- In a nutshell:AI is rapidly changing the business world, but many companies are not ready for it.
- Research shows that most enterprises are at a "Standardizing" state of readiness for AI.
- Building a strong data foundation is crucial for AI success.
- Alignment between data science teams and business objectives is essential.
- Best practices for AI readiness include clean data, executive buy-in, and continuous model monitoring.
AI is changing the business world at a rapid rate. But not every company is ready for it. Pecan recently co-sponsored "The State of AI Readiness 2024," a new research report and webinar from trusted data industry research firm TDWI. Sponsored by Pecan, MongoDB, and SAP, the study analyzes results from surveys of hundreds of data professionals. The research covers the fundamental areas for AI success: organizational readiness, data readiness, skills and tools readiness, operational readiness, and governance readiness. These are many of the challenges organizations are facing in getting ready for AI. Additionally, the research provides considerations and best practices for moving forward with AI. Overall, the enterprises responding to the 2024 survey had a median score of 62 out of 100, placing them collectively at what TDWI calls the "Standardizing" state of readiness. This means they're building a preliminary plan for using AI, examining potential use cases, and perhaps even building some proofs of concept.
In this post, I wanted to share some insights from participating in the webinar conversation about the research. The research results framed a discussion of how organizations are preparing to implement AI and their current challenges — or are failing to do so.
Building a data foundation
One critical takeaway was the need for a strong data foundation. Organizations are in different stages of AI readiness, with many still grappling with data silos and integration issues. In the TDWI research, 57% of respondents disagreed with or felt neutral about the statement, "My organization has systems in place to ensure that data is easily accessible and can be integrated from diverse sources." That means that only about 4 in 10 enterprises have achieved a level of data maturity that will prepare them to start using AI. And 53% don't yet have data engineers in place to help facilitate the data pipelines needed for AI projects. In my experience at Pecan AI, we've seen that companies often underestimate the complexity of consolidating data. It requires a significant shift in culture and technology to create a solid data infrastructure that supports advanced analytics like AI.
Business and data team alignment
Another point of discussion was the alignment between data science teams and business objectives. I've noticed that without clear alignment, projects can veer off course, focusing too much on the technical aspects rather than delivering business value. As I explain in the video below, it's crucial for data scientists and business leaders to communicate and ensure that AI initiatives are directly tied to solving business problems.
This is where data analysts and business analysts can play a significant role. They are closer to the business than data science teams, and they may better understand how to answer business questions and solve challenges with the available data. As the TDWI report states:
A recent trend in predictive analytics and AI is to democratize it, in other words to open up AI to a wider audience. In some cases, this audience can include business analysts (those who build dashboards and reports). These organizations typically have a data warehouse or data lake and analysts are running BI reports from this infrastructure. Many of these analysts are ready for AI; they are interested in growing their skill set. They may be bored with simply producing dashboards. They also understand the move to AI and want to go to the next step. They understand their data and the business. … Using data scientists to fill these roles simply won’t scale.
This observation is very much in line with our thinking at Pecan: data analysts are perfectly situated to take AI projects in hand, given the right tools for success. Data scientists have fantastic skills, but there simply aren't enough of them (nor can every business afford them) to cover all of the business opportunities out there.
Pursuing generative AI and predictive AI projects
A question that often comes up is whether to focus on traditional machine learning or large language models (LLMs) like those provided by Google and other tech giants. At Pecan AI, we believe both have their place. Traditional machine learning remains valuable for structured, predictive analytics, while LLMs, although trendy, require robust infrastructure and thoughtful integration to be effective in business applications. As the TDWI report notes:
Identify the business problems to solve with AI. AI should not be done simply for AI’s sake. The right value proposition is key at the enterprise level. It is important to know why and how to leverage AI in your business.
Best practices for AI readiness
Best practices for AI readiness from a technology perspective include starting with clean, well-organized data. Historical data is crucial for training accurate models, and keeping this data updated is essential.
Additionally, once models are in production, they must be monitored continuously to ensure they provide relevant and accurate predictions. Ignoring this aspect can lead to models that degrade over time and lose their effectiveness. Developing a plan for model maintenance is also an important part of AI readiness.
In terms of organizational best practices, we emphasized the need for executive buy-in and initial pilot programs to demonstrate AI's value. The TDWI report notes, "Enterprises succeed with analytics and AI when their executives support and evangelize it across the company." Executives need to understand that AI is not a magic solution but a powerful tool that requires strategic investment and alignment with business goals.
Moving into an AI-ready future
Overall, the discussion underscored that AI readiness is a multifaceted challenge. It involves data alignment, governance, clear business objectives, and continual adaptation to new technologies. At Pecan AI, we're committed to helping organizations navigate these complexities and unlock the full potential of their AI initiatives. I'm excited about AI's future and look forward to more conversations about how we can all prepare for this transformative technology. Thank you for joining us on this journey. Want to learn more about how Pecan can accelerate your AI journey? Get in touch with us today to learn more.
