Tech-Driven Performance Reviews

Do your performance reviews still feel like guesswork? The latest SAP SuccessFactors release quietly introduced something game‑changing: AI that actually helps you write feedback and plan goals. Here’s what caught my eye. • The new Performance & Goals module now suggests comments based on the skill and rating you choose. No more staring at a blank box wondering how to phrase constructive criticism. • It generates “performance insights” that sift through an employee’s data and summarize strengths, achievements, and areas to improve. In other words, you walk into one‑to‑ones with a clear picture and a fairer perspective. • Sentiment analysis flags negative or mixed feedback in 360‑degree reviews, so you know where to focus your coaching. • Preparation time for compensation discussions drops by 90% because AI surfaces the right talking points, and overall performance goal‑setting is 80% faster. What this really means is that AI is moving from buzzword to practical tool. It’s taking the busywork out of reviews and letting managers spend more time on real conversations. And it’s doing it while employees still feel seen and fairly evaluated. I’m curious: would you trust AI to help shape feedback and compensation discussions? Have you tried any of these tools yet? Share your experiences — or tag a colleague who should weigh in. #SAPSuccessFactors #PerformanceManagement #AIinHR #PeopleAnalytics #FutureOfWork

Meta and JPMorgan Chase announced they’d be using AI for annual performance reviews. As if the process couldn’t get any more performative and detached from reality. Now we’re adding AI slop to employee feedback. The goal is to summarize a year’s worth of work with the help of AI, but the challenge is that most AI doesn’t have the domain expertise required to write a high-quality performance review. Just because it can generate something in the format of an employee review or summarize notes in an employee review style doesn’t mean it has the knowledge to do it well. AI lacks the context to evaluate the intent and define the outcome. That’s why AI alone is rarely enough to support enterprise or customer use cases. AI doesn’t have the evaluation criteria for what makes a performance review high-quality. It doesn’t have access to comprehensive long and short-term employee outcomes to know what about a review improves them. It lacks context about how employees respond to the feedback it generates. AI doesn’t understand the intent of the manager writing the review or their management style. There are two levels of information required to support this use case: General Information: Domain knowledge about what makes performance reviews effective. Personalized Information: Domain knowledge about the employee and manager that personalizes the review to fit the unique nuances of both people involved. AI alone isn’t enough to support the performance review use case, so it’s only one part of the agentic workflow. The decision and workflow chains must be mapped. Resources must be provided at each link in the chain. Intent and outcomes must be fully defined upfront. Agents require new design paradigms that go beyond AI, or the result is slop.

The AI Assessment Effect Candidates often tend to adjust their answers or behavior to match what they believe the “ideal candidate” profile looks like. A new study published earlier this month found that when candidates believe they’re being assessed by artificial intelligence, they emphasize analytical skills and downplay their intuitive and emotional skills. This so-called “AI assessment effect” stems from the widespread assumption that AI-based evaluations prioritize rational, data-driven attributes over human-centric abilities. Researchers warn that if job seekers tailor their behavior to what they think AI values, their true competencies and personalities may remain hidden, undermining the integrity of the recruitment process. In addition if most candidates assume AI favors analytical traits, the talent pipeline could become increasingly uniform, limiting diversity and reducing the variety of perspectives within organizations. The researchers recommend 1) Radical transparency: Don’t just disclose that AI is used in assessments—be explicit about what it evaluates. Clearly communicate that your AI values a range of traits, including creativity, emotional intelligence, and intuitive problem-solving. Share examples of successful candidates who excelled by showcasing these qualities. 2) Regular behavioral audits: Go beyond demographic bias checks. Look for patterns of behavioral adaptation: Are candidates’ responses becoming more homogeneous over time? Is there a noticeable shift toward analytical self-presentation at the expense of other valuable traits? 3) Hybrid assessment models: Combine AI and human judgment to ensure a more balanced and holistic evaluation of candidates. See research published in the June issue of the Proceedings of the National Academy of Arts and Sciences. https://lnkd.in/ebtD4HBd

