How it works
The full cycle a Strivers review period runs — from setting goals to a reward number finance can defend, plus the live view leadership gets throughout.
Org admins set organizational goals. Managers and executives cascade team goals underneath them; individual contributors add their own. AI suggests goals grounded in the org's priorities and flags vague or misaligned ones before the cycle locks.
Executives review a set of hypothetical employees — each with a different mix of impact across the org's goals — and decide how they'd split a raise budget between them. A regression infers each goal's actual weight from those calls, revealed fresh each cycle instead of typed in once.
Self, peer, and manager reviews share one flow: a short interview per goal, an AI-drafted narrative, and an inline chat to refine it before submitting. Peer and manager reviewers also get an AI-suggested score per goal, with a rationale, to accept or override.
Once reviews are in, every employee gets an alignment score: individual contributors from their own goal results against the learned weights; managers from a blend of their own results and their team's average, rolled up through the org chart. One number, reflecting how well their output matched what the org values.
Reward shares are computed from alignment scores within each manager's group. An amplification dial you control decides how much more top performers pull from the pool — turn it up for winner-take-more, down for flat.
Leadership gets a standing dashboard of alignment scores by employee, team, and goal — the same live answer to what the org actually rewards that the reward math itself runs on.
Alongside the review cycle, org admins run recurring, org-wide pulse surveys — three standing prompts (what's going well, what needs to improve, what should stay the same) plus any custom questions, on a cadence as tight as weekly. Rounds auto-open, auto-dispatch by email, and auto-close; results are fully anonymized and visible to org admins and executives as aggregates or anonymized per-respondent bundles.
The full AI surface
Every AI surface reads from the same org-level writing spec, so generated content sounds like your org, not a generic assistant.
Goal suggestions
Grounded in the org's stated priorities, not generic templates.
Conversational drafting
Self, peer, and manager reviews — drafted from a structured interview, refined by chat.
Goal-quality critique
Flags vague or misaligned goals before the cycle locks.
Org writing spec
One standing voice profile every AI surface reads from.
Bring your own AI
Strivers runs its own Model Context Protocol (MCP) server, so Claude or any MCP-compatible AI client can read your goals, reviews, notes, and rewards data on your behalf — no export, no copy-paste, no separate API key. Sign in with your existing Strivers account; every tool call is scoped to your role and organization, same as the app itself.
OAuth-secured
Connects through the same login as the app — Google or your org's SSO. No API keys to manage or leak.
Role-scoped access
Every tool call runs through the same org and role checks as the app itself — managers see their team, admins see the org.
Goals, reviews, notes, rewards
Ask Claude about your goals, feedback history, team notes, or reward data directly, without opening the app.
Works with Claude Desktop and Claude.ai
Add Strivers as a remote MCP connector — find your server URL under AI Config in account settings.
Try now — tell us about your goals and we'll show you where they map.