It’s time to bring machine learning into the strategy room — here’s why!
Applying advanced analytics in corporate development
Over the years, advanced data analytics have transformed how many core functions operate in many companies.
Core processes in marketing and finance are leveraging machine learning and data analytics, for example, in dynamic price recommendations or accounting robot process automation.
The same, however, has not happened in strategy or corporate development … at least not yet.
While corporate development will always require creative and experienced executives to define scenarios and make thoughtful choices, advanced analytics using machine learning can give an edge in crafting strategies that lead to remarkable outcomes.
Advanced analytics can enhance strategic planning by uncovering growth opportunities that would otherwise be hard to spot — attractive markets or acquisition targets, ideas for new offerings, or even new applications for existing products.
The most effective tools use sophisticated network analysis (graph technology) and natural language processing (NLP) to parse and find connections among thousands of disparate text sources such as company descriptions, M&A data, patent filings, and academic papers.
Here are some cases I’ve been working on with a data analytics platform that highlights opportunities for any corporate development team:
- Uncover neighboring growth opportunities — clustering markets to complement traditional brainstorming methods to reveal hidden pockets of growth.
- Evaluate reputation of a target company in due diligence — identifying early-stage trends by painting what is the market or customer sentiment about a specific company you are considering to acquire or partner with.
- Assess price per employee in target acqui-hire transaction — understanding the market dynamics in a specific sector for pricing the value of an employee in a startup.
- Better allocate R&D resources by reducing bias in the decision process — betting on the right technology, based on insightful trends coming from patent and academic unstructured knowledge.
I’m going to go into some technical details in the next sections that may seem daunting for the uninitiated but bear with me, like any new approach, there is a learning curve but the rewards are well worth the effort!
Uncover neighboring growth opportunities : clustering markets
It can take a long time to translate information into valuable knowledge, especially when information growth outpaces our understanding of how to process it. Fortunately, we currently have new tools to help us augment our human limitations — tools that find hidden relationships in data.
Some of these tools use graph capabilities. They retrieve and present information so that relationships are more easily seen. When one uploads a dataset to a graph-based tool, every row of the data is treated as a node and every node is linked with other nodes based on discovered relationships.
Then, natural language analysis enables the identification of key content from a text and creates semantic relationships between them. In this case, I wanted to know how a company describes its business. So I did a content analysis of the text describing the company and then looked for relationships with other variables.
The text describing companies can be parsed and grouped into clusters (to group similarity) based on a color code, and, then, automatically rank clusters according to their size (amount of startups in this case), automatically labelling them using the most relevant terms defining a topic.
Here is an example for startups dedicated to reshaping how materials and information flow in supply chains.
In this example, I played with the “join/break” parameter to find more insightful clusters and then, created several sub-clusters and identified further granularity. For instance, in the ‘Warehousing’ subcluster one can identify the companies that address:
- Autonomous trucks and drones — from the words drone, vehicles, protect, trucks.
- Asset tracking — from the words device, tracking, GPS tracking, log.
- Warehousing operation — from the words human, video, machine, tape.
- Warehousing automation — from the words automation, warehouse, robotic, mobile.
Thanks to this technique, one can quickly identify companies that are likely to address the same market needs (nodes being closer) and short-list the players to deep-dive on for a more thorough analysis.
Evaluate reputation for a target company in due diligence
State-of-the-art artificial intelligence engines can analyze publicly available information spanning millions of web pages, news sources, tweet posts, earnings-call transcripts, and more.
By finding patterns in such disparate sources, they can help identify emerging trends by, for example, measuring the sentiment of the market or change in the frequency of a term appearing in the data.
In this example, I took all the tweets related to a proptech company to identify the sentiment of the tweets done since January 2020 and assess the reasons for the negative posts.
“Sentiment analysis” leverages trained algorithms to classify news and social-media content based on the event or topic, the companies involved, and the positive or negative sentiment associated with each company.
Quantitative investors seeking short-term market inefficiencies use this type of data to inform their trades.
Nowadays, any company developing partnerships can similarly tap such analyses for timely perspectives on a prospective partner’s customer sentiment or reputational risk.
With millions of tweets per day, information is no longer a scarce commodity; we have more of it than we know what to do with it. But relatively little of it is useful. We think we want information when we want knowledge. Advanced analytics will help you differentiate signal from noise.
Assess price per employee in target acqui-hire transaction
The most prominent technology leaders buy start-up companies whose primary motivation is to satisfy their intense demand for engineering talent in the latest emerging technologies.
This practice is called ‘acqui-hire’, a combined word of ‘acquisition’ and ‘hire’.
