Most behavioral-economics findings in this hub did not survive scrutiny. Tversky and Kahneman’s 1973 availability heuristic did. Across five decades of replication, in domains as varied as risk perception, jury decisions, consumer behavior, and media-driven judgment, the core finding has held. Here is the honest case for one of behavioral science’s most durable insights.
If you have been reading through this hub, you have watched a long parade of canonical behavioral-science findings get dismantled. Power posing did not survive Carney’s own recantation. Ego depletion collapsed under Hagger 2016. Money priming evaporated in preregistered replications. The marshmallow test shrank to something much smaller than its original claim once SES was controlled. Stereotype threat, the entire family of social-priming results, the facial-feedback hypothesis --- one after another, the most-cited demonstrations of “this is how the mind works” have either failed to replicate or eroded to a confidence interval that quietly straddles zero.
A rational reader by now might be tempted to conclude that all of behavioral economics is suspect. That conclusion would be wrong, and the availability heuristic is one of the cleanest counter-examples to it.
Because in the same five-decade window that has produced all of those replication failures, the availability heuristic --- the proposition that people judge the frequency or probability of an event by the ease with which instances come to mind --- has held up. It has held up across the original Tversky and Kahneman experiments, across the Lichtenstein lethal-events work that extended it to real-world risk perception, across Combs and Slovic’s media-coverage analyses, across Schwarz’s dual-process refinements in the 1990s, and across hundreds of subsequent studies that have probed when ease-of-retrieval functions as informational input to a frequency judgment.
This is the second anti-example article in a hub full of takedowns. The first was the default effect. The availability heuristic is the second. It exists for the same three reasons. First, calibration --- readers should leave this hub knowing that “behavioral economics is mostly broken” is wrong; the more accurate claim is that the field has produced a small number of robust, mechanism-grounded findings and a much larger number of fragile, contextually thin findings, and the failure mode has been treating those two categories as if they were the same. Second, decision-usefulness --- for an executive evaluating which behavioral concepts to actually deploy in marketing, sales, hiring, investment, or organizational decision-making, the availability heuristic is one of the highest-confidence inputs you have. And third, intellectual honesty --- a hub that catalogs the failures owes readers the parts that worked.
So here is the case for the availability heuristic, with the legitimate critiques included.
What Tversky and Kahneman 1973 Actually Demonstrated
The foundational paper is Tversky, A., & Kahneman, D. (1973). “Availability: A Heuristic for Judging Frequency and Probability.” Cognitive Psychology, 5(2), 207—232. DOI: 10.1016/0010-0285(73)90033-9.
The structure of the paper is unusual for behavioral economics in that it is not one experiment. It is a series of eight separate studies, each probing a different facet of the proposed heuristic, each using a different population and a different task, each yielding a result consistent with the same underlying mechanism. The cumulative weight of the eight studies is much harder to dismiss than any single demonstration would be, and this is part of why the paper has aged so well --- you would have to come up with eight different artifact explanations to dismiss the eight different studies, and no critic has ever managed to do that.
The setup is straightforward. Tversky and Kahneman proposed that when people are asked to estimate the frequency of a class of events or the probability of an outcome, they often do not enumerate base rates or consult a frequency table they have constructed. They instead reach for whatever examples come to mind, and they use the ease of that retrieval as a cue to frequency. Things that are easier to recall are judged more frequent. This is the availability heuristic.
The eight studies tested specific implications of this proposal. In one, subjects were given a list of names --- some famous, some obscure --- and asked to judge whether the list contained more men or more women. When the famous names were predominantly of one gender, subjects systematically judged that gender to be more frequent in the list, even when the actual count was the opposite. The famous names were more available; availability translated to higher judged frequency.
In another study, subjects were asked which is more frequent in English text: words beginning with the letter R, or words with R as the third letter. Most subjects said words beginning with R, even though the actual frequency runs the other way. Words beginning with R are easier to retrieve --- you can generate them by running through the alphabet of initial letters --- whereas words with R in the third position require a much harder retrieval strategy. Ease of retrieval drove the frequency judgment, in this case to the wrong answer.
