So you’ve decided you want to take an AI product or service to market.
Before you sell anything – you’ll have to decide what kind of product or service to develop.
- What specific problem will you solve, and for who?
- Which features and benefits will be most important for our buyers or users?
- What workflows and data sources will you integrate with?
- What kind of ROI measures will prove your value?
The number of AI product variations one could think of is endless. Whether we’re working with a startup or an larger firm spinning out an AI product, we encourage product teams to come up with many potentially viable ideas – and then to rank them based on the following factors:
- Internal Skill Sets and Expertise
- Internal Connections and Network
- Relative Market Size
- Relative Market Need
- Relative Competitiveness
In the remainder of this article we’ll explore each of these factors in greater depth.
This article is part two in a three-part series about AI product development. Read part one here: Developing AI Products: Winning in the Near-Term and Long-Term. In that article, we explore the process of generating viable AI product ideas, while article will focus more exclusively on ranking those viable ideas.
Our AI Product Development Roadmap process uses five unique scores, each of which can be applied to any near-term AI product idea. While the scoring criteria occasionally vary from client to client, the scoring “rules of thumb” below (all listed on a 0-4 scale) should serve as useful defaults for developing relative ratings for AI products and projects.
In practice, your project ranking might look something like the simple chart below:
Skill Sets and Expertise
Ask: “What are We good At?”
We encourage clients to generally focus on business processes and functions (i.e. email marketing, brick-and-mortar fashion retail display, international shipping logistics, construction materials, mobile shopping experiences, accounting and forecasting, etc).
For any AI product idea, score your own level of expertise in that functional business area on the following scale:
- 0 – No skill at all
- 1 – An rough understanding
- 2 – Familiarity with the process in concept and context, but not hands-on application
- 3 – Hands-on familiarity with the process area, but not mastery
- 4 – Deep subject-matter mastery (top 5% of skill among professionals in your field)
The Internal Skill Sets and Expertise score is scored by you, based on your own estimates.
Connections and Network
Ask: “Who Knows Us and Trusts Us?”
We encourage clients to generally focus on industries (auto insurance, lumber retailers, etc), departments (accounting departments, compliance department, etc), and titles (CEOs, heads of sales, etc).
For any AI product idea, score your own level of connections and network on the following scale:
- 0 – Zero connections
- 1 – Could get in touch with people through existing connections, but no direct connection
- 2 – Connected to buyers in a related industry or function
- 3 – Connected to buyers in that industry or function, but not deeply embedded
- 4 – Deeply embedded with buyers in that industry or function
The Internal Connections and Network score is scored by you, based on your own estimates.
Relative Market Size
Ask: “What is the Total Value For Solutions to Solving X Problem?”
While our research services clients sometimes work with Emerj to do accurate market size estimates, most market size estimates will be gleaned from a combination of:
- Professional intuition (which is only to be trusted if you have robust experience in this sector or function).
- Secondary data about the sector or function.
- Primary market research data from professionals and leaders in that sector or function.
Note that the 1-4 score is relative to the evidence of market size. This means that you will likely need to estimate the market size of three to five AI product ideas before having a sense of what a “4” is, and what a “1” is. The important lesson is to know what market sizes are larger and smaller, not necessarily to land on a specific number.
While it is possible to base Relative Market Size on professional intuition alone, we do not advise clients to do so.
For any AI product idea, score relative market size on the following scale:
- 0 – No evidence of market size
- 1 – Lowest evidence of market size
- 2 – Lower-middle
- 3 – Upper-middle
- 4 – Highest evidence of market size
Relative Market Need
Ask: “How Much Are Existing Buyers Explicitly Looking for This Kind of Solution?”
Ask: “How Keen is the Pain That We Are Explicitly Trying to Solve With This Solution?”
This is a relative score.
For any AI product idea, score relative market need on the following scale:
- 0 – No evidence of market need
- 1 – Lowest evidence of market need
- 2 – Lower-middle
- 3 – Upper-middle
- 4 – Highest evidence of market need
Note that the 1-4 score is relative to the evidence of market need. This means that you will likely need to estimate the market need of three to five AI product ideas before having a sense of what a “4” is, and what a “1” is.
The Relative Market Need score should draw upon professional intuition, secondary research, and primary research interviews.
Relative Competitiveness
Ask: “How Many Other Companies (AI or Not AI) Are Trying to Satisfy the Same Customer Need?”
For this exercise, we are examining the level of direct (vendors that sell the same kind of product to the same buyer) indirect (vendors or solutions that satisfy the customer’s need in a different way) competition.
This score is based on the following two factors.
- Number of companies aiming to satisfy this specific customer need.
- The relative maturity of the companies aiming to satisfy this specific customer need.
For any AI product idea, score your own level of connections and network on the following scale:
- 0 – Highest level of competitive solutions (in terms of problems solved, product features and benefits)
- 1 – Upper-middle
- 2 – Lower-middle
- 3 – Lowest level of competitive solutions (in terms of problems solved, product features and benefits)
- 4 – Little to no evidence of competition
The 1-4 score is relative to the competitive landscape. In a space of relatively high competition, having three direct competitors and two indirect competitors might be a score of 1 or 2, while in a relatively low competition space, it might be a score of 4. The objective here is to provide a relative range to score AI product ideas.
Note that the Relative Competitiveness score scores “Little to no evidence of competition” high. This is because low competition spaces are more desirable. As with the rest of the scores in AI Product Development Roadmap, the goal is to find high score opportunities.
The Relative Competitiveness score should draw upon professional intuition, secondary research, and primary research interviews.
In part three of this series, we share practical advice for how enterprises, mid-size firms, and startups should approach AI product development (and go to market strategy) differently. Read part three: Building Your AI Product Development Roadmap – Recommendations for Startups and Enterprise Leaders (Part 3 of 3).
