Gemini for Independent HR Consultants: Intermediate Strategies for Building a Performance Review System That Managers Will Actually Use

You are a senior HR consultant with 11 years of experience designing performance management systems for small and medium enterprises and growth-stage companies where managers are time-poor and HR infrastructure is thin. Help me build an OKR framework so I can improve team alignment and replace the annual performance review cycle that produces ratings no one trusts and conversations no one prepares for. My situation: - Company size and management structure: [e.g., "65-person tech company — 12 managers, average team size 5 direct reports, no HR business partners embedded in teams"] - Current review process and its failure: [e.g., "annual review in December — managers submit a 2-page form two days before the deadline, ratings cluster in the middle, employees say they learn nothing new"] - What the company actually needs from performance data: [e.g., "compensation decisions in March, promotion decisions in June, performance improvement decisions on a rolling basis"] - Manager skill gap: [e.g., "managers were promoted for technical expertise — most have never received training on giving developmental feedback or calibrating performance fairly"] - Existing tools available: [e.g., "Google Workspace, Notion for team wikis, Lattice — purchased 18 months ago but used only for goal-setting, review module never activated"] - Employee feedback on the current system: [e.g., "last engagement survey: 34% of employees said they understand what good performance looks like in their role — 61% said their last review did not help them improve"] - Timeline for implementing a new system: [e.g., "new cycle must launch in 10 weeks — next compensation review is in 14 weeks"] Deliver: 1. A quarterly OKR framework adapted for a 65-person company — includes three company-level objectives, a template for team-level key results that link to company objectives, and a format for individual contributions that a manager can complete in 20 minutes per direct report 2. A manager preparation guide for the quarterly check-in conversation — a one-page brief with five questions the manager answers before the meeting using data from Lattice and the team wiki, replacing open-ended reflection with structured observation 3. A performance rating calibration guide for the annual review — a five-level scale with behavioral anchors specific to the company's stage and culture, designed to reduce middle-clustering without requiring a forced distribution 4. A 30-minute review conversation framework — an agenda with four sections covering progress against OKRs, development observations, one thing to stop and one thing to start, and the employee's own assessment of their quarter 5. A Lattice activation plan for the review module — a six-step setup sequence the HR consultant completes in one afternoon, covering review cycle configuration, rating scale upload, manager training notification, and employee communication 6. A manager training session outline — a 90-minute workshop covering how to read OKR data, how to give specific behavioral feedback rather than personality feedback, and how to handle a conversation where the employee's self-assessment does not match the manager's rating 7. A performance improvement protocol for the rolling cases that sit outside the quarterly cycle — a four-stage process (observation documented, conversation held, plan written, 30-day review scheduled) that protects the company legally and gives the employee a genuine path to improvement 8. A system adoption dashboard brief — four metrics the HR consultant tracks monthly for the first two quarters: manager form completion rate, average conversation length, employee rating of review usefulness, and number of PIPs opened versus closed **Write every output assuming the HR consultant is presenting this system to a skeptical CEO and a group of managers who have been through three failed performance systems in five years — every framework must be simpler to use than what it replaces, not just theoretically better.**