Pricing psychology: what the research shows, and where Irish law draws the line

A price is a signal, not just a number

A price tells the customer more than what something costs. It says what kind of product this is, whether it is a good deal, and how much to trust you. Get the signal wrong and you leave money on the table, or put buyers off. Get it right and people pay a fair price and feel good about it.

That is not manipulation. It is understanding how people actually read numbers, which decades of behavioural economics have studied. Daniel Kahneman shared the 2002 Nobel memorial prize in economics for this line of work.¹ But the popular version of pricing psychology is full of effects quoted as laws from single studies, some of which have not survived being repeated.

So this article does three things for each effect: what the original study found, how well it has held up, and a worked example you can test on your own shop. Then it sets out where Irish law draws the line, because several "pricing tricks" are now plainly illegal. Every number in the worked examples is made up so I can show each step.

Anchoring: the first number pulls the rest

Anchoring is the tendency to lean on the first number you see when you estimate something. In Amos Tversky and Daniel Kahneman's classic 1974 demonstration, people first saw a number from a wheel of fortune, then estimated the percentage of African countries in the United Nations. Those who had seen 10 gave a median estimate of 25. Those who had seen 65 gave 45.² The number was obviously random, and it still moved the answer.

How well it holds up. Anchoring in estimates is one of the sturdiest findings in psychology. The Many Labs project ran 13 classic effects, including an anchoring task, across 36 samples and 6,344 people. Ten replicated consistently, and the two that failed were priming effects, not anchoring.³ The stronger claim, that an arbitrary anchor sets what people are willing to pay, is shakier: a set of new experiments by Maniadis, Tufano and List found much weaker anchoring in valuations and warned against building on a single striking result.⁴ So anchors move estimates reliably. How much they move a purchase is something to test, not assume.

In pricing, the anchor is usually the first price on the page: the most expensive plan shown first, the top hamper at the top of the range.

Worked example: the Grand Hamper at the top of the page

Northgate Hampers is an invented online hamper shop in Cork, at northgatehampers.example. Priya, who runs it, wants to know whether listing the €250 Grand Hamper first raises spending. She splits visitors to the Christmas range at random for four weeks.

• Group A sees the range sorted from the most expensive. 10,000 visitors, 400 orders, average order €72.
• Group B sees it sorted from the least expensive. 10,000 visitors, 410 orders, average order €66.

Revenue per visitor is orders times average order, divided by visitors. For A it is $400 \times 72 / 10{,}000 = €2.88$. For B it is $410 \times 66 / 10{,}000 = €2.71$. The anchor version earns 17 cent more per visitor, about 6%, while converting slightly less. That is the number to judge it on, and the test section below shows how to check that such a gap is not noise.

Where it backfires. Products with well known market prices, where an inflated anchor just looks odd. Repeat buyers who already know your prices. And any context where trust matters more than the sale.

The decoy effect: an option nobody buys

The decoy effect, or asymmetric dominance, is what happens when you add a third option that is clearly worse than one of the others. Joel Huber, John Payne and Christopher Puto showed in 1982 that adding such a dominated option can raise the share choosing the option it is worse than, which breaks a basic assumption of rational choice.⁵

The famous illustration is Dan Ariely's. An advert for The Economist offered an online subscription for $59, print for$125, and print plus online for $125. He gave the list to 100 MIT students and most chose the combined deal. With the print only option removed, another 100 students chose differently: the combined deal went from most to least popular.⁶ It is a classroom demonstration, not a controlled field study.

How well it holds up. This one is contested. Shane Frederick, Leonard Lee and Ernest Baskin ran a series of studies in which people could experience the options directly (tasting drinks, for example) and found no attraction effect at all. Quite often the decoy reduced the share of the option it resembled, which they called a repulsion effect.⁷ Huber, Payne and Puto replied that the effect remains robust when the original conditions are repeated, and that it is strongest when options are described on paper with clear numbers.⁸ The fair reading: a decoy can work on a pricing page with clear attributes, and can backfire elsewhere. Test it.

Worked example: a decoy plan at Tramore Bookings

Tramore Bookings is an invented booking software company in Waterford for salons, at tramorebookings.example. It sells Solo at €29 a month and Team at €79. The team tests adding Team Lite at €75, with fewer features than Team, on half of new trial accounts.