In another, subjects were given a brief story and asked to estimate the number of plausible paths or completions for a particular event. Subjects’ estimates tracked the ease with which they could actually generate paths --- if generation was easy, estimated probability was high; if generation was hard, estimated probability was low. The mechanism was operating not just on retrieval of stored memories but on the ease of constructive generation.
The eight studies, taken together, established something more durable than any single experimental demonstration. They established a regularity of cognition --- ease of retrieval functions as informational input to frequency and probability judgments --- and they showed this regularity across enough different task structures that it could not be dismissed as a quirk of one paradigm.
The follow-up paper, Tversky, A., & Kahneman, D. (1974). “Judgment Under Uncertainty: Heuristics and Biases.” Science, 185(4157), 1124—1131. DOI: 10.1126/science.185.4157.1124, embedded the availability heuristic in a broader framework of three heuristics --- representativeness, availability, and anchoring-and-adjustment --- and is the paper that gave behavioral economics its dominant analytical vocabulary for the next forty years. Both papers are still actively cited; neither has been retracted; neither has been the subject of the methodological critique that has dismantled other founding behavioral-economics work.
The Lichtenstein 1978 Lethal-Events Study
The original Tversky and Kahneman experiments lived in a relatively narrow paradigm --- subjects in a lab, judging frequencies of words or names. The question of whether the availability heuristic also drove real-world risk perception, with all the consequences that would imply, was settled by Lichtenstein, S., Slovic, P., Fischhoff, B., Layman, M., & Combs, B. (1978). “Judged Frequency of Lethal Events.” Journal of Experimental Psychology: Human Learning and Memory, 4(6), 551—578. DOI: 10.1037/0278-7393.4.6.551.
The Lichtenstein team asked U.S. subjects to estimate the annual frequency of death from various causes --- forty-one causes in total, ranging from heart disease and stroke (very common) to botulism and shark attack (very rare). They then compared subjects’ estimates to actual U.S. mortality statistics.
The structure of the bias was systematic and large. Subjects substantially overestimated the frequency of dramatic, vivid, easily-imagined causes --- tornado, flood, homicide, accident, fire, cancer --- and substantially underestimated the frequency of common, undramatic causes --- stroke, diabetes, asthma, emphysema, tuberculosis. The pattern was not noise; it was a regular distortion in the direction predicted by availability theory. Causes that received heavy media coverage, that produced vivid mental imagery, that were easy to retrieve as discrete dramatic events, were systematically over-weighted. Causes that killed quietly in hospital beds were systematically under-weighted.
The Lichtenstein paper went further than the Tversky-Kahneman 1973 work in three ways. First, it grounded the availability heuristic in a domain --- mortality-risk perception --- that has obvious consequences for public-health policy, regulatory design, and individual decision-making. Second, it documented the size of the distortion in real frequency space: subjects’ estimates of the most-overestimated causes were ten to a hundred times higher than the actual rates, and their estimates of the most-underestimated causes were ten to a hundred times lower. The bias was not a small perturbation; it was a major reshaping of the perceived risk landscape. Third, it explicitly tied the pattern of distortions to the pattern of media coverage of those causes of death, setting up the natural follow-up study.
That follow-up is the next paper.
Combs and Slovic 1979 --- Newspaper Coverage and Judged Frequency
The hypothesis raised by Lichtenstein’s results was that the systematic distortions in risk perception were caused, at least in part, by systematic distortions in media coverage. Causes of death that were heavily covered by newspapers were retrieved more easily and judged more frequent. Causes that received little coverage faded from availability and were judged less frequent than they actually were.
Combs, B., & Slovic, P. (1979). “Newspaper Coverage of Causes of Death.” Journalism Quarterly, 56(4), 837—849 is the paper that tested this. The team content-coded a year of newspaper coverage from two U.S. newspapers --- the New Bedford Standard-Times and the Eugene Register-Guard --- across forty-one causes of death, counting the number of articles and column inches devoted to each.
The pattern was clear. Some causes of death received massive coverage relative to their actual mortality contribution. Homicide received roughly three times the coverage that disease deaths received, in column-inch terms, despite causing dramatically fewer deaths than common diseases. Tornado, fire, flood, accident, and other dramatic-event causes all received coverage substantially in excess of their mortality contribution. Diseases of all kinds, despite collectively dominating actual mortality, received coverage substantially below their share.
