Growth Strategy: Generative AI Immediate Applications

The use of generative AI can, right now (at this very moment!), unlock both speed and production improvement in strategy work, accelerating strategy and M&A decisions, improving the quality of inputs into those decisions and, through reduced cost, increase the quantity of decisions receiving the right amount of research support.

What is Generative AI? Generative AI relies on deep learning models to learn how to understand written text requests and then produce an output from a large dataset (for example, a snapshot of all data on the internet) to meet that request. This AI is not actually “thinking”; it is more of a highly complex data search and organizational tool that sifts text, images, video, etc to serve up the user’s request. As a result, many early adopters of Gen AI have failed to acknowledge its limitations, including:

• It cannot analyze content and develop insights

• It is notoriously bad at math: it can’t be used to develop and run analytical models

• Most importantly, while it has access to vast quantities of information, it needs to be taught how to use the information the way the user needs it used

In short: Gen AI can’t just hand you the answer to your market analysis question. It can’t even produce the insights you need to develop your own answers off of a single question prompt.

However, if we recognize its limitations and build a process that leverages what it is good at (i.e., searching vast quantities of data to provide requested information in seconds) by pointing it specifically at the foundational research effort, then we can create quite a bit of value. 

Growth strategy work- including market sizing, competitive landscape analysis,  customer analysis and assortment/service offering- generally encompasses X phases of work:

1. Scope definition (defining the decisions that need to be made and what questions we need to answer?)

2. Foundational research (surfacing the information we need)

3. Insights development (turning that information into insights)

4. Options development (what routes can we select from with pros/cons for each)

5. Decision and execution roadmap

Gen AI is ill suited to assist with 4 out of the 5 above steps. Scoping, insights, options and decisioning all require a combination of qualitative and quantitative analytical ability mixed with contextual knowledge that make the brain of a professional expert vastly superior to that of an AI. 

However, where is most of our time spent and what is the bottleneck on such growth projects? It’s that 5th phase, the one Gen AI is fantastic at: foundational research. Organic and inorganic growth work usually requires developing up to date insights on customers, competitors, products/services and technology often in a space that is relatively new to the company. As a result, someone has to go out and collect the information needed to inform insights and options.

Part of that can come from syndicated research: the team goes and buys a report at $5k-$10k. These reports are never enough, though, as they are general overviews and not written in a bespoke manner to address the team’s specific challenge. As a result, the team must invest weeks of expensive talent on searching for and reading articles, websites, 10ks etc. and then more time in organizing that information, vetting it and curating it.

 Enter Gen AI. With the right Gen AI agents, we can produce a more complete and robust set of on target research in 30 seconds than a strategy team can produce in weeks and have that work summarized for expert review in a standard format, quickly surfacing the growth strategy research we need, including:

• Market sizing

• Market trends/innovations

• Product/service segments and sizing

• Customer segments and sizing

• Competitor segments and profiles

• Competitive dynamics, including how competitors decide to switch providers

• Risks and threats

Why aren’t more organizations using Gen AI in this manner? First, the standard large language models are not set up to address this sort of research effort right out of the box. One needs an AI agent that has been trained, over many similar requests, to know what sort of information we need, from what sort of sources and presented in what manner. 

Secondly, to get the research results we are looking for, we actually need multiple such agents, each handling a discrete part of the research question, often handing its output off to another agent to build on. In this manner, one creates chains of agents.

Thirdly, one still needs to have a growth strategy expert to review the AI output, rerun AI work as needed, identify new questions to ask based on early research and serve up the preliminary insights. 

Building the above requires expertise and an investment of time and resources. Strategy organizations- whether in house or at a firm- are already strapped for resources (which is why they need Gen AI in the first place!) and normally don’t have the capacity to spare to develop such a capability. However, the technology is there and it works. At Chronos Insights, our consultant led-AI powered process can produce robust, reviewed and curated research of the above topics in a single week… something that in my experience normally takes a month or longer. 

Change is coming. Organizations that figure out how to leverage Gen AI will drive faster, better informed decisions resulting a distinct and powerful competitive advantage.