Recency Bias

Recency bias sits at the intersection of memory and judgment. It is the systematic tendency to give disproportionate weight to information encountered most recently, whether that information is a quarterly result, a single client complaint or last week's market swing, at the expense of a longer and more representative record. The bias is not a failure of intelligence or effort. It is a byproduct of how memory retrieval works: recent events are easier to recall, arrive with less cognitive effort and feel more relevant simply because they are fresh. For consultants, executives and boards, the practical consequence is that forecasts, performance ratings, hiring calls and capital allocation decisions routinely drift toward whatever happened last, rather than toward what the fuller evidence supports.

The psychological roots of the bias

Recency bias traces back to well-established findings in cognitive psychology about how sequences of information are stored and retrieved. The serial position effect, first studied systematically through list-recall experiments, shows that people remember items at the end of a list better than items in the middle, a pattern researchers labeled the recency effect1. That memory advantage is adaptive in narrow contexts, such as recalling a short list moments after hearing it, but it becomes a liability when it silently governs judgments that should draw on months or years of evidence. Daniel Kahneman described the underlying mechanism in his work on the availability heuristic, the shortcut where people judge the likelihood or importance of something based on how easily examples come to mind.

For example, students of policy have noted that the availability heuristic helps explain why some issues are highly salient in the public's mind while others are neglected. People tend to assess the relative importance of issues by the ease with which they are retrieved from memory and this is largely determined by the extent of coverage in the media

That same mechanism operates inside a boardroom or a one-on-one review. A manager does not consciously decide to ignore ten months of an employee's work; the last six weeks are simply what comes to mind first, fastest and with the least effort. Academic work on the availability heuristic and serial position effects, published in the Bulletin of the Psychonomic Society, confirms that recency and availability interact to produce systematic distortions in judgment tasks well beyond simple memory recall2. The distortion is predictable, measurable and, importantly, correctable once a decision process is redesigned to counteract it.

The business cost of overweighting recent events

Forecasting and strategic planning

Corporate forecasting is one of the most exposed processes to recency bias because it typically starts with the most recent actuals and extrapolates forward. A team that just closed a strong quarter tends to build next year's plan on the assumption that the trend continues; a team coming off a weak quarter tends to build in unwarranted caution. McKinsey's research on the "outside view" in strategic planning shows that forecasts anchored purely to internal, recent trends miss the reference-class information that a statistical comparison to similar past situations would surface. As Dan Lovallo, a McKinsey adviser on decision science, put it in describing the alternative approach:

The outside view is statistical. That's the basic difference between the inside view and the outside view

McKinsey partner Tim Koller has observed that most corporate planning teams barely have time to consolidate numbers from business units, let alone build a systematic outside view into the process3. That gap is precisely where recency bias fills in:

absent a disciplined reference-class check, the most recent quarter becomes the de facto forecasting model, whether or not it deserves that weight

Performance reviews and talent decisions

Annual and semiannual performance reviews are a textbook setting for recency bias because managers rarely keep continuous notes and instead reconstruct a year of performance from memory at review time. The employee who delivered steadily for eleven months but stumbled in the final weeks before the review often receives a lower rating than the record justifies, while an employee who coasted early and rallied late benefits from the same distortion in reverse. This pattern affects promotion decisions, bonus allocations and, cumulatively, retention, because employees who sense that timing rather than sustained contribution drives their evaluation lose confidence in the process. Organizations that require managers to log performance observations throughout the review period, rather than reconstructing them from memory at the end, materially reduce this skew.

Recency bias in investment and financial decisions

Financial markets provide some of the clearest, most costly examples of recency bias in action. Investors who watch an asset class outperform for one to three years tend to increase their allocation to it just as the outperformance is most likely to mean revert, then retreat from asset classes that recently lagged, locking in losses at close to the worst possible time. Omar Aguilar, chief executive and chief investment officer of Schwab Asset Management, has warned that this pattern directly undermines long-term outcomes.

Short-term market moves caused by recency bias can sap long-term results, making it more difficult for clients to reach their financial goals

Certified financial planner Charlie Fitzgerald III frames the bias as something closer to instinct than error, noting that "people need to understand that recency bias is normal and it's hard-wired", calling it "a survival instinct."4 Financial services stocks rose 32% in 2019, drawing in investors chasing the trend, only for those late entrants to underperform through 2020 as the momentum reversed. The lesson for consultants advising clients on capital allocation is not that recent performance is irrelevant, but that it needs to be weighted against a longer base rate before it drives a reallocation decision.

Why consultants and project leaders are not immune

Consulting engagements and large capital projects carry their own version of this distortion. Oxford researcher Bent Flyvbjerg's analysis of behavioral biases in project management identifies recency as one of the recurring contributors to cost overruns and schedule slippage on major infrastructure and technology programs5. Planners frequently anchor a revised estimate to the most recent milestone update rather than to the project's full track record, which understates the cumulative drift that has already occurred. A consulting team advising a client on a multiyear transformation can fall into the same trap: a strong recent workshop or a positive stakeholder meeting can overwrite months of slower progress in the team's own assessment of how the engagement is going. The remedy is the same discipline applied at the individual level:

treat the most recent data point as one observation among many, not as the summary of the whole trend

Structural defenses against recency bias

Reducing recency bias is a process design question, not a matter of asking individuals to try harder. Several practices have demonstrated measurable effects across forecasting, hiring and portfolio management, including the structured pre-mortem exercise that Gary Klein popularized for surfacing risks a recency-driven plan would otherwise miss6.

1. Build forecasts from a documented reference class of comparable past situations before layering in recent internal data
2. Require continuous, dated performance notes throughout a review cycle instead of end-of-period recall
3. Set fixed rebalancing rules for portfolios and budgets that do not respond to any single recent data point
4. Run a pre-mortem before major decisions, asking what a longer time horizon would suggest that the most recent trend does not
5. Separate the person who generates a forecast from the person who reviews it against historical base rates

None of these measures eliminates the underlying psychological pull; the recency effect is a stable feature of memory, not a bug that training removes. What structured processes do is interrupt the moment where that pull would otherwise translate directly into a decision, forcing a comparison against a fuller record before judgment is finalized. Organizations that build these checkpoints into forecasting cycles, review calendars and investment committees see steadier decisions, because the process no longer depends on any one person's unaided recall of the past year.

• 1recency effect
• 2primacy, recency and the availability heuristic
• 3how to take the outside view
• 4recency bias: what it is and why it causes poor investment choices
• 5top ten behavioral biases in project management
• 6performing a project premortem