# Tell Whether Culture Investments Are Working You need to determine whether specific organizational investments -- a mentorship program, a tooling rollout, a process change -- moved the numbers before the next budget review asks for evidence. ## Prerequisites Complete [Demonstrate Engineering Progress](/docs/products/landmark-quickstart/) first. This page assumes you have a working Landmark setup with health views, snapshot data, and at least two GetDX snapshots spanning the period of your investment. ## List active and completed initiatives Start by confirming which investments Landmark is tracking. Initiatives are organizational programs linked to a target driver in your engineering standard. ```sh npx fit-landmark initiative list --manager alice@example.com ``` ```text Initiatives Code Review Standards 65% active 2025-09-30 Runbook Coverage 100% completed 2025-06-30 Mentorship Pilot 40% active 2025-12-31 ``` Each row shows the initiative name, completion percentage, status, and due date. If the output is empty, your GetDX data does not include initiative records -- check with whoever administers your GetDX account. ## Check the driver trend for a specific investment Before looking at initiative-level impact, see whether the target driver is trending in the right direction. Each initiative targets a driver from your `drivers.yaml` -- for example, a code review program targets `quality` and a runbook effort targets `reliability`. Track the driver across all available snapshots: ```sh npx fit-landmark snapshot trend --item reliability --manager alice@example.com ``` ```text Trend for reliability 2025-03-15 65 2025-06-14 71 2025-09-13 74 ``` A rising trend does not prove the investment caused the change, but a flat or declining trend after a completed initiative is a strong signal that the investment did not produce the expected outcome. Replace `reliability` with the driver ID your investment targets. Run `npx fit-landmark snapshot list` to see available snapshot dates. ## Compare a snapshot against benchmarks A trend shows direction; a comparison shows position. Check how the team's current score for the target driver sits relative to organizational benchmarks: ```sh npx fit-landmark snapshot compare --snapshot NzE4MmRk --manager alice@example.com ``` ```text Snapshot NzE4MmRk Driver Score vs_prev vs_org vs_50th vs_75th vs_90th quality 78 +6 +5 +8 -2 -10 reliability 71 +6 +3 +3 -5 -13 cognitive_load 82 +1 +10 +10 +1 -8 ``` Use the snapshot ID from `npx fit-landmark snapshot list`. The `vs_prev` column shows movement since the last snapshot -- the most direct measure of recent change. The `vs_org` and percentile columns place that change in context: a +6 that still sits below the 75th percentile means the investment is working but the team has not caught up yet. ## Measure initiative impact on driver scores For completed initiatives, Landmark correlates the completion date with before-and-after driver scores from the bounding snapshots: ```sh npx fit-landmark initiative impact --manager alice@example.com ``` ```text Completed initiatives — outcome correlation "Runbook Coverage" (completed 2025-06-30) Target driver: reliability Score before: 65 Score after: 71 Change: +6 percentile points "Onboarding Revamp" (completed 2025-04-15) Target driver: cognitive_load Score before: 79 Score after: 82 Change: +3 percentile points ``` The output pairs each completed initiative with the driver score from the snapshot immediately before completion and the snapshot immediately after. A positive delta suggests the investment correlated with improvement; a zero or negative delta prompts a deeper look. Two caveats to keep in mind when presenting this data: - **Correlation, not causation.** Other changes may have contributed to the score movement. Use voice data (see below) to add qualitative context. - **Snapshot timing matters.** If the initiative completed between two snapshots that are far apart, other factors had time to influence the score. ## Add qualitative context with engineer voice Numbers show direction; the team's own words explain why. Surface comments related to the investment's target area to ground the quantitative data in lived experience: ```sh npx fit-landmark voice --manager alice@example.com ``` ```text Voice: Alice Smith's team (latest snapshot) incident 3 comments "On-call handoffs are still rough" "Runbook coverage is improving but gaps remain" "Incident review meetings have been helpful" planning 2 comments "Sprint planning feels more realistic this quarter" "Design docs are getting better but still inconsistent" ``` Look for comments that name the investment directly or describe its effects. Three comments about runbook improvements alongside a +6 reliability delta tells a more convincing story than either data point alone. ## Verify The assessment is complete when you can answer three questions with data: 1. **Is the target driver trending in the right direction?** `npx fit-landmark snapshot trend --item --manager ` shows a rising score across the snapshots that span your investment period. 2. **Did completed initiatives correlate with score changes?** `npx fit-landmark initiative impact --manager ` shows a positive delta for each completed initiative. Initiatives still in progress show no impact data -- that is expected. 3. **Do engineer comments support the quantitative signal?** `npx fit-landmark voice --manager ` surfaces comments that reference the investment area, adding context that pure numbers cannot provide. If all three signals align -- rising trend, positive delta, supportive comments -- you have evidence that the investment is working. If they diverge, the data gives you a specific starting point for investigating why.