Human memory is not an audit trail. Yet in most Indian organisations, the annual performance review still depends on a manager's ability to recall twelve months of work in a 45-minute conversation. Research consistently shows that the last 6–8 weeks dominate that recall — a phenomenon called the recency effect. For distributed, fast-growing teams operating across languages and geographies, this is not a minor inconvenience. It is a systematic injustice.
The consequences compound quietly. The employee who delivered a critical project in February is rated on the crisis they managed in November. The one who was visible in Q4 outscores the one who quietly drove results all year. Over time, this erodes trust — not because managers are malicious, but because they are human, and no human brain was designed to carry twelve months of evidence in perfect fidelity.
Why Mid-Market India Is Most Exposed
Enterprise companies in India — with dedicated HR business partners, structured calibration committees, and documented performance journals — have at least some structural buffers against pure memory-based reviews. Mid-market companies (50–2,000 employees) typically have none of these. A single HR manager supports 40–80 employees. Managers conduct reviews once a year, sometimes twice. The review itself is a conversation, not a process.
Add the complexity of distributed workforces — field employees in Andhra Pradesh reporting to managers in Bengaluru, remote workers across multiple time zones, multilingual teams where nuance gets lost in translation — and the memory problem becomes structural. The same cognitive limitations that affect any manager are amplified by distance, language barriers, and the absence of continuous documentation.
In our internal analysis of 250+ performance review conversations facilitated through TARA, fewer than 22% of managers referenced any specific employee achievement from the first half of the review year without being prompted.
The Three Biases That Fill the Memory Gap
When memory fails, something else fills the gap. In our analysis of performance conversations, three biases appear consistently when managers lack documented evidence.
- Recency bias: Weighting the last 6–8 weeks of performance disproportionately. An employee who struggled in November but excelled all year gets rated on November.
- Affinity bias: Unconsciously rating employees who communicate in similar styles, share cultural references, or are more physically visible more favourably.
- Halo/horn effect: One highly visible success or failure colours the entire year's rating. A missed deadline in Q3 becomes the frame through which all of Q1 and Q2 is re-evaluated.
None of these are intentional. All of them are structural — predictable outcomes of a process that asks humans to do something brains simply cannot do reliably: recall and fairly weight twelve months of nuanced performance data from memory.
What the Evidence-Led Alternative Looks Like
The fix is not to train managers to have better memories. It is to build a process that does not depend on memory in the first place. Evidence-led performance management means: continuous OKR tracking so achievement is recorded in real time; structured review conversations that surface specific examples rather than relying on recall; and AI-assisted bias detection that flags when a rating pattern diverges from documented evidence.
At TalentSpotify, this is exactly what TARA does. It joins the review conversation, listens for bias signals in real time, and presents managers with the documented evidence trail before ratings are finalised. The goal is not to replace managerial judgment — it is to give managers better inputs so their judgment is actually grounded in what happened, not what they happen to remember.
The recency effect, affinity bias, and halo effect are not character flaws. They are cognitive architecture. The right response is structural — build processes that don't require perfect memory. Evidence-led reviews are not a luxury; for growing Indian companies, they are the only path to a review system employees will trust.