Combs and Slovic then correlated the coverage pattern with the judged-frequency pattern from the Lichtenstein 1978 dataset. The correlation was substantial and in the predicted direction: causes that were over-covered in newspapers were over-estimated in frequency, and causes that were under-covered were under-estimated. The relationship was not perfect --- there are other determinants of which deaths feel vivid, including direct personal exposure and dramatic-event imagery independent of news coverage --- but the media-coverage signal was a meaningful predictor of the perception-distortion signal.
This is one of those papers that, in retrospect, looks obviously right and even underwhelming --- of course people who read more newspaper stories about homicide think homicide is more common --- but at the time of publication, the empirical demonstration of a clean media-coverage-to-perception link, with both ends quantified and the correlation explicitly computed, was a significant advance. It also did something important for the credibility of the broader availability framework: it offered an external, observable mechanism (newspaper-coverage counts) that could be measured independently of the perceptual outcome it was supposed to explain. The whole loop --- newspaper coverage produces ease of retrieval, ease of retrieval produces high judged frequency, high judged frequency is measurable in surveys --- was now fully instrumented.
The Combs-Slovic finding has been replicated many times since. Subsequent work has documented analogous patterns for terrorism risk, crime risk, child-abduction risk, plane-crash risk, and several other heavily-mediated risk categories. The basic empirical regularity --- that the news-coverage profile of a risk meaningfully shapes the public perception of that risk’s frequency --- is one of the more durable findings in applied behavioral science.
Schwarz 1991 --- Ease of Retrieval As Information; The Dual-Process Refinement
The 1980s availability literature ran into a productive theoretical problem. The original Tversky-Kahneman formulation said that ease of retrieval was the operative cue --- but how was ease of retrieval being measured by the cognitive system itself, and could that measurement be dissociated from the content of what was retrieved? The paper that solved this and meaningfully refined the theory is Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer-Schatka, H., & Simons, A. (1991). “Ease of Retrieval as Information: Another Look at the Availability Heuristic.” Journal of Personality and Social Psychology, 61(2), 195—202.
The Schwarz team designed an experimental paradigm to dissociate the ease of retrieval from the number of instances retrieved. Subjects were asked to recall examples of times they had behaved assertively. Some subjects were asked to recall six examples; other subjects were asked to recall twelve. The team had pre-tested the task to confirm that recalling six was relatively easy for most subjects, and recalling twelve was relatively difficult. After completing the recall task, subjects rated how assertive they were in general.
The classical availability prediction would be that subjects who recalled more examples of their own assertive behavior should rate themselves as more assertive --- they had more evidence of assertiveness available to them. The Schwarz result was the opposite. Subjects who had been asked to recall twelve examples rated themselves as less assertive than subjects who had been asked to recall six. The reason: subjects who had to recall twelve examples struggled to do so, and the experienced difficulty of retrieval itself became the informational cue. “It was hard to recall many examples of my assertive behavior; therefore I must not be very assertive.” The ease of retrieval was being used as informational input, independent of the content of what was retrieved.
This was a significant theoretical refinement. It established that the availability heuristic operates through two channels --- the content of retrieved instances, and the meta-cognitive experience of retrieval difficulty itself --- and that these two channels can produce opposite predictions in the right experimental setup. The dual-process refinement was incorporated into the broader heuristics-and-biases program and has held up across many replications.
It also, importantly, is one of the older social-psychology findings to have survived the replication crisis intact. The Schwarz 1991 paradigm has been replicated repeatedly, including in preregistered studies in the 2010s. A 2017 meta-analysis of ease-of-retrieval studies (Weingarten and Hutchinson 2018 in the Psychological Bulletin) reported a robust pooled effect for the basic dual-process distinction, with appropriate caveats about moderators. The finding has not collapsed.
What Replicates And What Doesn’t
Now for the honest part, because every anti-example article in this hub owes one: not every claim ever made under the availability-heuristic banner has held up. The core finding is robust. Some specific applications have proven more conditional.
What has clearly replicated:
The core proposition that ease of retrieval functions as informational input to frequency and probability judgments. This is the central claim of Tversky and Kahneman 1973, and it has held up across hundreds of studies in dozens of paradigms.