• Without the decoy, 500 sign ups: 340 Solo, 160 Team. Team's share is $160 / 500 = 32\%$.
• With the decoy, 500 sign ups: 220 Solo, 15 Team Lite, 265 Team. Team's share is $265 / 500 = 53\%$.

Monthly revenue without the decoy is $340 \times 29 + 160 \times 79 = 9{,}860 + 12{,}640 = €22{,}500$. With it, $220 \times 29 + 15 \times 75 + 265 \times 79 = 6{,}380 + 1{,}125 + 20{,}935 = €28{,}440$. A chi squared test on the Solo and Team counts gives a statistic of about 51, far beyond chance. In this invented case the decoy worked. In your case it might repel, which is exactly why you run the test.

Charm prices: €19.99 and the left digit

Prices ending in 99 or 95 work through what researchers call the left digit effect. Manoj Thomas and Vicki Morwitz showed in five experiments that a nine ending price is judged smaller than one a cent higher only when the left digit changes ($2.99 against$3.00), and more so when the prices being compared are close together.⁹ In the field, Eric Anderson and Duncan Simester manipulated price endings in three experiments with a retailer: the $9 ending raised demand in all three, more for new items than for familiar ones, and less when a "Sale" cue was also shown.¹⁰

The idea that round prices suit emotional purchases and precise prices suit considered ones comes from a 2015 paper by Monica Wadhwa and Kuangjie Zhang. A 2018 replication by Harms and colleagues did not find clear evidence for it, so treat it as an open question.¹¹

Worked example: €49.99 against €50 on the Afternoon Tea Hamper

Priya tests €49.99 against €50.00 with 10,000 visitors in each group. At €49.99, 420 people order, a conversion rate of 4.2%. At €50.00, 380 order, 3.8%.

The difference looks like a win. Is it real? The two proportion test asks whether a gap this size is likely if the prices made no difference. The pooled rate is $800 / 20{,}000 = 4.0\%$. The standard error of the difference is $\sqrt{0.04 \times 0.96 \times (1/10{,}000 + 1/10{,}000)} = 0.00277$. So $z = 0.004 / 0.00277 = 1.44$, a two sided p value of about 0.15. Not significant. The test is explained in Foundations: hypothesis tests.

How many visitors would she need to detect 3.8% against 4.2% reliably? The sample size formula for two proportions is

$n = \frac{(z_{\alpha/2} + z_{\beta})^2 \left[p_1(1-p_1) + p_2(1-p_2)\right]}{(p_1 - p_2)^2}$

where $n$ is the visitors needed in each group, $z_{\alpha/2} = 1.96$ sets a 5% significance level, $z_{\beta} = 0.84$ sets 80% power, and $p_1$ and $p_2$ are the two conversion rates. It is set out step by step in Foundations: sample size. With $p_1 = 0.042$ and $p_2 = 0.038$ the answer is $7.84 \times 0.0768 / 0.000016 \approx 37{,}600$ visitors per group, nearly four times what she ran.

Even if the lift were real, compare revenue per visitor, not conversion: $0.042 \times 49.99 = €2.10$ against $0.038 \times 50 = €1.90$. Charm prices cost a cent an order. That is rarely the issue. Running a test too small to tell is.

Where it backfires. Premium positioning and services, where .99 can look like haggling. And B2B, where a quote of €4,999.99 just looks odd.

Price as a sign of quality

When people cannot judge quality directly, they use price as a clue. Sometimes that is sensible, because cost and quality often go together. The most quoted evidence is a 2008 brain imaging study by Hilke Plassmann and colleagues: 20 people tasted wines in a scanner, told one was $5 and another$45, or the same wine at $10 and at$90. They rated the "expensive" wine as more pleasant, and activity in the medial orbitofrontal cortex, an area linked to experienced pleasantness, rose with it.¹²

How well it holds up. It is one study of 20 people, with wine, which is about as subjective a product as exists. It shows that a price claim can change how something is experienced, at least in that setting. It does not show how big the effect is for your products.

