Hype to ROI: Applying AI to Your Transformation

Recent advances in AI are no doubt top of mind for those planning (or in the throes of) digital transformation. But sorting the hype from the ROI can be difficult. There is nothing wrong with technology-driven innovation, as long as it is tied to strategy, business outcome and provides validated results. However, implementing shiny tech for its own sake is like throwing darts in the dark; you might get lucky, but most likely you’ll be asked to leave.

Follow these steps to ensure you’re setting yourself up for success.

1. Start with strategy: Don’t put tactics in the driver's seat. Go back and look at your strategy. Remember what you’re trying to achieve? Clear on the major pain-points, opportunities and obstacles you’re facing? Great, you’re already in-front of the majority.

2. Understand what AI can do: Now you’ve got the big picture in mind, educate yourself and your team on the latest AI technologies, capabilities and applications to your industry. For my fellow Trailblazers, check out the lists of AI Trailmixes on Trailhead. Most of us are saturated in AI content. If you are starting here, see step 1.

3. Identify and validate use cases: The fun part for the innovative and entrepreneurially-minded. Assess priority processes, workflows, and data assets to identify potential AI use cases. Look for areas where AI can automate, predict, classify, navigate or generate. Bring together a diverse group of stakeholders and a great facilitator, and don’t forget to validate the value of your ideas.

4. Assess your data readiness: AI relies heavily on high-quality and relevant data, so you need to assess your data infrastructure, quality and availability. Understanding what you need to collect, cleanse, integrate and govern, brings your ideas into the realm of reality.

5. Proofs of technology: POTs are focused on technical validation, complimenting the data feasibility assessment in the previous step. They should cover the performance, compatibility and scalability of proposed ideas, and capture data that will form the foundation of the future projects (if feasible) or pivots (if unfeasible).

6. Prioritise and plan: A roadmap into the relative unknown should start with a quick-win to prove value. Your organisation is paying close attention. Pick something small that demonstrates positive value. Consider the change impacts and avoid boiling the ocean until you’ve got the political capital required to align the organisation.

7. Build your team: The specifics of your organisation (and your plan) will determine how this team comes together. You’ll likely need a diverse set of skills covering data science, AI engineering, domain expertise, design and business strategy. Avoid any empire building instincts where peak capacity or specialism is required. When picking platform partners, be careful. Be sure to understand how the data you own and are generating will be treated.

8. Monitor, evaluate and evolve: Many fear the accountability, but failing to measure restricts your ability to sing your successes and expand your efforts. Measurement also provides the opportunity to improve and pivot. This is a data initiative, no excuses.

While a simplified version of a full approach, the above will help you make the most of AI opportunities and avoid common pitfalls. If you’re working with the world’s #1 AI+Data+CRM platform (Salesforce), get in touch with our Professional Services team to ensure you achieve maximum value in the shortest amount of time.

Good hunting.

Simon Carter - Business Transformation Director (Salesforce)

Connect with me here