The Lichtenstein-style finding that risk-frequency perception is systematically distorted by the vividness, recency, and media coverage of the underlying events. This is one of the more reliable applied findings in behavioral economics, and it has been documented across many specific risk domains (mortality, crime, terrorism, financial losses).
The Schwarz dual-process refinement, that retrieval difficulty operates as a meta-cognitive cue independent of the content retrieved. This has held up in preregistered replications.
What has been more conditional:
Specific quantitative claims about the size of availability-driven biases in individual decision domains. The general direction of the bias replicates; the magnitude varies substantially with population, task framing, and the specific risk being judged. A growth-team practitioner who reads “availability biases produce huge effects” should not assume that they will personally produce huge effects in any specific application without measurement.
The “affect heuristic” extension by Slovic and colleagues, in which positive or negative affect attached to a risk is supposed to drive both probability and benefit judgments in a coordinated way. This has shown more mixed replication results, and several specific claims in the affect-heuristic literature have weakened under more rigorous testing.
The “availability cascade” framework popularized by Kuran and Sunstein, which extends availability theory to collective social dynamics around moral panics and risk amplification. This is theoretically interesting and observationally suggestive, but it has not been tested with the experimental rigor of the core availability work. Treat it as a useful frame for interpreting social phenomena, not as a confirmed quantitative regularity.
Some downstream claims in the consumer-behavior literature about availability-driven purchase decisions, particularly around the alleged size of recency effects in brand recall, have been pared back by more recent measurement. The direction still goes the way availability theory predicts; the magnitude is often smaller than the marketing-textbook version implies.
The honest summary, then, is that the core availability heuristic is one of behavioral economics’ most durable findings, but that specific quantitative claims and specific extensions deserve case-by-case evaluation rather than blanket endorsement. This is what a mature behavioral-science finding looks like: a robust core, a set of well-established applications, and a periphery of weaker claims that benefit from continued empirical work.
How Availability Shows Up In Business
For practitioners --- marketers, salespeople, executives, investors, hiring managers --- the availability heuristic shows up in operational decisions constantly, often without being recognized as such.
Testimonials and case studies. The reason a single vivid customer testimonial often outperforms a statistical summary of customer outcomes is, in significant part, availability. The reader who encounters one specific named customer with a specific outcome retrieves that customer as an instance when later evaluating whether the product works. The aggregate statistic, even if more representative, is harder to retrieve in the moment of decision. This is one of the better-evidenced reasons why testimonials and case studies remain a high-leverage marketing asset --- they are not just emotionally persuasive; they meaningfully alter the availability landscape the prospect uses to estimate probability of success.
News-cycle effects on investment behavior. Investor risk perception is heavily shaped by the most recent vivid market events. After a financial crisis, perceived equity-investment risk runs above the long-term frequency of crises would justify, and this elevated risk perception persists for years before normalizing. After a long bull market with few salient losses, perceived equity-investment risk runs below the long-term frequency would justify, and this depressed risk perception contributes to over-allocation. This is availability operating on the time domain rather than the media domain.
Hiring decisions. Hiring managers consistently over-weight recent vivid examples --- the candidate who reminds them of a successful past hire, the candidate who triggers memory of a recent disaster --- relative to base-rate predictive validity of the formal evaluation signals. The structured-interview literature in industrial psychology is, in significant part, an effort to constrain availability-driven bias in hiring decisions. (Worth noting: structured interviewing also has its own implementation problems, but the diagnosis of the availability problem in unstructured interviewing is well-established.)
Crisis response and incident communication. When something has gone wrong publicly --- a security breach, a regulatory action, a high-profile customer issue --- the perceived frequency of that class of issue spikes in the audience’s mind in a way that is dramatically out of proportion to the actual base rate. This is the basic mechanism behind why “one bad review online is worth ten good ones” --- not because people consciously weight them this way, but because the vivid negative example is highly available when the prospect is later judging brand reliability.
Risk decisions in regulated industries. Compliance functions, audit functions, and risk-management functions inside organizations are systematically prone to availability bias from recent enforcement actions. A company in an industry where a peer just received a large fine for one type of violation will over-allocate compliance attention to that specific type of violation for the next year or two, often at the expense of base-rate-larger risk categories that are not currently in the news.