An example: a Galway skincare brand that is new and unknown will not look premium just by charging premium prices. Without reviews, ingredients and proof, a high price reads as presumption. The heuristic matters most where quality is hard to check (wine, supplements, professional services) and least where specifications and reviews are everywhere.

Good, better, best: the compromise effect

Given three options, people lean towards the middle one, because it avoids both extremes. Itamar Simonson described this compromise effect in 1989, alongside the attraction effect, as choices people can easily justify.¹³ It is why so many software companies show three tiers. My rule of thumb: three tiers, four at most. Beyond that, the comparison gets harder rather than easier.

Worked example: which tier pays at Tramore Bookings

Tramore Bookings' three plans, with the example's own customer counts and monthly churn (the share of customers who cancel each month):

Tier	Price	Customers	Monthly churn
Solo	€29	1,200	8%
Team	€79	2,800	4%
Studio	€199	450	2%

Team has $2{,}800 / 4{,}450 = 62.9\%$ of customers and $221{,}200 / 345{,}550 = 64.0\%$ of monthly revenue. The middle tier carries the business, as the compromise effect predicts.

But the value of a customer depends on how long they stay. A simple lifetime value, with no discounting, is

$\text{LTV} = \frac{m}{c}$

where $m$ is the monthly margin per customer (not the revenue) and $c$ is the monthly churn rate, so $1/c$ is the expected number of months a customer stays. The full version with retention and discounting is in Foundations: lifetime value. Assume an 80% margin. Solo is worth $0.8 \times 29 / 0.08 = €290$, Team $0.8 \times 79 / 0.04 = €1{,}580$ and Studio $0.8 \times 199 / 0.02 = €7{,}960$. Studio has only 450 customers, but each is worth five times a Team customer. The top tier is not just a decoy for the middle one. Price it as a real product.

Where it backfires. Tiers that differ in name only, and a middle tier that is so obviously the only sensible choice that customers feel steered.

Reference prices, and the Irish rules on "was" prices

A struck through higher price next to the selling price, "was €60, now €45", gives the customer a reference point, and people judge prices against reference points rather than in isolation. That is also exactly why the law watches it closely.

In Ireland the rules on unfair and misleading commercial practices sit in Part 3 of the Consumer Protection Act 2007, as amended by the Consumer Rights Act 2022. A commercial practice is misleading if it gives false information, or would be likely to deceive or mislead the average consumer, about matters that include the price of a product or the existence of a specific price advantage, and would be likely to cause them to take a decision they would not otherwise have taken. Where it does that through information, it is no defence that the information is factually correct.¹⁴ For a "was" price the Act is specific: the question is whether the product was previously offered openly and in good faith at that price, at the same place, for a reasonable period of time before the claim was made.¹⁴ A "was" price that was hardly ever charged is the classic case. The Competition and Consumer Protection Commission (CCPC) enforces the Act, and a misleading commercial practice is a criminal offence: on indictment, a fine of up to €60,000 for a first conviction and up to €100,000 for a later one.¹⁵

EU law adds a hard number. Under the European Union (Requirements to Indicate Product Prices) (Amendment) Regulations 2022, Ireland's version of the EU price reduction rule, a trader who announces a price reduction must show the prior price in that announcement: the lowest price the trader applied to the product in a period of not less than 30 days before the reduction, or before the first of a run of consecutive reductions. Leaving it out is an offence.¹⁶ The CCPC adds that any discount, an amount or a percentage off, must be worked out from that prior price, and that the rule covers goods, not services.¹⁷

How rarely is too rarely? Britain has just had a test case. On 30 July 2026, in the British Competition and Markets Authority's case against the mattress seller Emma, decided by the High Court of England and Wales under the UK law that applied at the time, the court found that Emma had committed some infringements, but refused to impose the regulator's proposed rule that at least one sale at the reference price is needed for every two at the discount. The judge found real situations in which the average consumer would not feel misled even if few people paid the reference price.¹⁸ An English judgment does not bind an Irish court, and it does not make reference prices safe. In Ireland the test remains whether the average consumer, in context, would be misled, and on top of that the announced "was" price has to respect the 30 day prior price rule.

