The AI Maturity Curve: Measuring AI Adoption in your Organization

Nvidia’s Jensen Huang is calling 2025 “The Year of the Agent.”??

Marc Benioff is welcoming the “Agent Era” – remarking that Salesforce has seen enough productivity gains from AI that they’re debating whether they need to hire more software engineers this year.

Sundar Pichai announced Gemini 2.0 in December, calling it the beginning of a new generation of models built for the "agentic era."

But many others say they feel disappointed by AI adoption.?

According to a recent study from BCG Research, 74% of companies report they have yet to show tangible value from their use of AI. And Slack’s Fall 2024 Workforce Index indicates that AI adoption is slowing due to uncertainty and training gaps.

Executives want to know where their org falls on the spectrum – and they’re asking People Analytics to give them the numbers.? What % of the organization is AI-ready?? How much work is being augmented by AI?? Have we made productivity gains?

Putting a number to AI adoption and proficiency, however, is fraught with difficulty.? In many organizations, AI tools aren’t fully standardized, so one person may prefer ChatGPT while another likes Claude.? Even when companies have go-to AI tools in-place, the reporting from Google Gemini or Microsoft Copilot is limited – with minimal team or task-level breakdowns and no good way to export into a dashboard.?????

We’ve got a better way.? When it comes to AI, companies should measure themselves against the AI Maturity Curve.??

Here’s how that might look:

Stage 1:? Adoption

Companies in the Adoption stage of AI are focused on uptake.? Their central question:? “How many people are using AI each day / week / month?”

At this stage, your executives are looking for a straight count.? And you can probably piece that together using audit logs from AI tools like Copilot, Gemini, or Salesforce’s Agentforce.

To really understand adoption however, you’ll want to start cutting that usage by department, team, and role.? Here, we like to overlay AI audit logs onto HR data to do a cohort analysis that allows us to understand the impact of things like AI training or manager intervention on the team's adoption of AI features.

?

Stage 2:? Proficiency

Once you’ve got a reasonable chunk of people using AI on a routine basis, then you’ll want to understand the impact it’s having.? Now, your central question has shifted:? “What percent of work being done is assisted by AI?”

AI Proficiency is a more complicated metric, but it’s also one that gets closer to the heart of things – is AI making work easier?? If yes, then that’s going to start taking care of the adoption problem for you;? people see their colleagues being more productive so they follow suit.

To get to those success stories though, you’ll need to look at an overall set of activities:? Email, Slack, Teams, Meetings, Salesforce etc.? Then, you could calculate what % of those activities contain some form of an AI assistant.? For example, you could audit calendar data to see what % of meetings have an AI notetaker invited.??

Stage 3:? Leverage

As you get people adopting AI and proficiently using it, our mind turns to productivity gains.? The central question shifts to:? “Are we getting more done in the day with AI than we were without it?”

There are two potential ways to think about measuring productivity gains:

1. At the team level, you could look to detect if more is getting done in a given day with AI than was done when it wasn’t used.? For example, you might see 20% more documents written, 10% more emails sent, 40% more action items tracked.

1. Alternatively, you might estimate the time savings from AI.? For instance, has time spent writing emails fallen since the implementation of an AI email assistant? Or is there less busy work being done or less time being spent in redundant meetings??

Notably, these metrics miss any measure of quality.? But you could layer on quality signals from surveys and performance reviews to understand how AI has impacted the caliber of output, alongside the quantity & speed.??

To really understand how your org is using AI, tool log-in data isn’t enough.? You need to take a holistic view of how AI is being used across your team’s workflow.? And to do that, you’ll need to get access to the IT metadata that captures the story of how digital work is getting done in your company.

Getting Started

Unfortunately, accessing IT metadata can be challenging at first.? It tends to come from a variety of different sources:? Microsoft Office, Google Apps, Github, Salesforce, JIRA, and Slack to name a few.? Once gathered, the data often needs to be aggregated and normalized before it’s useful.

At Worklytics, we work with Fortune 500 companies to give executives easy access to their existing IT metadata.? Our pre-built application connectors make accessing and cleaning data easy.

For your organization to move beyond the adoption stage and fully capture the productivity benefits of AI, you must be able to develop targeted intervention strategies and accurately measure ROI —and this can only be achieved with clean, actionable data. Learn more about how to centralize your SaaS data here.