Isn't this kind of measurement reductive — engineers don't just ship PRs?

Tycoon Agent knows. She measures PRs *and* code review hours *and* incident response *and* mentorship signals (junior PRs they reviewed, docs they wrote). The score is multi-dimensional. The point isn't to reduce a person to a number — it's to give you signal you'd otherwise have to fly blind on. You still make the call.

What if I don't track output in Linear / GitHub / Apollo?

She works with what you have. With just payroll + Slack + email volume she can still flag flight risks and obvious outliers, but the cost-per-output analysis needs at least one objective output source per role. Most teams add Linear or Apollo within 30 days of running this for the first time.

Can I see Tycoon Agent's reasoning for a specific person?

Yes. Ask 'why did Tycoon Agent flag X as underperforming' and she shows the underlying numbers, comparisons, and time series. You can override any call ('J is on a special project — exclude from output ranking') and she'll respect it going forward.

Does Tycoon Agent share this with my team?

No. Never. This brief is a Lark DM only you can see. Tycoon Agent never tells team members their score, never logs it in a shared place, never references it in public chat. It's a tool for you to make calls, not for performance management theater.

Ask Tycoon Agent: Is my team's output worth what I'm paying?

You'd think this is the question only a CFO can answer — Tycoon Agent reads payroll, GitHub, Linear, and revenue data and gives you the unvarnished version.

The short answer

Tycoon Agent answers "is my team's output worth the cost" by joining payroll cost (from Brex/Gusto) with measurable output per role: PRs merged + reviewed (GitHub), features shipped (Linear), revenue moved (Stripe attributed to launches), customer support tickets resolved (Intercom), content published (CMS), deals closed (Apollo/HubSpot). She computes a cost-per-output ratio per person, normalizes by role and tenure (month-1 hires don't ship like month-12), and benchmarks against industry medians. Output is a confidential one-pager: who's outperforming their cost, who's matching, who's underperforming, and who's a flight risk based on Slack activity drop. She doesn't give performance reviews — she surfaces signal you can use to coach, reassign, or have a hard conversation. You stop guessing if your eng team is over-hired or your sales rep is worth the OTE.

How Tycoon Agent actually does it

Pull payroll + role data
Brex/Gusto for cost-per-person-per-month including loaded cost (benefits, equity vesting). Tag by role (eng, sales, design, ops).
BrexGusto
Pull output per role
Eng: GitHub PRs merged, code review hours, incident response. Sales: Apollo/HubSpot deals + revenue closed. Design: Figma files shipped + frontend impact. CX: tickets resolved + CSAT.
GitHubLinearApolloHubSpotIntercom
Normalize by tenure + role
First 90 days = ramp, weight 0.4x. Senior IC vs junior = different output expectation. Compare against the role's industry median, not against each other.
Compute cost-per-output
Ratio per person. Plus a flight-risk signal: Slack message frequency drop, calendar density change, last 1:1 sentiment if logged.
Write the confidential brief
Lark DM (just to you): outperformers, matchers, underperformers, and flight risks. Each with reasoning + one suggested action.
Lark

What it looks like in chat

YouIs my team's output worth what I'm paying?

Tycoon AgentConfidential — read once and decide. **Outperforming (keep, retain hard):** - Sarah (Eng, $185k loaded). 38 PRs/mo merged, owns the billing rewrite, 0 P1 incidents. Cost-per-PR is 40% below role median. Flight risk: low. - Marcus (Sales, $220k OTE). $1.3M ARR closed last quarter, 3.2x his loaded cost. Top of leaderboard. **Matching expectations:** - 3 of your 6 engineers. Solid, no red flags. **Underperforming:** - Junior eng J ($140k loaded, 4 months in). 8 PRs/mo, 60% bug fixes. Behind ramp curve. Action: 1:1 this week, set 60-day output target. - Marketing manager K ($165k loaded, 9 months in). 1 campaign shipped this quarter, no measurable revenue lift. Either reassign to growth experimentation or have hard conversation. **Flight risk:** - Designer D. Slack messages down 60% over 30 days, declined last 3 team events, opened LinkedIn 12x last week (visible in connection notifications). Worth a candid 1:1 before they leave.

What you get back

A confidential weekly Lark DM (only you see it) ranking your team by cost-per-output, flagging outperformers to retain and underperformers to address, plus flight-risk signals.

Cadence: Monthly deep-dive, plus weekly flight-risk pings if a signal triggers.

Is my team's output worth what I'm paying?