When most professionals think about “AI consulting” they tend to think about technical machine learning services, like: Building our data infrastructure, crafting and testing new algorithms, interesting AI systems into existing IT infrastructure.
However, as AI proliferates industry – and leaders realize just how much of AI adoption relies on non-technical skills – more and more AI services firms are involved in strategy and management consulting (no code involved). As AI fluency grows in the C-suite, there will only be more and more demand for non-technical, strategic AI leadership skills (we’ve covered this dynamic in a longer article titled The AI Career Gap).
This article highlights five non-technical AI service business models inspired directly by our community of Emerj Plus and Catalyst members.
If you’re an enterprise team member, this list might provide you with new ideas for how you might grow your skills and add more value to your organization. Most of these services that non-technical consulting firms provide will also have correlative internal roles in the enterprise.
If you’re entrepreneurially minded, you might also pick up some ideas about how to combine your own unique industry experience with AI strategy knowledge to start a services firm of your own.
There are no limits on the number of AI-related consulting services that non-technical professionals.
My aim is not to limit your imagination with the business model ideas listed below. Rather, you should use these business models to open up the art of the possible, and build a viable menu of possible futures for your business – based on business models that we’ve observed in the real world.
With that said, let’s explore the pros and cons of the five business models in question:
1. Speaker and Workshop Leader
Speaking clients might include industry associations, enterprises with strong internal education initiatives, conferences, or large AI vendors (who often host events of their own). Most speakers (myself included) begin with free or low-fee talks to build their reputation, gradually building up to substantial fees.
Speaking services might include:
- Keynotes on current AI trends and capabilities, on near-term AI impact, or on how to apply AI productively in an enterprise environment
- Panels and moderation of virtual or in-person events
Workshop clients are most likely to be corporations with explicit needs (i.e. a retail bank interested in building an AI strategy) that overlap with your workshop offerings. Industry associations are often open for workshop offerings as well.
Workshops might include:
- AI discovery sessions, showing enterprise deals how to find AI opportunities in their own businesses
- AI project management workshops, helping other non-technical leaders manage the complexities of AI projects to have higher rates of success
Pros:
- A great aide to personal brand-building and industry recognition.
- Can serve as a foot in the door to nearly any other AI-related service (consulting projects, market research, etc).
Cons:
- Often not lucrative enough to build a business on alone – reliant on a suite of other services in order to be financially viable.
2. Business Lead with a Technical Deployment Partner
This model involves a business-focused, non-technical founder who partners with a technical ML engineer – or more likely a team of ML engineers – and goes to market to sell work. Once the work is sold, the two parties (non-technical founder and ML team) come together.
This allows the non-technical founder to go to market offering technical AI services without needing to take on the expensive overhead of ML engineers early on.
The benefit of this model is that the majority of enterprise leaders believe that they want technical AI services.
The non-technical catalyst skills of project selection, expectation-setting, and understanding AI maturity are generally undervalued, and get included (as a necessity to the success of the project) along with a larger technical initiative.
Many consultants started their AI consulting businesses with outside partners – and only later hired data science talent when revenue was consistent enough to afford it. Dave Timm of Red Marble AI – an AI consulting firm in Australia – did just that. Listen to his full episode on the AI Consulting Podcast:
Pros:
- Many enterprise clients believe that they need technical support – and only as a tertiary benefit (i.e. the technical services are the spoonful of honey that helps the AI education medicine go down). Having technical delivery ability through delivery partners allows you to attract more AI projects (again, because enterprises are likely looking for technical delivery, even when they need much more) without taking on the overhead involved in hiring new consultants.
Cons:
- The complexities of technical AI projects are significant, and you’ll inevitably learn many hard lessons getting into the thick of actual AI initiatives and deployments (including issues around setting expectations, building AI maturity, and more). Projects can often be so consuming and hands-on that consulting leaders aren’t able to keep other prospects warm and build a pipeline of later work. Living project to project without such a pipeline is stressful.
