The Agentic Work Report (monthly)

The shift to agentic work is happening across thousands of engineering teams simultaneously, but most of them are making decisions in isolation. One team experiments with Claude Code and draws conclusions based on three developers’ experience. Another rolls out Copilot company-wide and measures success by commit volume. A third tries Cursor for a sprint and decides it is not ready because the initial learning curve was steep.

Each of these teams is generating valuable data about what works and what does not. Almost none of that data leaves the organization. The result is an industry that is collectively learning the same lessons over and over, team by team, with no shared reference point.

The Agentic Work Report exists to change that.

What the report covers

Every month, Dailybot publishes an analysis of how teams are using coding agents, drawn from aggregated and anonymized data across the Dailybot community. The report tracks several dimensions that matter for leaders making decisions about agent adoption and governance.

Adoption metrics

How many teams are actively using coding agents, and how is that number changing? The report tracks adoption rates segmented by team size, industry, and geography. It also measures the depth of adoption: are teams using agents for simple tasks like test generation, or have they expanded to complex workflows like multi-file refactors and architectural changes?

Early data shows that adoption is not uniform. Smaller teams tend to adopt faster because they have fewer governance constraints and shorter decision cycles. Larger organizations adopt more cautiously but scale more quickly once they commit, because they can spread best practices across multiple teams.

Agent-to-human contribution ratios

One of the most interesting metrics is the ratio of agent-generated work to human-generated work within a team’s output. This ratio varies dramatically across organizations, from teams where agents contribute less than ten percent of commits to teams where agents handle over half of the routine coding.

The report tracks how this ratio evolves over time within individual teams and across the industry. Early patterns suggest that the ratio increases steadily in the first three months of adoption, then stabilizes as teams find the natural boundary between what agents handle well and what requires human judgment.

Common workflows and use cases

What are teams actually using agents for? The report catalogs the most common workflows, ranked by frequency and by the satisfaction teams report with agent performance in each category.

Current top use cases include test generation, code refactoring, documentation updates, boilerplate generation, and bug investigation. Less common but growing use cases include architectural exploration, code review assistance, and migration projects.

Reporting and visibility patterns

Since Dailybot sits at the intersection of human and agent communication, the report has unique insight into how teams are handling agent visibility. It tracks how often agents report their progress, what format those reports take, and how teams that implement structured agent reporting compare to teams that do not.

Early observations suggest a strong correlation between agent reporting practices and team satisfaction with agent output. Teams where agents report regularly tend to trust their agents more, give them more autonomy, and report fewer incidents related to agent-produced code.

Why benchmarking matters

Engineering leaders are making high-stakes decisions about agent adoption with very little external data. Most of the available information comes from vendor marketing (which emphasizes best cases) or developer social media (which emphasizes edge cases, both positive and negative). Neither source provides the kind of representative, data-grounded perspective that leaders need.

The Agentic Work Report fills that gap by providing industry benchmarks that leaders can compare against their own organization’s experience. If your team’s agent adoption rate is significantly below the industry average, that might indicate organizational barriers worth examining. If your agent-to-human ratio is much higher than peers, that might signal a need for stronger review processes.

Benchmarks do not tell you what to do. They tell you where you stand, which is the starting point for informed strategy.

What we are seeing early on

While the report’s full dataset grows with each monthly edition, several early trends are worth noting.

Agent adoption is accelerating. The percentage of teams using coding agents daily has increased month over month since the report began tracking. This growth is driven by improvements in agent capabilities, but also by organizational learning: as more teams share their experiences, the barriers to adoption lower for everyone.

Visibility is the top concern for managers. When asked about their biggest challenge with agent adoption, engineering managers consistently rank visibility above quality, security, and cost. Managers feel they cannot effectively govern what they cannot see, and most agent tools do not provide the team-level visibility that managers need.

Teams with structured reporting outperform. Teams that implement structured agent reporting through tools like Dailybot report higher satisfaction with agent output, faster identification of issues, and better coordination between human and agent work. This correlation holds across team sizes and industries.

The learning curve is shorter than expected. Most teams report reaching productive agent usage within two to three weeks, shorter than the months-long adoption timeline many leaders feared. The primary bottleneck is not technical skill but organizational readiness: having clear policies, review processes, and visibility infrastructure in place.

How to use the report

The Agentic Work Report is designed to be actionable, not academic. Each edition includes specific takeaways organized by role.

For CTOs and VPs of Engineering, the report highlights strategic trends and benchmarking data that inform adoption strategy and investment decisions.

For engineering managers, the report provides practical insights about workflows, team structures, and governance practices that are working well for similar teams.

For individual developers, the report surfaces tips, tool comparisons, and workflow patterns that can improve daily productivity with agents.

Staying current

The agentic era is moving fast. Decisions made based on six-month-old data are decisions made on stale assumptions. The monthly cadence of the Agentic Work Report is designed to keep leaders current without overwhelming them with noise.

Each edition builds on the previous ones, tracking trends over time rather than presenting isolated snapshots. Over the course of a year, the report creates a longitudinal view of how the industry’s relationship with coding agents evolves, which is far more valuable than any single data point.

The Agentic Work Report is available through Dailybot. If your team is already using Dailybot for check-ins and agent reporting, your aggregated data contributes to the report’s insights, and you get early access to each edition.