These transactions are structured as acquisitions, but in many cases, the buyer has little interest in acquiring the startups’ projects or assets. Instead, the buyer’s primary motivation is to hire some or all of the startup’s engineers.
After the transaction, the buyer redeploys the newly hired talent into its existing projects and jettison the startup’s projects and customers.
The so-called ‘aqcui-hiring’ has permeated to companies beyond the largest company in Silicon Valley, and companies from different geographies and sizes have adopted it with more or less success this strategy.
There is no perfect data set on acqui-hires because many of these transactions are never announced. Nevertheless, I’ve taken a stab by analyzing the relevance of artificial intelligence startups that may fall in such M&A category. I searched all the startups dedicated to artificial intelligence — in North America, Europe and Israel that had less than 50 employees, the typical startup size after a series A. From January 1st 2010 to August 1st 2021, there were 711 artificial intelligence companies acquired with less than 50 employees, representing a total of 12,850 employees!
These numbers likely underestimate the total number of acqui-hires in every period. Still, we have enough foundations to get an idea about the transactions per year taking place, whether there’s a specific field of AI driving more activity, or considering that a M&A transaction is more expensive than just poaching employees, how many employees are justifying such transactions?
Many buying companies price these deals on the basis of a per engineer on the team for an early-stage deal. One would assume that the earlier the startup, the lower the price is. Buyers might give a premium if the team has been around a longer period of time, has built some hard-to-build proprietary technology, or has some customer traction. It’s also interesting to consider if the location of the team weighs in the final price?. We can speculate all on all of this but advanced analytics should be able to get the signal out of the noise.
Subscribe here if you want to learn the full picture in my next essays.
Better allocate R&D resources by reducing bias in the decision process
The leadership team will gather several times a year to set lofty goals with the best of intentions. They hope to assess their situation, prospects and mount a decisive, coordinated response toward a shared ambition. Then, reality hits the room with everyone’s egos and agendas competing!. Bringing advanced analytics into the mix can start changing the dynamics in the room by reducing bias in the decision process.
One of the most strategic decisions relates to technological choices in the R&D roadmap and the identification of potential new applications.
By gathering publicly available information, including patents and academic publications, one can use a text-clustering algorithm and network analysis to pinpoint potential growth areas within a selection of clusters, while at the same time shortlisting acquisition prospects depending on the outcome of their build-partner-buy reflection (check out my essay on how the best companies decide their build-partner-buy programs).
Take the example of a company trying to decide which semiconductor technology to invest in. This is not a trivial decision — the capital expenditure alone would run into millions of dollars and likely lock the company into a specific technology for many years.
Using advanced analytics, one can benefit from knowing how associated trends evolve and when a specific technology is likely to have a clear advantage.
One can gain these insights by tracking patent and academic publications, announcements, and investments across different technologies.
In a world of increasing uncertainty, companies need to be dynamic in setting and managing their strategic plans. That requires combining no-regret moves that work in any condition and can be executed immediately with a few bigger, bolder bets that would be executed once the executive team is comfortable that a seductive scenario is unfolding. Using advanced analytics to track emerging trends can trigger contingent moves before the competitors do.
My last thoughts on advanced analytics for corporate development
Developing impactful strategies start with changing the dynamics in how the strategy room operates.
What is the right approach to achieve that?
In my view, diversity in the room is key.
If I borrow Philip Tetlock “Foxes vs Hedgehogs” research, a key contributor to deliver a better result in forecasting and analysis will be those that embrace the attitudes of a Fox:
- They are multidisciplinary: they incorporate ideas from different disciplines, regardless of their origin
- They are adaptable: they find a new approach — or pursue multiple approaches at the same time — if they aren’t sure the original one is working
- They are self-critical: they are willing to acknowledge mistakes and accept the blame for them
- They are tolerant of complexity: they see the universe as complicated, perhaps to the point of many fundamental problems being irresolvable or inherently unpredictable
- They are cautious: they express their predictions in probabilistic terms and qualify their opinions
- They are empirical: they rely more on observation than theory
It may be worth making sure you have a Fox in the room in this field of data applied to strategy and corporate development!
A note from the author, Albert Vazquez-Agusti: Since I was a teenager working with my father at his engineering office, I’ve seen firsthand how technology and innovation impact our work. We have reached a crucial acceleration point where technological change, education, and inequality are involved in a kind of race. I’ve come to realize that the real bottleneck to taking advantage of innovation is the lack of relevant managerial skills to impact business models through new technologies. That’s why I promote the development of people and organizations to support technology adoption to solve small to big problems based on my experience in Fortune 500, SMBs, Private Equity, Start-ups and Venture Capital organizations.
