AI and Your 2025 Operating Plan

(NOTE: The following is excerpted from a white paper I recently co-authored with Frank Fillmann of Salesforce .)

AI is on the mind of every enterprise, from the board of directors on down, and rightfully so.  The most pressing challenge right now is to determine what role it should play in next year’s operating plan.  Specifically, as an enterprise leader, you need clear answers to the following five questions:

1. What can AI do today?
2. What kind of infrastructure is necessary for AI to work?
3. What kinds of use cases warrant consideration?
4. Where is the ROI, and what are the critical metrics?
5. How important is it to act fast?

What can AI do today?

AI brings three game-changing capabilities to the table that can be incorporated into a 2025 operating plan:

1. Discover, detect, and control process deviations from operational guidelines.  This is the domain of predictive AI.
2. Engage in conversation, respond to questions, and give guidance while relying on natural language with no user training required.  This is the domain of generative AI.
3. Follow directions, process information, take actions, learn from errors, and continuously improve its performance.  This is the domain of agentic AI.

What kind of infrastructure is necessary for AI to work?

For predictive AI, the primary requirement is a comprehensive transaction history, both at the database level (for content) and the log file level (for time-dependent correlations).  The infrastructure must refresh this data on a consistent and timely basis and ensure its quality and integrity.  Where multiple disparate systems are involved, as is frequently the case, the infrastructure must extract, aggregate, and normalize all participating data sources to enable AI correlations to proceed properly.

For generative AI, the primary requirement is to have a corpus of validated information suitable for exposure to end-user queries.  Then one needs to license a Large Language Model to provide the conversational interface as well as a prompt management system which can assist end users in providing the AI with useful context as well as develop a library of increasingly effective and efficient prompts for recurrent use.

For agentic AI, you need to meet all the requirements listed above.  In addition, you need to have proven automated workflows that can be invoked by AI agents via application programming interfaces (APIs) as well as agent-building capabilities for both subject matter experts and IT professionals so that the system can adapt to changes effectively and efficiently.

What kinds of use cases warrant consideration?

For predictive AI, the use cases will likely involve process management where there is a continual flow of transactions that needs to be managed or a given set of control limits that need to be maintained.  Optimizing transportation and logistics is a good example.  Digital marketing is a second.  Financial crime detection is a third.

For generative AI, the use cases will likely involve connecting people with questions to applications with answers.  Contact center services are a prime example.  Personalizing retail interactions is a second.  Business intelligence is a third.

For agentic AI, the use cases will likely involve offloading repetitive procedures or accelerating workflow.  Regulatory compliance is one obvious example.  Contract compliance with changing terms or pricing is less recognized but potentially a more valuable second.  Improving customer and/or employee experience with faster and more accurate responses is a third.

Where is the ROI, and what are the critical metrics?

To get the ROI from an investment in AI clearly in focus, it helps to look at the enterprise through the lens of zone management.  In this framework, there are four zones of activity, each with its operating model, each of which will be looking for a different ROI from an AI investment:

1. The Performance Zone.  This is where you sell what you make and make what you sell.  It is where you deliver your value to the world.  ROI comes from increases in competitive advantage and customer success, leading to higher win rates, better retention, greater market share, increased leverage over ecosystem partners, and lower discounting.  The critical metrics focus on effectiveness.
2. The Productivity Zone.  This is where you deliver the systems that allow the Performance Zone to operate effectively at scale.  ROI comes from faster response times, better quality, more accurate information, successful compliance, and lower costs.  The critical metrics focus on efficiency.
3. The Incubation Zone.  This is where you engage with next-generation disruptive innovations, be they in technology, business model, or operating model. ROI comes from making better informed decisions on which ones to invest in, which to ignore.  The critical metrics focus on adoption and product/market fit.
4. The Transformation Zone.  This is where you go to make a material change either to your portfolio of offerings, your operating model, or your business model.  ROI comes from transitioning from being a marginalized legacy player to being an active participant in the next big thing.  The critical metrics focus on inertial momentum getting to a tipping point.

How important is it to act fast?

The answer depends on how far the disruptive innovation has progressed through the Technology Adoption Life Cycle in your geography and industry.  There are four stages in the life cycle that matter, and each one comes with its own imperatives:

1. The Early Market.  This is a time when disrupters need to act fast in order to gain first-mover advantage.  Exploiting this kind of opportunity is best suited to start-ups funded by venture capital and unencumbered with any other obligations.  It is not critical for established enterprises to participate at this time, but if there is interest and bandwidth, they can activate their Incubation Zone to explore the possibilities.
2. Crossing the Chasm.  For a handful of use cases, it is important to act fast, but not for technological reasons.  Rather, it is because there is an urgent need to address a mission-critical business process problem, and the current solutions are simply not cutting it.  Start-ups can focus on a single use case in a single industry in order to put their business on a solid footing, no longer dependent upon raising another round of funding.  Established enterprises can free their operating model from a trapped-value bottleneck that is bringing down their overall performance, initially by tackling it in their Incubation Zone, and then once the solution has proven out, transitioning to either the Performance Zone or Productivity Zone, depending on where the bottleneck resides.
3. Inside the Tornado.  This is when the technology has caught on, adoption is accelerating like crazy, and either you are going to catch the wave or the wave is going to catch you.  Now, everyone must act as swiftly as they can.  Vendors need to capture as much market share as possible, and customers need to modernize their operating models as fast as possible.  The pressure is on the Productivity Zone to accelerate deployment and adoption and on the Performance Zone to deliver value realization.  If inertial resistance is stubborn and you are falling behind, this warrants activating your Transformation Zone. 
4. On Main Street.  The technology has now become reliable infrastructure, and the urgency to act is replaced with the opportunity to optimize.  This is a great time to do a trapped value analysis and target processes that have been holding you back for years.  This is what the Productivity Zone does best, and the bulk of all ROI from technology investments is captured at this stage.

Summing Up

To recap, here are the five questions you want to develop clear answers to in order to incorporate AI into your 2025 operating plan:

1. What can AI do today?  Specifically, what can it do that might be material to our success in the coming year?  Who can we talk with to get a deeper understanding of our opportunities?
2. What kind of infrastructure is necessary for AI to work?  Do we have this infrastructure in place today, or do we need to invest, and if so, in what exactly and for how much?
3. What kinds of use cases warrant consideration?  Specifically, have we identified one or more use cases that could materially impact our success in 2025, and if so, what are they?
4. Where is the ROI, and what are the critical metrics?  Specifically, what zone will host our AI initiative, and to what metrics are we going to hold ourselves accountable?
5. How important is it to act fast?  Where do we think we are in our local technology adoption life cycle, and based on that conclusion, what time-to-value objectives do we intend to set?

That’s what I think.  What do you think?

Follow Geoff on LinkedIn | Geoffrey Moore Main Mailing List | Infinite Staircase Mailing List

Geoffrey Moore | Infinite Staircase Site | Geoffrey Moore Twitter | Infinite Staircase Twitter | Facebook | YouTube