58 | Getting on Board with AI: Top-Down *and* Bottom-Up
Preface
Happy New Year!
StrefaTECH is back, and I’m excited to share some of the thoughts, info, and cool tools I ran across during the holiday hiatus.
Without further ado, Andiamo (we’re off)!
Introduction
I’ve been thinking a lot about the questions I keep hearing about AI. They’re truly all over the place, which has influenced my somewhat eclectic choice of topics to explore in these StrefaTECH articles. I’m working on cataloging them into themes (yes, with the help of GenAI), which I hope will provide a more deliberate structure for upcoming articles.
In the process, though, something occurred to me about organizations and their struggles with what to do with AI … and that is that the answers to many of the questions depend on the perspective of the person asking. In particular, what makes sense often depends on whether you’re coming from a leadership perspective, wanting to guide AI use in your organization responsibly, or you’re on staff looking to use AI to get your work done better.
I’m going to collect those in leadership positions into having considerations related to a “top-down” approach to AI adoption. Those more in the trenches, so to speak, may be in positions to promote a “bottom-up” approach. Either way, many of the fundamental questions and needs related to this new technology apply; but there are also numerous notable differences!
Top-Down and Bottom-Up
In a top-down approach, AI implementation is a strategic decision made by the organization's leadership. This approach is characterized by its structured nature, with AI tools being deliberately integrated into specific areas to enhance overall performance and align with the organization’s objectives.
Conversely, the bottom-up approach is more spontaneous and employee-driven. It empowers individual staff members or teams to explore and adopt AI tools for their specific needs. This is marked by its flexibility and potential for grassroots innovation, potentially leading to unexpected and highly beneficial applications of AI within the organization.
I advocate for building roadmaps for each of these. If you can establish a firm understanding at all levels that both approaches are valuable, you can accelerate the broad and safe use of AI. And if you create roadmaps for the areas of the organization that will be exploring AI, you can coordinate and capitalize on the benefits of both the top-down and bottom-up approaches.
The Importance of Leadership Understanding and Guidelines
Regardless of the approach your organization is taking, it’s vital for leaders to understand both the potential and the limitations of AI. This includes recognizing that AI is a tool to aid human decision-making, not replace it. Setting clear guidelines is crucial, too. These guidelines should address ethical concerns, data privacy, and ensure that the use of AI aligns with the organization's mission and values.1
The Top-Down Approach – Deliberate and ROI-Focused
In the top-down approach, decision-making is centralized, with a clear focus on maximizing the return on investment (ROI) of the change in processes and technology. Think of a sports team—there needs to be a game plan, led by a coach who understands the needs and capabilities of the team, leads the creation of the plan, and collaborates with the team for effective execution of the plan.
Notable considerations:
Your organization absolutely must have an AI Acceptable Use Policy in place2, and it should be reviewed and updated regularly based on experience with each notable initiative.
It’s ideal to start with AI projects with demonstrable benefits and achievable targets, and remember … smaller is better.
The pros and cons of AI tools under consideration should be weighed based on impact and costs. The timing of major tool investments can be intricate, given the rapid changes in AI capabilities. This is an area where securing advice from someone knowledgeable can be worthwhile.
Monitor early rollouts carefully for risk areas such as data privacy exposure and bias.
Projects should have broad visibility and clear endorsement from senior leadership.
Including goals related to increasing the organization’s understanding of AI technology is not only appropriate but highly valuable. This isn’t just implementing another database; it’s potentially changing how staff and clients relate to information in radical ways.
For example:
AI for Donor Management: By utilizing AI to analyze extensive donor data, nonprofits can gain insights into donor behaviors, preferences, and patterns. This strategic application not only enhances fundraising efforts but also enables more personalized donor engagement. It could be justified by leading to better donor retention and increased funding but it also may provide useful unexpected insights that should be cataloged when assessing the value of the initiative. Note: Any use of AI, but particularly when dealing with sensitive information such as donor records, must be vetted for protecting privacy.
AI Chatbots: An organization might invest in the development of a chatbot that can be placed on its website to handle common questions. Commercial tools to create these are becoming easier to configure and less costly to license, but any such new offering would require an investment of time and money, as well as careful testing and risk assessment.
A standardized approach in the top-down method also ensures that everyone in the organization is on the same page regarding AI tool usage, facilitating easier management and evaluation of these technologies. In particular, communications in both directions—from leadership and the project team to the organization and from the users back to leadership—and training should be considered crucial elements of each project.
The Bottom-Up Approach – Flexible and Innovation-Driven
The bottom-up approach, in contrast, is less about structured implementation and more about organic exploration and adoption of AI. This approach encourages individual initiative and can lead to innovative uses of AI that may not have been considered in a top-down strategy.
Notable considerations:
Your organization absolutely must have an AI Acceptable Use Policy in place, and there should be clear channels for individuals to provide feedback on the policy as they consider using AI tools.
Managers need to have a solid grasp of what’s acceptable and what’s not allowed when it comes to “trying out” new tools.
Striking the balance between allowing (or even encouraging) experimentation and avoiding potentially costly mistakes is difficult, but crucial to pay attention to. There should be an environment of partnership in learning, which sometimes may come from mistakes. The organization’s culture should drive how much latitude individuals are given to experiment—again, within the bounds of the AI Acceptable Use Policy—and how both triumphs and failures are capitalized on to advance the organization’s maturity in using new technology.
Communicating is key. No one should feel that they need to hide their use of a tool, and everyone should benefit from the learning that comes from innovation and experimentation.
For example:
AI for Personalized Communication: Individual staff members might employ AI-powered tools to create more personalized and effective communications with beneficiaries, volunteers, or donors. These tools can help tailor messages based on individual preferences, leading to higher engagement rates.
AI for Efficient Administration: Employees may choose AI-driven tools to improve their day-to-day work, such as using AI for scheduling, task management, or even automating routine administrative tasks. This grassroots adoption can significantly boost productivity and job satisfaction without the need for large-scale organizational changes or investments.
Conclusion
When adopting a top-down strategic implementation or a bottom-up organic integration, your organization’s operations and impact can be significantly enhanced by AI; and in fact, you should be taking on both adoption paths. Thoughtful planning, ethical safeguards, and an eye toward broad buy-in can lead to AI boosting nonprofit productivity and innovation. The key is matching the strategy to your organization’s culture and needs.
And when it comes to deciding what’s acceptable, what’s out-of-bounds, and how much you’ll do to plan a top-down strategy and to foster an environment of bottom-up innovation…
Make Good Choices
See StrefaTECH article about creating and rolling out an AI Acceptable Use Policy.
Also, listen to this Nonprofit Nation podcast where host Julia Campbell interviews Kim Snyder of RoundTable Technology about “Why You Need An AI Use Policy.” They cover a host of relevant topics—it’s the best summary of this topic I’ve run across!