Goodbye JBTD - Hello ODJS for AI Products!

Why JBTD could break down for complex products such as AI Products and Outcome Driven Job Stories could drive success of AI Products.

The core idea behind JTBD for AI Products is that customers "hire" products to get specific jobs done in their lives. They are looking to make progress and achieve a desired outcome in a particular circumstance. However, as we will find out, AI Products are unique in that not all jobs are done by the customer.

As an AI product manager, the key is to deeply understand the jobs your target customers are trying to accomplish and then build AI products that help them get those jobs done better than existing alternatives. However, AI Products experience high failure rates and even if the JBTD are well defined, they often break down for AI Products.

• When trying to create truly disruptive, visionary AI products. JTBD is focused on identifying existing customer needs and jobs they are already trying to accomplish. However, breakthrough AI innovations often create entirely new markets and enable jobs that customers couldn't conceive of before.
• When exploring radically new AI capabilities. If an AI product or feature is extremely novel, customers may not have an existing "job" that directly maps to it. Their current processes for accomplishing goals may look very different from what an AI could enable. In these cases, rigidly adhering to JTBD may constrain innovative possibilities for AI.
• When AI spans many potential jobs and industries. Some AI technologies like large language models are "horizontal" - they can be applied to a huge variety of possible jobs and use cases. Studying existing jobs may provide limited insight compared to envisioning completely new capabilities. The jobs don't exist yet for the AI to be "hired" for.
• When speed of AI development outpaces speed of JTBD research. Studying customer jobs in depth takes time - interviews, observation, analysis. But AI product development can move very quickly, especially for digital products. Insights from JTBD research may be stale by the time new AI features are ready to ship.
• When AI is not directly customer-facing. Some AI provides backend optimization, analytics, automation, etc. rather than directly helping end users accomplish their jobs. In these cases, the "jobs" are more technical problems to be solved rather than customer jobs-to-be-done.
• As mentioned earlier, for AI Products, some jobs are not done by the customer at all which is the automation part. Some customer executed jobs are augmented by the capabilities of the AI Product creating some dependency.

To overcome these issues, alternate approaches such as Outcome Statement, merging user stories to create Job Stories have been experimented with. However, they have had limited success in facilitating Outcome Driven Innovation for AI Products.

Hence I propose and have applied an "evolved" framework called the Outcome Driven Job Stories that seem to particularly relate well to AI Products development.

Outcome Driven Job Stories are a framework used to understand and articulate customer needs by focusing on the context, causality, motivation, and measures of success. This approach helps teams build products that deliver real, measurable value to users.

As one can see, there are key differences between ODSJ and JBTD, as ODJS goes beyond the experience part of JBTD and evolves to Experience + Outcomes + Signals of Success.

Key Differences:

1. Focus on Outcomes vs. Jobs

JTBD focuses on the specific job or task the customer wants to accomplish.

Outcome Driven Job Stories extend this by emphasizing the desired outcome and how success will be measured.

2. Context and Motivation:

JTBD captures the context and motivation but often remains at a high level.

Outcome Driven Job Stories provide a detailed context, including the situation, motivation, expected outcome, and the business implications.

3. Incorporation of Business Outcomes:

JTBD primarily focuses on the user’s perspective.

Outcome Driven Job Stories link user needs to business outcomes and measurable signals, ensuring that the product development aligns with both user satisfaction and business goals.

4. Measurable Signals and Data Instrumentation:

JTBD does not typically include a direct link to how success will be measured.

Outcome Driven Job Stories explicitly include measurable signals and how these will be tracked and validated through data and experimentation.