I grew up watching machines go rogue🤖 Now I help companies stop that from happening in real life. 🦾 Growing up, I loved watching sci-fi movies. In the 90s, the theme was always the same: man creates a scientific marvel, man loses control over said marvel… cue the running, screaming, and inevitable bloodshed. As a kid, I lapped up those stories, which always hammered home one moral: humans messing with the laws of nature never ends well. Fast forward to today, and I find myself advising companies on a very real version of that narrative, which is using AI in HR. With AI tools increasingly used to monitor performance and even flag employees for dismissal, the question isn’t just “can we do this?” but “should we? And how do we do it fairly?”. I recently shared my views on this topic with HRD Asia (link to article in the comments below). In general, HR teams must get the following right: 🔹 Transparency: Employees should know how their performance is being assessed and what data is being used. 🔹 Human Oversight: AI should assist human judgment. It can never replace it. Accordingly, a meaningful review process is essential. 🔹 Vendor Accountability: Employers must understand how third-party tools work and ensure they don’t produce biased outcomes. 🔹 Appeal Mechanisms: Employees need a way to challenge decisions influenced by AI. 👨⚖️ In my practice, I’ve already seen clients ask whether an AI-generated score is enough to justify dismissal. My answer? Not without human validation and a clear explanation of how the score was derived. Implementing a Human-In-The-Loop approach to any automated scoring tools would also ensure that any employment decision is validated by an employee who can justify the AI-generated recommendation. This is especially important in employment decisions relating to summary dismissal which carry significant legal risks, such as wrongful dismissal claims. While there is no hard and fast rule when it comes to determining the appropriate level of intervention, the key principle is that the reviewer must be able to understand how the AI arrived at its decision and the individual must have the authority to override it if necessary. The review process should not be a mere formality or rubber-stamping exercise; it must serve as a meaningful check to ensure fairness and accountability. As the use of AI tools in HR is increasingly becoming popular, the time to get familiar with the legal issues surrounding its use is now. Build internal safeguards, update your policies, and make sure your HR team understands the tools they’re using. Because if those 90s sci-fi movies have taught us anything, it’s that leaving machines to make human decisions rarely ends well. Would love to hear how you are balancing AI efficiency with fairness, do share your thoughts below! #AIinHR #WorkplaceFairness #SingaporeHR #HRCompliance #AIethics #HumanOversight #EmploymentLaw #SciFiMeetsReality

How much does AI augment employee performance and creativity, and which employees benefit the most? Some compelling research suggests that AI may provide less uplift to lower skilled employees, instead mostly increasing the output of higher skilled employees. This has important implications on how organizations should be thinking about talent management, as well as some not hard to imagine societal consequences. A comprehensive field experiment published in the Academy of Management Journal involving over 3,000 customers and 40 sales agents demonstrates that AI assistance can dramatically boost human creativity and performance, but the improvements are not evenly distributed. When AI handled routine lead generation tasks, allowing human agents to focus on complex sales interactions, the results were remarkable. Agents with AI support were 2.33 times more successful at handling novel, untrained customer questions compared to those working alone. This translated directly to better business outcomes, with AI-assisted teams achieving significantly higher sales conversion rates. However, the benefits of AI assistance were heavily skill-biased. High-skilled employees showed 2.81 times greater improvement in creative problem-solving with AI assistance compared to their lower-skilled colleagues. Through detailed interviews, researchers found that while high-skilled employees experienced increased motivation and creativity when freed from routine tasks, lower-skilled employees often felt more pressure and stress when handling only complex interactions. These findings have major implications for how organizations should approach AI implementation. AI-driven performance augmentation may be less of a talent equalizer and more of a talent divider. If augmentation of labor happens primarily via delivering efficiency around more clerical elements, freeing up time for higher order/more creative functioning, the beneficiaries will be those with the comparative performance advantages on the higher order tasks. Study: https://lnkd.in/ek-ugetF Talk: https://lnkd.in/erpRm77S #ArtificialIntelligence #FutureOfWork #TalentManagement

