Why AI Product Managers should do periodic wellness check for their AI Product Strategies

I was recently engaged to brainstorm and figure out why an AI Product was not doing well and getting less traction with customers. The AI Product in focus was actually an innovative physical AI Product; well designed, cutting edge features and a great form factor. It was a puzzle to both the company and myself initially as to why the AI Product was not experiencing growth and success.

Going back to the drawing board and tracing through the Product Lifecycle history, I asked about the AI Product Strategy, Mission and Vision. Right away, I could tell that something was not jelling

So I assumed the role of Doctor of Strategy to detect and diagnose any "issues" with their AI Product Strategy. The focus was not so much on "ills" but on the rational, objective and structured approach to the evaluation.

Adapting the principles from the book "Good Strategy, Bad Strategy" by Richard Rumelt, here is what I came up with for envisioning the hallmarks of a Good AI Strategy / Bad AI Strategy.

Good AI Product Strategies:

• Align AI initiatives with overall business objectives. Identify specific areas where AI can make a significant impact, such as improving customer experience or creating new revenue streams.
• Adopt a customer-centric approach. Conduct thorough market research, gather user feedback, and continuously iterate based on customer insights to ensure AI products deliver meaningful value.
• Foster cross-functional collaboration between data scientists, engineers, designers, and business leaders to effectively align technical capabilities with business objectives.
• Embrace an agile and iterative development approach to adapt to changing market dynamics and user needs more effectively.
• Prioritize responsible AI adoption by ensuring transparency, fairness, and privacy safeguards are built into AI systems.
• Maintain a AI Product Management approach by balancing automation with human creativity, iterating and refining, and staying informed about the latest AI advancements.

Bad AI Product Strategies:

• Lack of clear objectives for how AI will solve specific business problems or deliver value to users.
• Failure to adopt a change management strategy to address the organizational and cultural shifts required for successful AI adoption.
• Overestimating the capabilities of AI and having unrealistic expectations, without understanding its limitations and the substantial input and management required.
• Ignoring ethical considerations and privacy concerns, which can damage a company's reputation and lead to legal complications.
• Neglecting data strategy and failing to ensure data is clean, organized, and accessible for AI systems to function correctly.
• Treating AI as a one-time project instead of an ongoing initiative that requires regular maintenance, updates, and fine-tuning.
• Going overboard with automated AI-generated content without human oversight, leading to errors and loss of credibility.

It all sounded great in theory, so I was then challenged to present what I would consider is an example of a "Great AI Product Strategy" in action today.

Using Starbucks as an example, I highlighted what I consider the top three standout factors of their AI Product Strategy.

• Customer-centric approach
• Aligning AI initiatives with business goals
• Leveraging AI for personalization
• Enhancing operational efficiency
• Starting small and scaling

Starbucks' mobile app and rewards program capture rich data on customer preferences and behaviors. This powers a sophisticated recommendation engine called DeepBrew that suggests products tailored to each customer based on factors like order history, weather, location, and more. Even when customers visit a new Starbucks, the app identifies them and provides their favorite order to the barista, enabling a seamless personalized experience across all locations. Targeted offers and discounts are individually personalized based on each customer's unique purchase patterns to drive loyalty and re-engagement.

AI helps optimize store labor allocations, inventory management, and equipment maintenance. For example, predictive analytics is used to forecast demand and deploy staff efficiently. AI solutions to intelligently sequence mobile orders so drinks are made at the optimal time based on customer ETA. And so on.

The next step was to do a "wellness check" to figure out how "healthy" their AI Product Strategy was currently.

I broke the Wellness Check into 6 dimensions and associated diagnostic criteria:

• Clarity
• Feasibility
• Adoption
• Adaptability
• Impact
• Governance

Clarity:

1. How clear is your AI Product strategy?
2. How sharp is its focus?
3. Is it clear why you have an AI Product strategy at all?
4. Is it obvious what changes your organization will need to undergo for your strategy to succeed?

You can say your AI Product Strategy is clearly-defined if it:

Defines a worthwhile purpose that the strategy seeks to serve AND defines a destination that, once reached, enables that strategic purpose to be served AND defines a handful of core methods which will change the organization sufficiently for the strategy destination to be reached.

