This article was a request from one of our Catalyst members. The Catalyst Advisory Program is a coaching program involving one-to-one advisory, weekly group Q-and-A with other Catalyst members, and a series of proprietary resources and frameworks based on insights from AI leaders at the world’s largest enterprises (AI adopters and buyers), and successful AI vendors and service providers. Learn more and apply emerj.com/catalyst.
In a previous Emerj Plus article, I went into detail about how to find the companies most likely to have strong budgets for AI projects. In that article, I discussed specific industries that are adopting AI more than others, and I explore why larger firms (companies with well over $1 billion in revenues) are the strongest candidates for buying AI products and services, or having strong in-house AI initiatives.
One of our Catalyst members asked me about what smaller firms (companies under $1 billion in revenue) might be able to buy. Like most of our Catalyst group, this member runs a small to mid-size IT and AI consulting firm, but what I’ll be covering in this article will be important for both buyers and sellers:
For AI Vendors and Consultants: You’ll learn more about how to deliver value (and win deals) with companies who may not have the budget to engage in robust, custom AI projects, but still have budget for executive education, and can grow into their first AI projects with you.
For Leaders in Smaller Enterprises: You’ll learn about some of the right first steps to take with AI vendors and consultants – even before you hire data science staff or engage in larger AI projects. Some of these steps are particularly important for Executive AI Fluency, which should be the first step in the AI journey of any enterprise.
Selling AI Services to Firms Under $1B in Revenues
Selling AI services often requires the client (typically a large enterprise) having a certain amount of AI understanding, a certain amount of in-house data, science talent, a certain amount of R &D budget, and a stomach for risk. When it comes to offering services and determining first AI steps for companies under 1 billion in revenue, we need to take some important factors into consideration.
1. Lack of In-House Data Science Talent
Established companies (not fast-growing tech startups) under a billion dollars in revenue often have no genuine in-house data science talent. They might imply or claim they do, but they generally have none at all.
At some point in the last four years, we’ve seen an increasing number of companies at around $300MM in revenues start labeling some of their staff “data scientists.” Sometimes these are folks with some middling data science background, sometimes they’re simply developers or IT working on some nominal AI project with a new cool-sounding title, but no actual skill.
I can’t blame companies for doing this, AI is “cool” and someone in leadership had some kind of toy AI project in mind and decided to give the task force cool-sounding titles. That said, small, misguided projects and terrible data infrastructure will never attract or retain strong data science talent, and this is where most smaller enterprises remain stuck.
Real in-house data science talent means academically credentialed or business experienced people, preferably from marquee tech companies doing AI. This would be people with hands-on experience with data science who are also involved in actual projects. In smaller firms, they are not going to have that talent.
What this means: We need to prepared to educate leadership on fundamental AI concepts, and we need to aim for projects that won’t require in-house data science talent (or that will require minimal in-house talent). For vendors and consultants, this opens up an opportunity to provide educational services and strategic guidance until the client is in a position talent-wise to clean up its data infrastructure and move into more significant technical AI projects.
2. Very Little Executive Understanding of AI
Smaller firms also tend to have very little executive understanding of artificial intelligence and of AI use cases. This might also be true for much larger companies, but because smaller companies often have has less experience doing AI even at the level of pilot projects or having legitimate in-house AI talent, executives will have even less access to the fundamental understandings about AI.
As I’ve mentioned in past articles and in social posts (see below), executives rarely focus on AI fluency education on their own. Rather, they gain this fluency only through interactions with real-world initiatives. Conversations with vendors, feedback on real projects in their business, and conversations with other in-house talent working on real problems – it’s these factors that make executives AI fluent. With no or few past AI projects, a company is likely to have relatively little understanding of AI and AI use-cases in the C-suite.
In other words, with all things being equal, between the CTO of an insurance company that does $10 billion in revenue and one that does $200 million in revenue, the CTO of the company doing $200 million in revenue would likely have less exposure to the conceptual understanding of AI and its viable use cases for the reasons stated above and also because of another factor, which is very little research and development budget to test or deploy AI solutions.
What this means: Basic AI concepts (such as how algorithms work at a conceptual level, or why data needs to be harmonized), and fundamental AI deployment considerations (importance of feature engineering, etc) will need to be translated to leadership before significant AI projects should be undertaken. For vendors, this again opens up an opportunity for AI education and strategy as a first step in “setting the table” for future services.
3. Very Little R&D Budget to Test or Deploy AI Solutions
It is just a given that a $10 billion company is going to have some serious money kicking around somewhere to potentially work on their AI readiness, try pilot projects, and work with vendor companies. They have the resources to stomach some risk and try things, so this is going to educate functional and technical leaders of that company. It is otherwise with smaller firms that might not have the resources to have much to do with research and development activities.
