AI Readiness at Ippon, Part 2 : What We Learned

In Part 1 of this series, we established a clear baseline for AI adoption at Ippon. If you missed it, read the full article here. We discussed the design of our study, the measures we used, and early indicators of return on investment. Now, let’s take a deeper dive into what the data actually revealed about how consultants used AI day to day.

Over the twelve weeks, patterns emerged that went beyond hours spent or saved. We saw shifts in behavior, changes in sentiment, and differences across practices and project environments. The overall trend we found was that AI use correlated with measurable productivity but also revealed meaningful variation across individuals. While some consultants used AI heavily throughout, others found usage fluctuated based on project type, client policy, or workload.

Still, the directional impact was significant. Statistical analysis showed a positive trend over time (average p-value 0.011, significance at p < 0.05), with the most substantial improvements emerging in the final month. This suggests that as consultants gained familiarity with AI tools, their efficiency increased.

Beyond the numbers, the study captured how AI tools affect the daily experience of our consultants. The sentiment data indicate that the benefits of AI extend beyond pure time savings to include reduced frustration and improved focus on high-value work.

Perceived Impact of AI Tools

When participants were asked about the perceived impact of AI tools, 70% agreed or strongly agreed that AI tools made them more productive and less frustrated. A similar share felt AI helped them focus on higher-value work. Slightly fewer reported that AI helped with repetitive or “stuck” tasks, suggesting opportunities for more targeted training or better tool integration. The data showed moderate correlations between AI use and increased productivity (0.40), reduced frustration (0.33), improved focus (0.36), and faster resolution of roadblocks (0.30). This connection between tool usage and improved work experience is a critical component of successful AI adoption.

How Ippon Uses AI

Our consultants are leveraging a diverse set of more than 15 AI tools to tackle different tasks. While the toolkit is broad, usage is highly concentrated in two main categories, which together account for over 93% of all tool mentions. The majority of usage fell into two main categories: conversational AI & chatbots (around 60% of mentions) and code assistants (about 33%).

Conversational AI tools, such as ChatGPT and Gemini, were primarily used for writing, editing, and ideation tasks. This included drafting proposals, reports, and emails; editing or refining technical documentation; and summarizing notes, creating outlines, or brainstorming ideas. Code assistants, by contrast, tools such as GitHub Copilot and Cursor, were used for writing and refactoring code, debugging and testing, and generally for understanding unfamiliar codebases.

Beyond these two categories, consultants also leveraged AI for research, data analysis, and learning new technologies, which are tasks that reflect AI’s growing role as both a problem solver and a knowledge accelerator.

The differences across the areas of practice were also apparent: product engineering showed the highest average AI usage over the twelve weeks and the greatest diversity of tools used, whereas product management showed the highest time spent using AI per participant, approximately 6 hours per week.

Interestingly, the depth of engagement varied. Conversational tools tended to be used in short bursts, minutes at a time, while code assistants often supported longer, deeper sessions throughout the week, as might be expected from tools processing deeply to produce code.

Benchmarking Against the Industry

To put Ippon’s results in context, we compared them with data from the 2025 DORA State of AI-Assisted Software Development report, which surveyed nearly 5,000 professionals.

The data shows a clear comparison in engagement. At Ippon, the largest group of users (45%) spends 1–3 hours per week using AI. This is similar to the largest industry segments that report using AI "a moderate amount" (37%) or "a little" (30%). The comparison also showed that Ippon’s consultants use AI in a manner similar to the industry average. For example, more than 60% of Ippon consultants used AI for code development or debugging, compared with 50% across the industry. Over one-third used AI for writing, documentation, or content creation, above the benchmark but very comparable, and research and summarization use also matched or exceeded peer levels.

These findings reinforce Ippon’s position as an early adopter in the enterprise AI landscape and as a company that is ahead of the curve, already experimenting across use cases that others are only beginning to explore.

Barriers and Realities

While the study underscored clear benefits, it also revealed important barriers to sustained AI usage. Some weeks simply didn’t offer tasks where AI provided obvious value. This reflects natural variation in consulting projects rather than disengagement. In certain engagements, client or government policies explicitly restricted the use of AI tools; for example, a few consultants reported not using them for certain weeks because “AI tools (are) not permitted on state machines.” These limitations highlight the need for ongoing education and collaboration with clients around responsible AI adoption. Consultants noted that during high-intensity weeks, such as migrations or client kickoffs, AI use declined, not from disinterest but from a lack of time to explore. A few users reported tool access restrictions or incompatibilities that prevented AI from being fully integrated into their workflows. Together, these factors illustrate that becoming “AI-ready” is not just about tool proficiency; it’s about aligning people, policy, and process.

In Part 3, we’ll examine the road ahead and the strategic recommendations we can now make based on our findings.