This article is the third in a series part in a series about AI product development.
In the first installment in this series, we covered how to develop AI product ideas with both near-term adopt-ability and long-term potential.
In the second article in this series, we covered how to score and rank AI product ideas.
In this last installment, we’ll talk about how to approach your AI product development roadmap – with unique advise for established firms and for startups.
Once you have brainstormed your product development ideas and comparatively scored these ideas using our system from the previous article, you can select which long-term AI advantages to aim for as a company. Additionally, you can now decide which short-term AI projects or products you will develop in order to obtain a short-term ROI and help us reach those long-term advantages.
It’s critical to keep the future competitiveness of your company in mind as you go through this exercise. AI is a long-term investment, a new set of skills, and a new way of doing business. Because of this, the ROI of your first AI product may not be the immediate financial return of that product (see Emerj’s three-part AI ROI model).
Monetary gain is still something to aim for and try to ensure, but the initial ROI will still include development of the skills and technologies that serve as the foundation of AI competence. It is very hard to invest money into something without an immediate financial return unless we understand that we are building towards long-term advantages. These include such critical capabilities as building data infrastructure, hiring data science talent, forming strong teams, and establishing data dominance.
The next step is to determine the long-term AI advantages you want to gain based on the previous exercise, and then to identify the AI-related initiatives you should be starting now.
Determining the best long-term advantages in AI is a process that varies depending on the kind of business you are:
- Established Enterprise: A company founded well before they ever considered AI. A well-established company where AI is not a core part of processes or products.
- AI Startup: A company built with the idea in mind that AI would be how it would serve its customers better than the competition.
We provide some recommendations on how to think about the relative value of the scores (outlined in our previous article) – tailored respectively for enterprises and startups.
Recommendations for Established Enterprises
Established enterprises should be looking for a sweet spot between their own abilities and the market need and size. Competition is still a factor, but it is less of a concern here because of the high financial resources, in-house expertise (or the ability to hire said expertise), and connections to customers and partners in your market. Enterprises need to focus on their core capabilities and how they might fulfill the current market need.
Additionally, you should emphasize the market size and how much of it your new product might be able to satisfy. As an enterprise, it makes sense to primarily focus on ways to enhance and augment your existing offerings. While it is possible for an established enterprise to develop an entirely new AI product separate from their core competency, it is generally not advisable to do so.
It is more advisable to extend your existing product suite. You have thousands of customers who know how you can help them. Use that branding to enhance the value you are known for with AI.
It is possible to spin out an entirely new company that is disconnected from your current business initiatives, but it is ill-advised. This is because you are not going to have the momentum and agility to become a completely new company the way an AI startup might. However, you will have existing expertise and connections within your company, and so it is more advisable to play to your strengths as an enterprise than to split off for unrelated initiatives.
- If you are selling point-of-sale software for high-frequency retail, you will want to focus on potential customers in large retail areas such as department stores, grocery stores, and businesses with numerous mall locations and standalone stores.
Jim Collins’ bestselling book Good to Great includes an entire chapter about “technology accelerators” – and his 2002 essay on this topic rings as true for AI as it did for the transition to the internet in the early 2000s.
Scores by Order of Importance
- Expertise and Connections Scores: These are the most important for established enterprises because they enable more options for how to move forward with AI initiatives. Enterprises should keep in mind that they should not be as concerned about competition because of their access to experts and other important business connections.
- Market Size and Need Scores: The next most important scores are for market size and need. Enterprises should consider just how many people or businesses will need the product and how they can satisfy as many of them as possible.
- Competitiveness: Competition is least important for enterprises in that they have enough funds and customer connections to be able to focus on other areas in more depth. While it is important to keep the intentions of competitors in mind, enterprises should make sure they have a good understanding of the market size and market need first.
Strengths of an Established Enterprise
- Existing funds, connections, and expertise
- The ability to pounce on the biggest market size problems which can drive long-term competitiveness
Weaknesses of an Established Enterprise
- You are relatively locked into serving who you already serve
- The inability to steer the entire company around AI-focused market need
Recommendations for AI Startups
Market need should be the highest priority for AI startups when considering how to move forward with ideas for AI products and initiatives. As a new company, you need to pay your bills. You will most likely try to do that by getting revenue from your customers, in which case you will need to find a market need that someone is willing to pay for soon. Alternatively, you could attempt to get venture capital funding. The best way to accrue venture capital is to get traction, or customers starting to use and adopt your product. Initial traction based on market need is the most important asset for an AI startup.
Even within a startup, you may have a lot of expertise on your team. While this can help inform your choices, skills and connections are less valuable when your company is new. It is more important to capitalize on the most prominent market needs and derive early revenue and traction from them. Many startups miss out on this because they are focusing on the specific expertise of the first few people to join their team.
AI startups are the most sensitive to competition. You will likely not be looking at the biggest market size segment right away, nor the segment with several well-established competitors.
Sometimes the expertise will come directly from your company’s founders. Marc Benioff, for example, had experience with some B2B software at Oracle before founding Salesforce. Jeff Bezos, on the other hand, had no eCommerce experience prior to Amazon. The real winning factor between these two is finding and taking advantage of a crushing market need. Connections and expertise are always helpful, but ultimately market need is going to be the biggest focus area.
- A company that offers an AI platform for building chatbots may market their product towards business areas that deal with numbers and calculations along with words. This includes customer service, simple inquiries, financial advisory, and mobile banking.
- An AI startup offering automated credit scoring or lending may want to find areas that have a high market need within that sector. An example of this would be providing scores based on alternative data such as social media posts in cases of underbanked customers.
Scores by Order of Importance
- Market Need: Identifying a product or service that your market is in desperate need of is the most important factor for AI startups to consider when deciding on their first initiatives.
- Competitiveness: Consider what might be missing from current market offerings and how AI can deliver that on top of services competitors already offer.
- Market Size, Expertise, and Connections: AI startups tend to lack the level of expertise and connections required to give such a new company an edge over simply providing a better product that fills an important niche. Because of this, they are limited to fulfilling market needs on a smaller scale and thus should not concern themselves with large market size opportunities.
Strengths of an AI startup:
- You can focus just on where market need is strongest. Additionally, you can build connections and expertise around that and transmute it into a much larger long-term advantage.
- Agility to lean into more AI initiatives and explore different business areas that may be able to benefit from your product.
Weaknesses of an AI startup:
- They are unable to focus just on market size, and must focus on market need.
- They are unable to compete toe-to-toe with companies that start with talent, data, and revenue from day one.