The ODJS approach is strategically designed to align AI Product features with business objectives and user needs, while also embedding data-driven validation and adaptability into the product development lifecycle. At the core of ODJS is the focus on context and user motivation. Traditional user stories often concentrate on user actions without delving into the situational context or underlying motivations. ODJS starts with "When <a situation>", allowing teams to explore the complexities and specific circumstances in which AI features are to be used. Understanding the user's intent ("The user wants to <motivation>") is essential for designing AI systems that effectively address real needs and enable desired outcomes ("So they can <expected outcome>").Incorporating business outcomes into the narrative ("We believe this capability> will result in this <business outcome>") ensures that the product's functionality is directly tied to achieving specific goals, such as improved efficiency or customer satisfaction. This alignment is particularly critical for AI Products, which require significant investment. By requiring teams to define measurable signals for success, ODJS fosters a culture of accountability and continuous improvement. The framework insists on identifying quantifiable indicators ("We will have confidence when we see <measurable signals>") that offer objective benchmarks for tracking progress. ODJS's compatibility with the Objectives and Key Results (OKRs) methodology is a testament to its strategic utility. It translates the product features into quantifiable objectives and key results, providing clear direction for product development and allowing for better alignment with the company's strategic vision. The framework's insistence on measurable outcomes ensures that AI features are continuously evaluated against their ability to meet both user and business objectives.

Lets us see an illustration of the ODJS framework applied to the Gen AI Mental Health Management Product that I am working on.

1. Experience:

Format: When <a situation>, The user wants to <motivation>, So they can <expected outcome>

Example: When Michael feels overwhelmed, the user wants to express his emotions through art, so he can find relief and manage his anxiety.

Purpose: This captures the context of the user's situation, their motivation, and the expected outcome they hope to achieve.

2. Experience + Business Outcome:

Format: When <a situation>, The user wants to <motivation>, So they can <expected outcome>. We believe <this capability> will result in this <business outcome>. We will have confidence when we see <measurable signals>.

Example: When Michael feels overwhelmed, the user wants to express his emotions through art, so he can find relief and manage his anxiety. We believe that providing an intuitive AI art tool will result in increased user satisfaction and retention. We will have confidence when we see high user engagement and positive feedback.

Purpose: This adds a layer connecting the user’s experience to the business outcome, providing a hypothesis about how a specific capability will drive business success and identifying measurable signals to validate the hypothesis.

3. Experience + Business Outcome + Signals:

Format: When <a situation>, The user wants to <motivation>, So they can <expected outcome>. We believe <this capability> will result in this <business outcome>. We will have confidence when we see <measurable signals>. We will measure and track this <signal> by instrumenting these <data> with this <experimentation framework>.

Example: When Michael feels overwhelmed, the user wants to express his emotions through art, so he can find relief and manage his anxiety. We believe that providing an intuitive AI art tool will result in increased user satisfaction and retention. We will have confidence when we see high user engagement and positive feedback. We will measure and track user engagement by instrumenting session duration and frequency data with our user analytics framework.

Purpose: This includes the most comprehensive view, tying the user experience to the business outcome, and defining how success will be measured and tracked using data and experimentation frameworks.

Here are some tenets related to ODJS:

• The unit of analysis is no longer the customer or the product. It’s the story of a Job well dome with measurable outcome that the customer is trying to get done.
• Needs aren’t vague, latent and unknowable. They are aligned with metrics customers use to measure success when getting a job done.
• A ODJS is stable over time and solution agnostic. AI Products can evolve rapidly with technology, features and capabilities; however it can continue to deliver a good Job.
• A deep understanding of the customer’s needs and wants through ODJS makes marketing more effective — and innovation far more predictable.
• AI Product innovation becomes predictable when “needs” are defined as the metrics customers use to measure success when getting the job done.

AI Product Leaders that I have worked with and show the ODJS framework seem to appreciate how the ODJS can help them:

• Align the organization around a common vision
• Achieve a shared understanding of customer needs across the organization
• Determine which firms to acquire and partner with
• Discover new markets to enter
• Adopt innovation practices that lead to greater success

When AI Product Managers think about their customers and markets and possibly Product-Market Fit from the ODJS perspective, they are much more likely to create and deliver extraordinary AI Products and services. Why? AI Products come and go, but the customer’s ODJS is stable over time. With a stable unit of analysis, you can define customer needs that are stable over time, too, which gives you unique, robust targets for value creation. In short, using a ODJS lens, you can take your understanding of customer needs to the next level. And with the right customer inputs, you can orchestrate the systematic creation of customer value across the organization.