July 31, 2024
In the digital world,
MVE is the best route to ROI
Is your company learning faster than the competition?
IDC’s Worldwide C-Suite Survey 2023-2024 showed that over half of C-Suite executives say that GenAI is a top investment priority for their company. An EY study published in July found 95% of senior executives saying their organizations were currently investing in AI.
The majority of this frenized pursuit of GenAI by companies of all sizes, across all industries, is based on the fear of missing out (FOMO). Like most transformative digital technologies, to adopt and deploy them successfully requires a clear understanding of what end result you want them to achieve. It also requres a realistic assessment of what resources and budgets you need to achieve that outcome.
The minimum viable experiement (MVE) approach allows you to test and learn before you adopt and deploy. The major benefit of the MVE is you quickly learn what you don’t know or didn’t aniticipate about the potential impact and value of a specific GenAI application or software tool in meeting end user needs. It also allows you to priorotize progress over perfection.
This approach reverses the roles of the product and the data. Data traditionally captures responses and reactions to individual product features. Now, it is the data about a desired user experience or solution that dictates the product development process.
Companies taking the MVE approach learn faster and have much higher confidence that the early versions of their product successfully address unmet user needs and desires. It will also help mitigate unrealistic expectations for immediate ROI impact.
Learning at the speed of discovery
Speed to adoption and time to utilization are the new metrics of competitive success. New product lifecycles have shrunk from years to quarters to months to weeks to days to hours. Those companies that have successfully adapted to this new cadence are continually distancing themselves from their competition.
Jeff Bezos has built Amazon around learning on a massive scale via experimentation. As he said, “if you double the number of experiments you do per year, you are going to double your inventiveness.”
When Satya Nadella took over as CEO of Microsoft in 2014, he declared that the new game was to be a “learn-it-all company rather than a know-it-all one.” A learn-it-all company makes the decision that every day will be a new day, with learning, exploring, and experimenting the norm.
How quickly can you learn what you don’t know?
Learning faster than the competition is one of the only sustainable competitive advantages. This requires that companies from top to bottom embrace and endorse the value of learning at the speed of discovery.
There is no playbook you can study that provides the right answers to all the unknowable outcomes of GenAI. As such, being able to create a culture and discipline of learning, discovery, and experimentation is the new competitive imperative for success.
Here are some steps & questions to get you started:
Step 1: Validating your desired business performance outcome and ROI
- What is the business problem we are trying to solve or the business opportunity we are trying to seize?
- What must GenAI deliver to achieve that outcome?
- What is the fastest and easiest way to test your assumptions?
Step 2: Understanding your risk assumptions
- What are the major risks of GenAI?
- What are the minimum costs to see if we can avoid or overcome these risks?
- What are the risks of not experimenting and learning?
Step 3: Implementing your MVE approach
- What minimal experiments can we conduct to learn what we need to know?
- What does an MVE roadmap look like?
- What does success look like?
MVEs can be used to create a common understanding and agreement as to what is important to the end user. Getting real-time, end-user responses and reactions to an MVE helps clarify what to keep and what to discard. As Kenny Rogers sang, “You’ve got to know when to hold ‘em and know when to fold ‘em.”
How can you avoid the sunk cost fallacy?
The reason so many GenAI projects have failed to deliver their desired outcomes is because they weren’t in service to solving a specific problem or taking advantage of specific opportunity. They were mostly about deploying a new technology before anyone really understood what it could do and what kind of organizational and operational preparation you must have in place to achieve the desired outcome.
Here are some of the reasons they failed:
- Open-ended outcomes that lack focus
- Specific metrics to measure desired results
- Unlimited time frame for seeing results
- No data management strategy
- No GenAI governance process
- No rigorous GenAI risk assessment process
- Lack of talent to effectively manage the projects
For example, if the GenAI pilot project doesn’t have a clearly defined timetable with agreed upon performance milestones it inevitably falls prey to the sunk cost fallacy. The project team has so much vested investment and effort into the project that they want to keep going because they believe it will achieve its desired outcomes.
One of the major benefits to an effectively run MVE is it allows the GenAI project team to fail fast and learn fast while avoiding the misuse of scarce resources and budgets to chase unrealizable ROI results.
|