3. Market Researcher or Analyst
Emerj Artificial Intelligence Research is a good example of a firm of this kind. Services include:
- Competitive intelligence on the AI investments and results of similar enterprise firms.
- Vendor selection and vendor comparison research.
- Feasibility research for the success of a specific AI initiative.
This business model requires some familiarity with market research methodologies, and the ability to generate a rich rolodex of AI experts across the vendor and enterprise ecosystems.
Pros:
- Services can be relatively productized.
- Some market research assets can be sold many times over to multiple clients.
Cons:
- Market research is less tangible and understandable than many others kinds of services, and requires a marketing and sales approach that make the benefits of the service quantifiable and understandable to the buyer.
4. Coach or Retained Advisor
A coach or retained trainer works with clients (either individually or in groups) to help them with their specific goals on an advisory-only level, usually without involvement in bespoke and complex projects.
Emerj offers a program called the Catalyst Advisory Program to AI consultants and AI services company leaders. This program includes a series of monthly live calls and one-to-one calls, along with a Slack community and custom curriculum designed to help AI consulting founders win more sales.
Examples of other coaching offerings might include:
- A group coaching program for AI project managers to help them grow their career-relevant skills and deliver success with AI projects.
- A coaching program for retail executives to help them find and execute on high-ROI AI projects.
- A coaching program focused on building a culture of innovation for project teams getting involved in data and AI projects.
- Etc…
Another on of our Emerj Plus members, Karen Silverman, was a career attorney at a large firm in the Bay Area. She never learned to code, but she did learn about AI use-cases and best practices. By combining this with her legal experiences she launched the Cantellus Group, and today she consults with large organizations on the legal risks of their AI and data applications. Listen to her AI consulting founding story:
Pros:
- As a productized service, coaching programs are “bounded” in terms of involvement, and can provide a reliable monthly stream of revenue for a predictable amount of work (unlike bespoke, hands-on consulting projects).
- Coaching programs can serve as a focus group for the development of other services – allowing the coach to develop other service offerings based on the lessons learned from helping clients (for example, Emerj Plus – which can be offered at scale – is developed based on the requests and demands of Catalyst members, allowing an non-scalable coaching program to help grow a scalable subscription offering).
Cons:
- While it is possible to train a team of coaches and reach a broader audience, coaching programs are hard to scale and often rely on the personal brand of the founder.
5. Sales Enabler for AI Vendors
Vendor companies have a predictably common motive: driving sales. Any service that helps AI vendors reach their audience or drive additional sales can be a potentially viable service.
Examples of vendor sales enablement programs might involve:
- Sales coaching for AI product and service companies – training individual or sales teams about the unique considerations for selling AI services.
- Creating thought leadership and co-written reports for vendor companies (we do this at Emerj with our Creative Services offerings).
- Offering exposure for AI vendors to reach a wider audience and deliver leads (we do this at Emerj with our Creative Services offerings).
Pros:
- Vendors have faster sales cycles that enterprises, and unlike many enterprise – vendors are open to working with smaller, nimble firms (as opposed to known entities and established brands) – potentially making them an ideal first client for some new AI services firms.
Cons:
- AI vendor engagements are higher volume, lower dollar-value – while enterprises can sustain much larger (and even longer-term) contracts.
Action Steps for Service Providers
Think through the business models that might be best suited for your experience, skillset, and desired lifestyle.
For example, if you have the desire to grow a large services firm, the speaking and workshop leading business model may be a viable start, but would need to pivot quickly into more robust management consulting or technical consulting engagements. Or, of you have the desire for control over your time, having productized services such as retained coaching or virtual workshops might be a strong fit.
Ask yourself:
- What is the business and lifestyle I want to have in five years?
- Which of the above business model ideas is most in line with my skills and experience?
- Which of the above business model ideas is most in line with my goals?