Sales pipeline forecasting. Sales leaders consistently over-weight recent vivid deal outcomes --- the big close, the big loss --- when estimating future close rates, relative to the statistical base rate of similar deals. This is a direct application of availability bias to forecasting and is part of why pipeline forecasts are systematically biased toward the recent emotional signal rather than the base rate.
News-cycle effects on the founder psyche. Founders read TechCrunch and overestimate how often companies in their category get acquired for nine figures, because acquisition stories are over-reported relative to the base rate of acquisitions per company-year. The same founders underestimate how often companies in their category just quietly run out of money over four years, because that outcome is under-reported. This availability distortion shapes ambition-calibration, fundraising-strategy, and exit-expectation in systematic ways across the venture-backed startup population.
The general pattern in all of these is the same: humans, including sophisticated decision-makers operating in professional contexts with significant economic consequences, lean on ease-of-retrieval as a cue to frequency and probability, and they do so systematically rather than randomly. This is a productive lens for diagnosing why specific decision processes go wrong and where to intervene.
What This Means For Strategists
The practical takeaways for someone making real decisions about marketing, sales, hiring, investment, or operational design are:
Availability is exploitable in marketing, and the exploitation is well-evidenced. Vivid customer stories, specific named case studies, concrete success-outcome imagery, and any other input that increases the ease with which a prospect can retrieve a positive use-case for your product will move purchase probability. The marketing-textbook claim here is, for once, supported by the underlying behavioral science. The pricing-anchoring article in this hub treats anchoring more skeptically; the testimonial-and-case-study leverage in availability deserves the opposite treatment --- it is real, it is large, and the marketing-and-content investment to produce vivid examples that retrieve easily under purchase-consideration conditions is well-justified.
Availability is defendable against in your own decisions, but only if you build the defense into the process. The way to reduce availability bias in your own decision-making is not to try harder to ignore vivid examples in the moment --- that does not work, because the bias operates pre-consciously on the retrieval-ease cue rather than at the deliberative-weighting stage. The way to reduce it is to insert a base-rate-consultation step into the decision process before any vivid examples enter. For hiring, this is structured interviewing with pre-specified evaluation criteria. For investment, this is checklists and pre-mortems that force consultation of historical base rates before considering the specific recent example. For pipeline forecasting, this is conversion-rate base rates by deal stage and segment, surfaced in the forecasting tool itself, rather than the rep’s gut estimate. The structural intervention works; the willpower-based intervention does not.
Pay particular attention to availability in your own crisis-response posture. When something visible has just gone wrong --- yours or a competitor’s --- your stakeholders will over-weight the frequency of that class of problem for some period after, regardless of the actual base rate. The implication is not to ignore the signal but to communicate the relative frequency explicitly. (“This kind of issue has occurred X times in the last five years across our customer base, which represents Y percent of operations.”) The base-rate framing partially repairs the availability distortion in the audience’s mind. The alternative, which is to lean into the apologetic framing without ever quantifying, leaves the distortion in place.
Be skeptical of your own news-cycle-driven strategic priorities. If your strategic priority list this quarter is heavily reshaped from last quarter’s list by a single recent vivid event --- a competitor’s announcement, a regulatory action, a tech-press story --- the chances are good that you are responding to an availability-distorted signal rather than to a meaningful base-rate shift. The disciplined version of strategic-priority setting is to consult your actual leading indicators (renewal rate, win rate by segment, top-of-funnel composition, NPS, account-expansion rate) rather than your news feed when deciding what to allocate attention to.
Use availability deliberately and ethically. The same mechanism that makes vivid testimonials effective makes manipulative testimonials effective. The line between “I am giving prospects a vivid example of a real outcome” and “I am cherry-picking the most vivid outliers and presenting them as typical” is the line between ethical marketing and consumer deception, and it has real legal consequences in addition to the reputational ones. The version of availability marketing that survives ethics review and consumer-protection scrutiny is the version where the vivid example is statistically representative or where the typicality is explicitly disclosed. The version that fails both is the version where the vivid example is being used to mislead about base rates.
The general framing: availability is real, large, and well-evidenced. It is genuinely useful as a marketing input and as a diagnostic lens for decision-making. It is also one of the bias categories that you should expect to be operating in your own decisions, and the defense against it is structural rather than willpower-based.
