [18]
or blue oceans [77], see Chap. 8 for more details) 4. Action plan
The above analysis delivers a better understanding of customers, competitors, innovators, and the overall potential of the client’s new venture. Based on these insights, the assignment can then advance to a refinement and implementation stage:
integrate new features in the business model by reviewing in turn the resources, processes and values, engage with customers through surveys and promotional campaigns (e.g. free trials, demonstrations) and articulate milestones that will bol- ster the integration of the value proposition in the overall industry value network.
5.8 New Ventures and Startups
73
© Springer International Publishing AG, part of Springer Nature 2018 J. D. Curuksu, Data Driven, Management for Professionals, https://doi.org/10.1007/978-3-319-70229-2_6
6
Principles of Data Science: Primer
Let us face it. Statistics and mathematics deter almost everyone except the ones who choose to specialize in it. If you kept reading and reached this far in the book you are probably now considering skipping the chapters on Data Science and moving on to the next on Strategy because, well, it sounds more exciting. Thus, let us start this chapter on statistics by a simple example that illustrates why it is worth reading and why consultants may increasingly use mathematics.
Suppose you gathered information on the demographics and psychographics of thousands of customers with extensive surveys. Suppose the data includes numeri- cal variables such as age, income and mortgages, plus categorical variables such as education, health problems and travel preferences. Your client provided you with all transaction records made by credit card over the past 5 years. In order to think about your client’s growth options, you would like to answer a simple question: who are the best customers? The [big] data is here, you have intelligence on tens of thou- sands of transactions themselves associated with particular income levels, emo- tional preferences, etc. You could barely hope for more data. But you start to wonder… what does “best customer” mean? Seemingly the ones who purchased the most in the database are potential candidates. But in your effort to understand these customers, you start to further wonder what are the characteristics of these custom- ers? What features correlate or associate with purchase levels? Does education background for example have anything to do with purchase levels? Maybe not, but in the category of high-income customers, now, does it? Can one make better predictions by using only the income levels, or will a combination of income levels and education background deliver better results? At the end of the day, what subset of features might best predict purchase levels and represent a good starting point for your team to brainstorm strategic options? How confident may you be in all these predictions?
The above problem is as simple as it is important – it asks: what features have something to do with purchase and by how much. But this problem is impossible to solve mentally. Some data are missing (purchases made in cash), some data cannot be directly compared (continuous variables versus binary yes/no answers), and
74
some data will turn irrelevant. And here is why data science is worth learning. With the mathematical framework of machine learning [89] introduced in Sect 6.2, if there exists a set of say three features in this database that correlate with purchase levels, with degree of confidence >95% (p-value <0.05), it will be possible to find these three features within just a few minutes on a computer. This is what data sci- ence has to offer. Once one understands the basics, what no rational thinking could ever hope to accomplish even in a lifetime may be delivered by a computer within the hour.
This Primer
This primer is not an introduction to statistics, which would require at least an entire book1. The goal of the data science chapters is to overview the basic mathematical tools and concepts that may be used in business management, some typical work- flows, and some applications. Common sense and focus are certainly required in these chapters at least as much as in any other. But armed with this commitment (of focus—we will assume everyone possesses common sense…), there is indeed noth- ing in these chapters that shall be out of reach intellectually. It shall, however, increase your confidence when you analyze data in the future.
A typical data analysis project starts with exploring and cleaning the data, then moves on developing theories to interpret the data, and ends with communicating takeaways and/or predictions based on intelligence gathered from the data. What happens next, be it either discussions with stakeholders or applications of the model to new data, both represent an ultimate refinement phase where theory meets prac- tice, where the model interacts further with real-world data. Indeed, acceptance vs.
resistance from stakeholders and success vs. failure of an application to new data, may both lead to valuable knowledge and refinement. This real-world feedback is increasingly leveraged in most data analysis projects due to the emergence of big data and the recent revolution in the economic of information, as discussed later in this chapter.
A note on equations
In the data science chapters the key equations are provided because, as the popular saying has it for pictures and graphs, an equation may be worth a thousand words.
But each equation has been stripped to its minimum formulation, and no technical background is assumed. Thus with a fair amount of focus and commitment, the fol- lowing material is accessible to all readers irrespective of their left-brain/right-brain inclination. An exception might be the repeated use of the integral symbol ∫ where the reader could expect the sum symbol ∑, so let us get that out of the way: ∫ and ∑ are completely equivalent for the purpose of this discussion and ∫ was chosen only because it is visually more elegant than ∑ when repeated multiple times. A formal difference exists (∫ applies to continuous variables, ∑ to discrete variables) but is irrelevant for the purpose of an introduction to data science.
1 One may recommend “Naked Statistics” from Charles Wheelan [89], which introduces the overall field of statistics in a simple and humoristic way …technical expertise not required.
6 Principles of Data Science: Primer