I audited 50+ performance programs. Here’s what I found. After interviewing to people leaders at companies sized 50 to 5,000 employees in tech, healthcare, AI, consulting, construction, manufacturing, finance about their programs—the patterns are the same. Want to see how you stack up? Comment AUDIT and I’ll send you my link 1) Tools exist. Engagement does not. Templates, cycles, and docs live next to the work, not in it. Managers see “another form,” or an "extra thing" not a tool that makes them better managers. 🛠️ The fix: Add lightweight checkpoints in the flow of work; auto-prompt managers & employees on real milestones (1:1s, project/sprint end); use AI to surface likely evidence from notes/goals so feedback isn’t a blank page. 2) Foundations for fair promotion decisions are still lacking. Promotion gates aren’t tied to clear, leveled behaviors, so calibration becomes a lengthy and costly debate. On top of it, most employees can’t see the bar. 🛠️ The fix: Publish transparent levels (scope, autonomy, outcomes) and a leveled rubric; rate against competencies (with 2–3 evidence bullets), not just an overall label; performance feedback monthly; Stop 9-boxing. 3) Individual performance ≠ company results. Most companies have some version of goals, but most employee goals are often bottom-up and unverified (yet performance is still measured against these). 🛠️ The fix: Use a light cascade (company → function → team → individual) OR stop at team; combine goal attainment and competency rating as separate, weighted inputs to an overall score. 4) Managers don’t see value (and it’s an expensive process). Hours spent writing narratives for their reviews then sitting in calibration to justify gut feel. Most of this effort does not improve business outcomes. 🛠️ The fix: Pre-calibrate folks against clearly defined performance rubrics; Use "calibration" as-needed, not after every review; leverage AI to flag outliers and synthesize themes for managers to verify. 5) “Continuous” is the goal but still not operationalized. Most programs still run in bursts; the system doesn’t generate small, in-flow signals between cycles. 🛠️ The fix: Make feedback embedded, prompted, and auto-aggregated from the work you already do. Continuous = ongoing signals, not more meetings. TL;DR Less form, more signal. A level-based structure. Embedded prompts. Short, regular performance (feedback) loops. If you want a quick, no-fluff audit with a maturity score and top 3 priorities—comment AUDIT and I’ll send a calendar link.

Performance Management Is Changing. Are You Ready? In the next three years, the way we assess performance will be fundamentally redefined by one major force: AI integration in everyday work. As AI becomes a baseline expectation, not a bonus skill, we must evolve how we define, measure, and reward performance. Historically, performance ratings emphasized behaviors, results, and goals. But now? The “how” -- which includes leveraging technology to amplify impact -- will be just as important as the “what.” How will future performance evaluations shift? 1 - Integrated Skills Assessment: Evaluating both human expertise and how effectively employees use AI tools (e.g., ChatGPT, Copilot, Replit, Claude, etc.) to improve work quality, efficiency, and innovation. 2 - Job-Based AI Expectations: Different jobs require different levels of AI fluency. A marketer using AI to generate customer insights is different from a software engineer automating testing scripts. Leaders must tailor benchmarks. 3 - Rewarding Adaptability: The speed at which employees adapt to new tools and workflows will be a key differentiator in performance. 4 - Performance Calibration Will Evolve: Managers will need to assess not only results but how AI helped achieve it. Was it used ethically? Was an employee’s judgment applied appropriately? Future-Focused Performance Rating Scale: 1. Not Meeting Expectations = Struggles to complete job responsibilities, avoids using new tools, and resists tech-enabled workflows. Example: Continues using outdated manual processes despite available AI support; missing deadlines and quality standards. 2. Partially Meeting Expectations = Some responsibilities met, but inconsistent application of AI tools limits impact. Learning curve is still steep. Example: Tries using AI but produces work that needs frequent rework; hesitant to explore new tech features. 3. Meeting Expectations = Meets job goals, uses AI/tech tools appropriately to support tasks, and demonstrates foundational digital agility. Example: Uses AI to draft content or summarize reports; integrates output with sound judgment and team input. 4. Exceeds Expectations = Proactively uses AI and digital tools to improve quality and productivity; mentors others in effective use. Example: Automates data workflows, reduces turnaround time by 30%, and helps peers adopt similar approaches. 5. Consistently Exceeds Expectations = Expertly integrates AI into work to drive innovation, transformation, or measurable business impact. Example: Creates AI-driven customer engagement model that increases conversion rates; pilots new tools for cross-functional use. As tech becomes the partner for most jobs, we must redefine excellence. Are your performance frameworks ready for that shift? #PerformanceManagement #Compensation #HR #HumanResources #AI #FutureOfWork #TotalRewards #SHRM #WorldatWork #CompensationConsultant #Pay #PerformanceFeedback https://shorturl.at/915OT

