Welcome back to our series on data biases. We’re breaking down the different ways data skews in traditional market research; specifically, how they can corrupt the data you base your biggest decisions on. We’ve discussed the ways consumers subconsciously want to control how they’re perceived with social desirability bias and the ways acquiescence bias can sneak unusually agreeable feedback into survey responses. Today we’re unpacking recency bias, where recent perceptions cloud consumer judgement.
Why is recency bias such a thing right now? With rapid trend cycles and social media generating a constant flow of new content, brands, ads, and campaigns — every swipe and scroll is saturated with newness. It's no wonder consumers have a hard time staying rooted in reality. So if you ask a survey respondent whether they prefer a cheaper version of a more expensive item, and they just saw a glossy Instagram ad for the cheaper version but they're vaguely aware that a higher-quality product exists that would be more cost-effective and provide a more satisfactory experience overall, they're more likely to say they prefer the cheap one.
Because of this, recency bias is more pervasive than ever, especially in surveys. But what is recency bias?
It's the bias that causes individuals to emphasize recent events, or novel data and experiences over historical, long-term evidence to the contrary. Essentially, recency bias makes people think that the recent past is a more reliable indicator of what's to come than patterns that have repeated over time.
Brands know all about the challenge of cutting through the noise, and recency bias plays a role in that, but with the right approach, you can mitigate its effects. We'll go through a few of the most common ways that recency bias can mess with your bottom line, and how in-the-wild testing gets around it.
The rise of "dupe culture" caused consumers to actively seek cheaper lookalikes of premium products.
On the surface, that looks like a serious shift in preferences, but recency bias may be causing a skew in that data.
When a dupe goes viral on TikTok or gets a wave of glowing reviews, it floods the consumer's recent memory. Survey them in that window and they'll report strong preference for the dupe. Survey them six months later, after the novelty has worn off, and the quality gap has made itself known – the story often changes. The problem is that most brands only catch the first snapshot and are left wondering why their approach doesn’t seem to be working.
Think about the explosion of dupes for high-end skincare and Stanley cups. Search interest and stated purchase intent spike hard around the viral moment, but longer-term repurchase rates and legacy brands need to meet the moment differently. Brands that remain clear and consistent in their mission are able to retain loyalty even with the existence of cheaper alternatives. Survey respondents captured mid-trend may not reflect what they actually buy down the line. Testing in the wild, which tracks actual behavior rather than stated preference at that moment, captures true behavior rather than individual skewed perceptions.
The protein boom
The protein-everything trend has made high-protein positioning feel like an easy win, but brands that build strategy around peak-trend survey data risk optimizing for a moment, not a market.
Right now, protein is everywhere: protein coffee, protein ice cream, protein pasta. In this trend environment, survey respondents asked about purchase drivers in food and beverage are primed to say protein matters to them because they've been marinating in protein-forward content and products. That doesn't mean they'll pay a premium for it in six months, or that it will meaningfully drive repeat purchases once the novelty fades. Plus, there still remains a question around whether consumers actually want protein where it didn’t exist before. They might say that they want coffee with protein cold foam, but that hasn’t been proven as a fact without behavioral data. This dynamic creates the perfect storm for flash-in-the-pan marketing, which can be costly and short-sighted.
The low-fat craze of the 1990s is a useful historical parallel. Consumer surveys at the time showed overwhelming preference for low-fat options; and it was a signal that drove massive reformulation across the industry. Within a decade, the trend had reversed, and many of those products quietly disappeared. Using this example, with real-world testing, you could observe what and where people actually want low-fat options in their carts, and where they aren’t as concerned or interested, which would reveal which parts of your product line produce stronger or weaker signals before pumping a bunch of dollars into something that doesn’t land.
Whether it's a viral moment or a booming macro trend, survey data captured around peak cultural saturation will almost always overstate the durability of that preference.
The brands that get burned aren't always the ones that noticed the trend and invested in it. But it is more commonly the ones that mistook a loud moment for a long-term shift that can make less informed go-to-market decisions that sting down the line.