What This Anti-Example Tells Us About Behavioral Science Overall
The replication crisis is real. The catalog of canonical-then-collapsed findings in this hub is long. A reasonable executive could read all of that and conclude that behavioral economics is mostly an academic vanity project --- interesting theory, weak evidence, low practical reliability.
That conclusion would be wrong, and the availability heuristic, alongside the default effect, is the best counter-example.
What behavioral science can produce, when it does the work properly, is a small number of robust, mechanism-grounded, replicable findings that have meaningfully advanced our understanding of how decisions actually get made. Availability is one of them. Defaults are another. Loss aversion, in its non-inflated form, is a third. There are perhaps a dozen findings in the broader behavioral-economics literature that meet this standard --- well-replicated, well-mechanism-specified, well-measured at scale outside the convenience-sample lab paradigm.
The job of a sophisticated reader of this literature is to distinguish the dozen-or-so robust findings from the much larger catalog of fragile findings. The hub you are reading is a guided tour of that distinction. The availability heuristic is one of the answers you take away. The default-effect article is another. The collapsed findings catalog (power posing, ego depletion, money priming, the marshmallow test in its inflated form, facial feedback, social priming, and the rest) is the third part of the picture.
The synthesis: behavioral economics is more useful than the replication-crisis critique alone would suggest, and less useful than the original consulting-deck version implied. The mature, calibrated, post-replication-crisis behavioral economics is in better shape than either the original version or the pessimistic post-crisis version. Availability is one of the findings that demonstrates this is possible.
Sources
- Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5(2), 207—232. DOI: 10.1016/0010-0285(73)90033-9
- Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124—1131. DOI: 10.1126/science.185.4157.1124
- Lichtenstein, S., Slovic, P., Fischhoff, B., Layman, M., & Combs, B. (1978). Judged frequency of lethal events. Journal of Experimental Psychology: Human Learning and Memory, 4(6), 551—578. DOI: 10.1037/0278-7393.4.6.551
- Combs, B., & Slovic, P. (1979). Newspaper coverage of causes of death. Journalism Quarterly, 56(4), 837—849.
- Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer-Schatka, H., & Simons, A. (1991). Ease of retrieval as information: Another look at the availability heuristic. Journal of Personality and Social Psychology, 61(2), 195—202. DOI: 10.1037/0022-3514.61.2.195
- Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux. (Chapter 12-13 give the popular-press synthesis of the availability literature, including the dual-process refinement.)
- Weingarten, E., & Hutchinson, J. W. (2018). Does ease mediate the ease-of-retrieval effect? A meta-analysis. Psychological Bulletin, 144(3), 227—283. DOI: 10.1037/bul0000122
Related
Browse the full Replication Crisis Hub for other behavioral-science findings, including:
- The Default Effect --- the other anti-example; what a robust nudge actually looks like
- Halo Effect --- a related judgment-formation bias with its own evidence profile
- Sunk Cost Fallacy --- robust in some applications, conditional in others
- Pricing Anchoring --- the other Tversky-Kahneman heuristic, with a different replication trajectory
- Loss Aversion --- robust-ish but smaller than the consulting-deck version
FAQ
How do I reduce availability bias in my own decision-making?
Not through willpower. The bias operates pre-consciously on the retrieval-ease cue rather than at the deliberative-weighting stage, so “I will try harder to ignore vivid examples” does not work. The intervention that works is structural: insert a base-rate-consultation step into the decision process before any vivid examples enter. For hiring, this is structured interviewing with pre-specified criteria. For investment, this is checklists and pre-mortems that force consultation of historical base rates. For pipeline forecasting, this is conversion-rate base rates surfaced in the forecasting tool itself before the rep’s gut estimate. The structural defense is durable; the willpower defense is not.
What about base rate neglect more broadly?
Base-rate neglect and the availability heuristic are closely related --- both involve a failure to consult the underlying frequency distribution when judging probability. The classic Kahneman and Tversky representativeness work documents base-rate neglect as a separate phenomenon, and it has also held up reasonably well in replication, though with more nuance than the original framing suggested. Practically, the same intervention (force base-rate consultation as a structural input to the decision process) addresses both biases at once.