We’ve seen this movie before. And it didn’t end well. Some organizations are beginning to incorporate AI usage into performance evaluations. On the surface, this makes sense: Encourage adoption. Increase productivity. Accelerate transformation. But there’s a deeper risk emerging—one that feels very familiar. When you tie performance to a metric, you get more of that metric. Not necessarily better outcomes. We learned this the hard way. Wells Fargo incentivized account openings… and got millions of them. Just not in the way they intended. Now we’re seeing early signs of a similar dynamic with AI: - Employees encouraged to maximize usage - “Tokenmaxxing” and activity becoming proxies for productivity - Output measured… but impact unclear This creates two systemic risks: 1️⃣ Behavioral distortion People optimize for the metric (AI usage), not the outcome (better performance). 2️⃣ Hidden cost escalation More AI usage = more tokens = more cost You may be rewarding behavior that actively increases your expense base without improving results. And yet many organizations are struggling to show meaningful ROI from AI investments. That shouldn’t be surprising. If we measure AI success by how much it’s used rather than how work is improved, we’re solving the wrong problem. The real opportunity isn’t individual productivity. It’s redesigning how work gets done: - Reengineering processes - Improving team-based workflows - Elevating decision quality - Reducing friction AI is a system-level tool—not a personal productivity scoreboard. A recent article reinforced this caution: using AI-related metrics in performance reviews requires extreme care, or organizations risk reinforcing the wrong behaviors. 📎 https://lnkd.in/gp7jDq4j Before adding AI usage into your performance management system, it’s worth asking: Are we measuring meaningful outcomes—or just creating a new version of “accounts opened”?

What is "good performance" in the age of AI? When a person and an AI build something together, there are three things worth measuring: (1) how the person did (2) how the AI did (3) how the two did as a pair Most companies only measure the first one, because that's how performance management has always worked. Except agents are starting to do manager things: sitting in on every call, handing out the action items, nudging you when you miss one, deciding what's worth escalating. Once software plays that role, there's an argument to be made that you should evaluate it the way you'd evaluate a human manager. A new Harvard Business Review piece proposes a three-layer performance framework for exactly this: (1) Human-contribution metrics: Shift away from outputs AI can inflate, toward what it can't replace → Boundary judgment: does the person catch it when the AI is out of its depth, and what do they do next? → Orchestration: can they use AI to lift the whole team's throughput, not just their own? → Learning velocity: are they adapting as the tools and workflows keep changing? (2) AI system and agent metrics: Uptime and accuracy aren't enough once agents are in the loop → Objective attainment: did the agent do what it was asked, inside the limits it was given? → Explainability: can you tell why it did what it did, and prove it? → Escalation quality: at the edge of its scope, does it hand off to a human or plow ahead? (3) Human-AI system metrics: The part almost nobody measures yet, and perhaps the most interesting → AI substitution rate: how much of the work have humans effectively dropped out of? → Complementarity index: how often did a person actually change the outcome vs rubber-stamp it? → Value attribution: how much of the final result traces to AI execution vs human judgment? I'm not saying I agree with every metric here, and nobody has the solution yet. But I agree with the notion that performance in the AI era needs a redesign, and doing so requires real-time visibility into the work people and agents actually produce. Ex: You can't compute a complementarity index from a form someone fills out from memory once a year. The signal for every one of these metrics lives in the work itself, continuously, across both the people and the agents producing it. That's part of the context layer we're building at Windmill. Full piece: https://lnkd.in/eqU_scDx If you're already folding AI into performance management, how are you doing it? What challenges are you facing?