All these factors limit the ability of smaller firms to sell AI services to big enterprises. Less in-house real data science talent, less executive understanding, and fewer opportunities to do research and development are the realities preventing these companies from executing or deploying artificial intelligence projects in a big way.
However, that does not mean that those working in smaller firms are not as smart as those that work in larger firms, or that they are incapable of gaining the same level of understanding and experience. They might just need to go about it using a different route.
What to watch out for: We need to be extra careful of “toy” AI projects – projects done for no reason other than saying “we did AI.” Firms with limited budgets (who feel pressured to “do something” with AI) are often pulled into this trap, and spend without any perspective on building AI maturity or reaching towards strategic goals.
What to do: Consider executive education of paramount importance. If we move forward into AI projects – consider smaller projects that will (a) not involve robust in-house AI talent (i.e. use APIs or similarly accessible tools, rather than intensive, custom-built solutions or more complex vendor solutions), and (b) will develop some important elements of AI readiness (see Emerj’s eight Critical Capabilities).
Selling AI Services to Smaller Enterprises – Leading with Education
Selling AI services to smaller enterprises will almost always involve leading with executive education, rather than with robust AI development projects. “Education” typically involves:
- Familiarizing the client with how AI works, and with a range of relevant AI use-cases.
- Opening the client’s eyes to AI potential – giving leader the ability to see connections and determine potentially valuable projects themselves.
Only when leadership has a realistic (not overly optimistic, not overly pessimistic) perspective on AI itself, and only when they can connect the dots of what AI can do in their own business – can initial AI projects or AI maturity efforts be brainstormed, decided upon, and executed.
The Gradient of AI Project Complexity
The benefit of educating functional and technical leaders in AI is it often serves as a platform to sell other services to smaller companies. It pays to think of selling AI services to smaller enterprises along a gradient.
AI in SaaS: One way to do this is bake AI into software where it is working in the background and the user barely knows or might not even know it is there. Most people are familiar with the term “AI-powered” as part of software, and may vaguely understand that it is really an AI deployment, but has no need for building skills in-house. It is light exposure to AI in a way that doesn’t really require any extra effort on the part of the user.
AI APIs and Tools: The next part of the gradient with a slightly deeper level of user involvement is using application programming interface (API) or building blocks for email support tickets for customers, for example. This would require in-house IT and data science skills to a limited degree for understanding and sorting of a sentiment analysis API or entity extraction API to allow a little bit more data on top of that system. Amazon, Google, Microsoft, and IBM have APIs for vision, NLP, and other AI applications – most of which require minimal AI experience, and often don’t ever require an ML engineer.
Vendor or In-House AI Solutions: A more robust version of this gradient would be more custom solutions requiring deeper vendor integrations into existing systems. There might be a need to tie it into in-house data in order to train a historical model, which would then run in a pilot project to eventually develop into an AI deployment roadmap for the company. The process for fully deploying such a solution is complex (see our AI strategy report called The AI Deployment Roadmap).
Smaller enterprises will typically need to begin with simpler projects, and projects more geared toward AI maturity building than radical changes to actual workflows or data infrastructure. This capability ROI is something that won’t be understood without leading with AI education. For outside AI product vendors and AI consultants – or in-house innovation leaders within smaller enterprises – we recommend education and capability-oriented projects a good guideline for getting started with AI.
The specifics will vary from client to client, and project to project, but that general order will serve smaller enterprise much better than the alternatives – which usually involve aimless “toy” AI projects, or overly complex vendor solutions that the client doesn’t have the budget or risk tolerance to handle.
Smaller capability-oriented projects allow smaller firms to:
- Learn more about how AI works (particularly for company leadership)
- Build team skills, data infrastructure understanding, a cultural value of data, and other key factos for AI maturity (again, see our Critical Capabilities framework)
- Gradually grow in-house data science talent at a rational pace
Smaller enterprises who squander their early AI efforts will be left with less confidence in AI, less willingness to try future projects, and less AI maturity to stand on for future applications. Smaller enterprises that develop executive AI fluency and build AI maturity early on will be vastly more prepared for the future – and will have a stronger ability to act on future AI innovation efforts. In-house and external (consultants, vendors) advisors who can build trust and steer these smaller enterprises the right way will be rewarded in the long run.
Emerj’s Catalyst Advisory Program
Smart AI and IT consultants and services leaders don’t just become good salespeople, they become trusted experts who build long-term relationships with their clients.
The Emerj Catalyst Advisory program is designed to help services leaders land and expand new AI business, and deliver maximal AI value for their clients. Get one-to-one guidance on developing your AI practice – and discover proven frameworks for estimating AI ROI, finding AI opportunities, and more.