What about testimonials in marketing --- are they manipulative?
The mechanism is the same as the availability heuristic in any other context, but the ethics depend on whether the testimonial accurately represents the typical outcome. A vivid customer success story that reflects the actual outcome distribution of your customer base is informational --- it gives prospects a clearer mental model of what success with your product looks like. A vivid customer success story that is the statistical outlier, presented as typical, is misleading. The FTC’s endorsement guidelines explicitly address this distinction. The marketing-ethics version is “use vivid examples to make typical outcomes more retrievable, not to make atypical outcomes seem typical.”
Is the availability heuristic universally exploited in news media?
In the descriptive sense, yes. News editors have long understood that vivid, dramatic, narrative-shaped events get more attention than statistical aggregates, and the editorial selection process therefore systematically over-represents the dramatic end of the risk distribution. This is partly a market-driven incentive (dramatic stories get more attention, which gets more revenue) and partly a structural feature of journalism (an event that happened is a story; an event that did not happen is not a story). The result, predicted directly by Combs-Slovic, is that media-coverage profiles of risks diverge systematically from actual frequency profiles, in the direction that drives availability bias in audiences. This is not a conspiracy; it is an emergent property of news as a market.
How does the availability heuristic interact with social media specifically?
Social-media algorithms amplify the same selection process that traditional news editors carry out, often with higher intensity --- vivid emotionally-charged content has higher engagement, gets surfaced more often, and produces a more distorted availability landscape than traditional news did. The empirical work on social-media-driven risk-perception distortion is still developing, but the basic mechanism (algorithmic amplification of high-engagement vivid content, which then shapes user availability landscapes, which then drives audience risk perception) is well-established. The practical implication for individuals is that anyone relying heavily on social media for their picture of the world should expect their availability-distorted picture of base rates to diverge substantially from reality.
How do I distinguish robust behavioral-economics findings from fragile ones?
The four-question diagnostic from the default-effect article applies here. Is the mechanism over-determined (multiple plausible mechanisms predict the effect, so attacking any single mechanism does not dismantle the prediction)? Are the conditions of application well-specified (so failed replications can be diagnosed rather than dismissed)? Is the effect size large enough to detect without statistical heroics? Has it been measured in high-stakes field experiments with administrative-data outcomes, not just convenience-sample lab work? The availability heuristic and the default effect both pass all four. Most of the failed findings in this hub fail at least three. The diagnostic is not perfect, but it is a useful first-pass filter.
Is the availability heuristic actionable in B2B and enterprise sales contexts?
Yes, with the same caveats. The mechanism does not change between B2C and B2B contexts --- decision-makers in enterprise contexts are also human, and they also retrieve vivid examples preferentially when judging probability. The practical implication is that enterprise-sales materials that include vivid named-customer case studies tend to outperform purely statistical claims about customer outcomes, and that enterprise procurement decisions are likely to be partly driven by which competitor’s recent successes are most retrievable to the buying committee. The same ethical and accuracy considerations apply --- the vivid example should be representative of actual customer outcomes, not a cherry-picked outlier.
What is the single best deployment of the availability heuristic in a startup marketing program?
Customer success stories, with specific named customers, specific named outcomes, and specific named use-cases. These do double duty: they provide social proof (a real customer chose you), and they alter the availability landscape that prospects use to judge “will this work for my company?” The most-leveraged version is a small library of three to five specific case studies across the modal customer segments, each with enough specificity to actually retrieve as a mental image. This is one of the highest-confidence behavioral-marketing investments you can make, and it has the additional advantage of being a marketing asset that compounds over time rather than degrading.
replication-crisis availability-heuristic tversky-kahneman decision-making evidence-evaluation
Free Tool
Built for Experimentation Teams
GrowthLayer is the experimentation platform I built for CRO teams --- test management, AI-powered insights, and pattern recognition across your entire program.
Explore GrowthLayer → (opens in new tab)
Share this article
LinkedIn (opens in new tab) X / Twitter (opens in new tab)
Copy link
Go deeper
Methodology The PRISM Method Case Studies $30M+ in Results Work Together Services & Mentoring
Experimentation and growth leader. Builds AI-powered tools, runs conversion programs, and writes about economics, behavioral science, and shipping faster.