AI dollar impact

The AI dollar impact tab on the AI Impact report translates your organization’s AI investment into a concrete dollar figure. Rather than debating whether AI is “worth it” in the abstract, leaders can see a financial picture of what AI is producing, what it costs, and what the net is—making board-level reporting and budget conversations significantly easier.

This tab requires both AI Code Insights (deployed to developer machines) and AI time savings responses from DX snapshots.

Dollar impact tab in DX

When to use AI dollar impact

This report helps leaders:

  • Quantify the financial return on AI tools — Move beyond adoption metrics to a net dollar figure that accounts for both the value AI creates and the costs it introduces.
  • Justify AI tool spend — Use concrete, defensible numbers for budget conversations, board reports, and ROI reviews.
  • Identify where AI delivers value vs. where it doesn’t — Break down impact by team to see which parts of the organization are getting the most return and where oversight costs are eating into gains.

How net AI dollar impact is calculated

The headline metric is an annualized net dollar figure that combines four components:

Net AI dollar impact = Efficiency gain + Agent output − Oversight cost − AI tool spend

Efficiency gain

The dollar value of time saved by developers using AI tools. This is derived from DX snapshot responses where developers self-report how many hours per week AI tools save them.

The calculation factors in:

  • Average FTE salary (configurable in report settings, default $150,000)
  • Developer count from snapshot respondents
  • AI adoption rate — the share of developers actively using AI tools (based on your configured “active AI user” tags)
  • Percentage of time spent on development (configurable, default 65%)
  • Self-reported hours saved from snapshot behavioral responses

Agent output

The dollar value of work completed autonomously by AI agents, estimated by treating agent-authored PRs as human-equivalent work.

The calculation factors in:

  • Agent PR count — PRs where the code is AI-authored (bot-authored PRs count fully; partially AI-authored PRs are prorated by their AI code percentage)
  • Median human work time — the median time from first commit to ready-for-review across human-authored PRs in the same period, used as a baseline for what this work would have cost a human
  • Hourly salary derived from the configured FTE cost

Oversight cost

The cost of reviewing and fixing AI-generated output. This is subtracted from the total because human oversight of AI work is a real cost that offsets the gains.

The calculation factors in:

  • Review time — estimated from the number of review comments on agent-authored PRs (approximately 5 minutes per comment)
  • Rework from reverts — reverted agent PRs multiplied by the median human cycle time, representing the cost of undoing and redoing failed AI work

AI tool spend

The annualized cost of AI tool licenses, configured in report settings. This is split proportionally across the organization based on headcount in scope.

Breakdown by team

Below the summary, a breakdown table shows all four components and the net impact for each team (or other aggregation). Use this to compare:

  • Which teams are seeing the highest efficiency gains
  • Where agent output is most concentrated
  • Where oversight costs are highest relative to output

The table supports CSV export and can be expanded to drill into sub-teams.

Breakdown modal

Clicking View breakdown on the net AI dollar impact figure opens a detailed receipt-style view showing each component with its inputs:

  • Human efficiency gain — engineer count, AI adoption %, and speed boost multiplier
  • Human-equivalent agent work — agent PR count and average cycle time
  • Human oversight & remediation — average review time and revert rate
  • AI tool costs — configured tool names and annualized spend

Report settings

The Dollar impact calculation depends on several configurable inputs, accessible via the settings icon on the AI Impact report:

Setting Description Default
Average FTE salary Fully loaded annual cost per developer, used in all dollar calculations. $150,000
Dev time % Percentage of a developer’s time spent on development work. 65%
AI tools List of AI tools with their annual license costs. $0 each
Active AI user tags Tags that define which developers are considered active AI users. —

If AI tool costs are all set to $0, an admin banner will prompt you to configure them so the total AI spend component is reflected accurately.