Cost Analysis & Salary Modeling

Engineering is the most expensive department in most tech companies. ShipLens connects salary data to commit-level impact analysis, giving CTOs visibility into the real cost of engineering output — not just headcount, but cost per unit of value delivered.

Purpose

Answer questions like:

• "What does a point of impact actually cost us?"
• "Which squad delivers the most value per R$ spent?"
• "Are we over-investing in a domain that isn't delivering proportional impact?"

These are uncomfortable questions. But a CTO who can't answer them is managing by intuition, not data.

Salary Configuration

Salary data is stored per contributor as a monthly value in the company's local currency.

Field	Type	Description
contributor_id	reference	The contributor this salary belongs to
monthly_salary	integer	Monthly salary in local currency (e.g., BRL)
effective_from	date	When this salary became effective

Encryption: Salary values are stored encrypted at rest using Cloak with AES-256-GCM. The encryption key is separate from the database credentials.

Access control: Salary data is visible only to users with the admin role. Non-admin users see cost analysis pages with salary data redacted — they can still see relative efficiency comparisons (percentages, rankings) without seeing absolute salary figures.

Cost Metrics

Cost per Impact Point

The primary efficiency metric. Lower values indicate more cost-effective engineering.

$$\text{cost\_per\_impact}=\frac{\text{total salary cost (period)}}{\text{total impact points (period)}}$$

Where impact points are the sum of all commit scores (V2) in the period.

Cost per Impact Point	Interpretation
Decreasing over time	Team is becoming more efficient
Stable	Consistent output relative to cost
Increasing over time	Output declining relative to cost — investigate

Impact per R$1,000

The inverse view — how much value does each unit of spend produce? Higher is better.

$$\text{impact\_per\_1k}=\frac{\text{total impact points}\times \mathrm{1,000}}{\text{total salary cost}}$$

This metric is useful for board-level reporting where "we generate X impact points per R$1,000 invested in engineering" is more intuitive than cost-per-point.

Cost per Domain

How much of total engineering salary goes to each business domain:

$$\text{domain\_cost}=\sum _{\text{contributors}}(\text{monthly\_salary}\times \frac{\text{commits in domain}}{\text{total commits}})$$

A contributor's salary is proportionally attributed to domains based on their commit distribution.

Cost per Squad

Aggregate salary cost per squad, compared against total impact delivered:

$$\text{squad\_efficiency}=\frac{\sum \text{squad member salaries}}{\sum \text{squad impact points}}$$

Cost per Sprint/Period

Total engineering cost within a user-defined period, broken down by:

• Squad
• Domain
• Commit type (feature vs fix vs refactor vs chore)

Squad Cost-Effectiveness Comparison

Squads are compared on a normalized basis using impact per R$1,000:

Squad	Monthly Cost	Impact Points	Impact / R$1,000
Payments	R$ 80,000	450	5.63
Onboarding	R$ 45,000	310	6.89
Platform	R$ 60,000	280	4.67

TIP

Cost-effectiveness comparisons between squads should be interpreted carefully. A platform team maintaining critical infrastructure may have lower impact scores but higher organizational value. These numbers inform conversations — they don't replace judgment.

Over-Investment Detection

ShipLens flags domains where cost is disproportionately high relative to impact:

$$\text{over\_investment\_ratio}=\frac{\text{domain cost \%}}{\text{domain impact \%}}$$

Ratio	Interpretation
< 0.8	Under-invested — high impact, low cost allocation
0.8 - 1.2	Balanced — cost proportional to impact
1.2 - 2.0	Potentially over-invested — review allocation
> 2.0	Significantly over-invested — investigate

A ratio of 2.0 means the domain consumes twice the cost share that its impact share would justify.

Privacy

Salary data is the most sensitive information in ShipLens. The privacy model:

Data	Admin	Manager	Viewer
Individual salaries	Visible	Hidden	Hidden
Squad total cost	Visible	Visible	Hidden
Cost per impact point	Visible	Visible	Hidden
Relative efficiency (%)	Visible	Visible	Visible
Over-investment flags	Visible	Visible	Hidden

Routes

/c/:slug/cost-analysis       # Cost metrics dashboard
/c/:slug/settings/salaries    # Salary configuration (admin only)

The cost analysis page shows:

• Company-wide cost per impact point (trending over time)
• Impact per R$1,000 (trending over time)
• Squad efficiency comparison table
• Over-investment alerts
• Period selector for sprint-level analysis