Businesses are eager to tap into the power of AI of late. According to a PwC study, AI is projected to contribute $15.7 trillion to the global economy by 2030.

Despite this huge potential, a Gartner Inc. study (subscription required) highlighted that AI projects are taking twice as long to pivot from the planning stage to full-scale launches. This delay not only increases costs and reduces ROI but also impacts the time to deliver value.

In my experience of interacting with hundreds of subscription business leaders, here are the most common challenges I’ve observed that are delaying AI projects from getting off the ground.

1. Identifying The Best Use Case To Target

Deploying AI in your business requires your C-suite executives to have a deep understanding of the technology and the ability to identify key business areas that would benefit most from its application.

Subscription businesses often struggle to determine the most suitable AI use cases to kick-start their AI transformation, as they find it challenging to justify the potential ROI of the targeted use case. This difficulty can hinder financial backing for their AI project.

To secure wider buy-in, you must first build a robust business case that outlines unit economics and potential lifetime value gains from each customer. Developing such a business case demands meticulous planning and collaboration among your leaders in data, analytics, IT and BI, which can be time-consuming.

2. Deciding On Building, Buying Or Outsourcing

Amid the rapidly expanding landscape of world-leading vendors and AI partners, making a strategic decision to build, buy or outsource your AI solution requires meticulous planning—leading to delays in bringing AI into your enterprise.

Building your AI solution in-house requires a significant time investment to train your existing workforce in AI technology and acclimate them to new workflows.

Opting to buy an AI solution from an external provider requires additional time for customization to fit your specific business needs.

Furthermore, outsourcing involves a complex transfer of knowledge from external providers to your internal IT, BI or data teams, necessitating thorough planning.

3. Managing Change And Driving Lasting Business Impact

Creating business value from your AI project each year requires embedding it into your workflows with full engagement and support from your employees.

In my experience, subscription businesses often struggle to manage the human side of change to ensure their AI projects are seamlessly integrated into business processes and employees’ day-to-day activities. This challenge typically arises from the absence of a strategic change management plan. Developing a robust change management plan requires a thoughtful approach to prepare, implement and sustain the change, which is a complex task.

4. Hiring And Training Necessary Talent

If you’re confident that your AI project has a strategic competitive advantage and decide to build it in-house but lack the necessary expertise, you will need to acquire and train new resources. This talent includes technical knowledge in specific AI technologies, data science, data quality maintenance, domain expertise, and skills to monitor, maintain and govern the AI environment.

This process requires a cultural shift and takes time for your teams to embrace the change and build trust.

5. Building Necessary Infrastructure For Data Integration

Most subscription businesses focus on developing machine learning models and often encounter challenges in integrating predictions into operations. This issue frequently arises due to a lack of dedicated expertise—especially data engineers and developers—needed for building the data architecture required for integrating models into data flows.

Success in AI business cases and plans relies heavily on a robust data and analytics infrastructure. Planning for this infrastructure needs to happen well in advance to implement AI effectively without compromising data security and compliance.

6. Developing An Action Plan For Operationalizing AI

Businesses often invest heavily in building AI models and generating predictions, mistakenly assuming that this marks the end of their efforts. They frequently neglect the crucial step of devising an action plan to translate these predictions into actionable strategies for implementation across customer-facing channels.

To showcase tangible revenue gains using AI, it’s imperative to develop an actionable implementation plan that zeros in on either a specific use case or channel, enabling you to effectively use these predictions and drive customer lifetime value growth.

Next Steps
AI offers great potential to businesses that implement it effectively. To realize this potential, subscription businesses should focus on addressing the above key challenges and developing a strategy with clear outcomes. They can start by concentrating on three main tasks:

• Build an ROI-driven business case. Demonstrate the financial benefits of AI by creating a robust business case that highlights the expected return on investment.

• Implement AI on a small scale. Begin with a small-scale implementation of AI, targeting either a specific growth use case or a sales/marketing channel. This approach allows for manageable, incremental gains and valuable insights.

• Facilitate a seamless transition. Develop a results-driven road map to ensure a smooth transition from planning to implementation. This road map should focus on tangible outcomes and continuous improvement.

This article was originally published in Forbes