The Gap: AI Expectations vs. Execution
Many executives and managers are beating the drum on using AI for more tasks, yet the lack of clarity around how, when, and where to rely on AI is causing confusion, anxiety, and mixed messages for employees who feel pressure to rack up AI success stories to demonstrate that they are using it.
In some cases, that’s manifesting as haphazard or poorly thought through outcomes.
For example, an AI-generated marketing plan that misses the mark or makes recommendations that aren’t relevant to how the business develops warm leads or generates revenue. Or “vibe code” generated by AI that contains random specs such as security certificates that expire after 180 days, causing unexpected and confounding problems months after new features are put into production.
If the people charged with executing your AI strategy aren’t set up to succeed with a clear framework for the right place and time to use AI, which platforms to use, and appropriate levels of checks and balances, your AI strategy probably won’t deliver the results you expect.
Worse, it could cost your business lost revenue, higher expenses, crisis response, and erosion of customer satisfaction and brand loyalty.
Here are some ways to narrow the gap between your long-term AI goals and the practical in-the-trenches reality of how to execute against that strategy:
Audit & Discovery — Begin with a thorough assessment of your company’s current practices, processes, and needs. Which tasks are manual and redundant, and would benefit from some automation or AI enhancement? Where are teams aligned, and where are there disconnects or pockets of resistance? Ask groups of employees where they think AI could be most helpful to them, and how, and why. Learn which managers are the most on board, and which ones are skeptical. Ensure you have a clear and realistic view of how, when, and where AI can be most helpful and meaningful for your company both now and over the longer term.
Consider Readiness, Not Just Adoption — Managers may feel pressure or be tempted to focus on key performance indicators (KPIs) like deployments and usage metrics. It’s important to also track readiness, confidence in using the tools, skills that are being developed and acquired, attitudes and perceptions, and whether the AI tools are making people more effective and helping them do better work.
Set the Strategy — Start developing internal champions and get managers involved in the strategy so they have a sense of ownership and can execute the plan more effectively. Let them and their teams propose workflows, ranked priorities, and plans for how and when they can start using some of these AI tools. Determine how much human oversight and fact checking there should be. Consider adding more rigorous reviews for high-priority or sensitive functions. Create plans for small beta tests, with time to pause and analyze results for effectiveness.
Create Space & Time — Reduce the workload for managers and employees to create more space and time for them to learn how to use AI tools effectively so it doesn’t feel like a new burden and time suck. Give them grace to learn, test, experiment, explore, and iterate. If you don’t create this capacity, even the best AI tools and plans may fail.
Create Feedback Loops — Honest, candid, thoughtful feedback on how things are actually going is vital. Create formats and channels for employees and managers to tell you what’s working well, and what isn’t. Stay open to modifications, iterations, and progress over perfection. Reassess whether the plan for rolling out AI tools matches reality, and be open to change. Any setbacks, mistakes, and problems you discover in your AI deployment are an opportunity to learn and grow.
Pay Attention to Resistance — There may be pockets of resistance to using AI, and that’s an indication that you’ve missed the mark on something. Perhaps readiness isn’t where you thought it was. Perhaps the rollout strategy is flawed. Perhaps employees are seeing errors, mistakes, and unexpected landmines and have lost trust in the outputs of AI. These caution flags provide valuable data and feedback so you can modify your approach and have greater success.
Check out this great article in The Harvard Business Review about how leaders in companies don’t always agree on how, when, and where to use AI, and that discrepancy is costing their businesses.