Feasibility:

How well-resolved are your AI Product Strategy goals?

An AI Product Strategy is the sum of its strategic goals. Strategic success will be the sum of the success of these goals, amplified by the synergies between them. For an AI Product strategy to be feasible, the strategic goals needed to achieve it must be well-resolved.

Your AI Product strategy is more likely to be feasible if your key strategic goals are well-resolved – that means they are:

Recognized AND Understood AND Realistic AND Aligned.

Adoption:

Are all stakeholders engaged with, and committed to the AI Product Strategy?

Is everyone engaged with, and committed to, your strategy? For your AI Product Strategy to be fully adopted across your organization, it is crucial to ensure the active engagement of all stakeholders and their willing commitment to your strategic goals.

The AI Product Strategy can be said to be effectively adopted if there is:

Active engagement of all stakeholders AND their willing commitment to the achievement of strategic goals.

Impact:

How well do AI Product Strategy targets drive and guide strategic change?

There are two ways that AI Product Strategy can be said to be impactful. Firstly, if it is successful in achieving its overall objectives; if it reaches its destination and serves its purpose. This ‘ultimate impact’ is measured against targets set for one or more of the highest level strategic goals. This handful of strategic key performance indicators (KPIs) are usually ‘lagging’ indicators – measures of outcome, which may only be measurable after their causes have had time to take effect.

Secondly, strategy can make incremental impact as the strategic plan progresses. This is measured against targets set for all strategic goals. Since these targets are of more value as indicators of progress, they can be ‘leading’ indicators – measures of input and hence immediately measurable.

An AI Product Strategy can be said to be impactful if it can be successful in achieving its overall objectives, and if it can make incremental impact as the strategic plan progresses. To do this, your strategic targets need to guide and drive strategic change by being:

Actionable AND Measurable AND Purposeful.

Adaptability:

How well equipped is the AI Product Strategy to adapt to changing circumstances?

A key feature of an AI Product strategy should be its endurance. The AI Product Strategy acts as a constant, long-term navigational beacon to align action and guide decisions across the organization. AI Product Strategies, however, need to adapt in response to change. This reveals a conundrum: how can the AI Product Strategy be a constant ‘North Star’, yet still adapt in response to changing circumstances? To resolve this, we need to understand that while AI Product Strategy provides a compelling vision of the future, strategic planning devises the transformational change program to get there.

The AI Product Strategy is adaptable if it:

Clearly separates your strategy from your strategic planning AND has built-in resilience AND is nimble.

Governance:

How well is your strategy reviewed and reported?

The AI Product Strategy needs to be ‘governed’ to maintain momentum and to be guided and steered towards strategic success. Adjustments and corrections need to be made. Resources need to be allocated and re-allocated. Priorities need to be refined. AI Product Strategy governance, is the means by which your organization ensures that everyone knows the status of your strategy and are both committed and empowered to contribute to its success.

The AI Product strategy is considered well governed if there are:

Clear, timetabled opportunities to review and discuss strategy AND clear ways to communicate strategy insights and decisions AND evidence-based evaluations of progress and the impact of decisions.

After doing a subjective evaluation for the AI Product Strategy and doing the first wellness check:

The best consensus I could get was there was some parts good and some parts of the AI Product Strategy that could use improvement. Bad Strategy is not something that stakeholders would align with.

The wellness check results visualized below, points to the fact that their initial strategy and vision were very food and they created an initial great product .However, the AI Product Strategy kind of remained stagnant and the organization could not adapt to changing consumer, market and competitive forces. The AI Product did not get the desired growth due to internal organization misalignments.

All of these provided fodder for thinking about possible AI Product Strategy re-working and applying the 80-20 rule to priority areas.

As a self proclaimed Doctor of Strategy, I was able to recognize the symptoms, perform tests and checks, make a diagnosis and subsequent treatment path..

This experience leads to make the statement that AI Product Leaders and Managers would do well to periodic wellness checks of their AI Product Strategies and apply the necessary pivots. With the pace of tech innovation, dynamic market conditions and fast changing customer expectations, AI Product wellness check will become needed and necessary to ensure continued success of AI Products.