Worked example: a fair "was" price at Northgate Hampers

Priya wants to run "was €60, now €45" on the Cheese Board Hamper in January. The saving is $(60 - 45) / 60 = 25\%$. Before she shows it, she checks three things the law cares about. Is €60 the prior price, the lowest price she charged for the hamper in the 30 days before the reduction, so that the 25% is worked out from the right number? Was the hamper genuinely on offer at €60, openly and in good faith, for a reasonable period, and did people actually buy at it? The CCPC's view is that the reduced price should then run for no longer than the higher price did.¹⁷ And does the sale end when she says it ends? If the answers are yes, the reference price is information. If not, it is a misleading practice, and a reduction announced without the prior price is an offence in itself, however common the practice is.

Pain of paying: why the payment method matters

Paying feels like a loss, and some ways of paying feel more like one than others. Drazen Prelec, who developed the idea with George Loewenstein, describes the research question as how payments diminish the enjoyment of what is bought.¹⁹ In 2001 Prelec and Duncan Simester found, in studies with genuine transactions of potentially high value, that people told to pay by credit card were willing to pay more than people told to pay in cash, by up to 100%.²⁰ It is a small set of studies without a direct replication, so treat the size as an upper bound. A brain imaging study by Brian Knutson and colleagues found that excessive prices activated the insula and deactivated the mesial prefrontal cortex before people decided not to buy.²¹ The insula is also active when people anticipate physical pain, which is where the "paying hurts" headline comes from. The study shows a link with negative arousal, not that paying is literally painful.

The practical lesson is about friction. Every extra step between "I want this" and "it's done" is a chance to reconsider.

Worked example: revenue per checkout start, by payment method

An invented month at Northgate Hampers, 1,000 checkout starts per method:

Method	Completed	Average order	Revenue per checkout start
Bank transfer	620	€145	€89.90
Card, typed in	780	€138	€107.64
Saved card	890	€142	€126.38
Apple Pay	920	€151	€138.92

Revenue per checkout start is completions times average order, divided by starts: for Apple Pay, $920 \times 151 / 1{,}000 = €138.92$. The ranking is the point: fewer steps, more completed orders. These are invented figures, and in a real shop the methods attract different customers, so a gap like this shows a correlation, not proof that switching methods would cause it.

One caution on automatic renewal, the payment that needs no action at all. It is convenient, and it is regulated. Before a consumer is bound by a distance contract that contains a subscription, the Consumer Rights Act 2022 requires the trader to give the total costs per billing period and, for a contract that extends automatically, the conditions for terminating it.²² A term that renews a fixed term contract automatically, with a deadline to opt out that is unreasonably early, is on the Act's list of terms presumed unfair.²³ Build reminders and easy cancellation in from the start. A subscription people forget they have is a complaint waiting to happen.

Bundles, and why mental accounting matters

People keep mental accounts and dislike paying separately for things they think of as one purchase. Richard Thaler set this out in 1985 as mental accounting.²⁴ A bundle sells the set at one price. Unbundling does the opposite and advertises the lowest possible entry price, with add ons on top.

Worked example: bundling Tramore Bookings' add ons

The base app is €29 a month. Advanced reports are €20 and priority support €15. Sold separately, 35% of customers take reports and 22% take support. Revenue per customer is $29 + 0.35 \times 20 + 0.22 \times 15 = 29 + 7 + 3.30 = €39.30$.

As a test, offer a bundle of all three for €49. Say 58% choose the bundle and the rest stay on the base app. Revenue per customer is $0.58 \times 49 + 0.42 \times 29 = 28.42 + 12.18 = €40.60$, 3.3% more. But more customers now have priority support, which costs money to provide, so check the margin before the revenue.

Partitioned prices, delivery charges and drip pricing

Should you show €30 plus €5 delivery, or €35 delivered? Vicki Morwitz, Eric Greenleaf and Eric Johnson found in 1998 that splitting a price into parts can lower the total customers remember and raise demand, depending on how the surcharge is shown. A later review of the research found the effect weakens, and can reverse, when the surcharge looks unfair or excessive.²⁵

In Ireland this is no longer only a question of psychology. An invitation to purchase must give the price inclusive of taxes and any freight, delivery or postal charges, or, where those cannot reasonably be calculated in advance, the fact that they will be payable.²⁶ The CCPC's pricing guidance for businesses says "There should be no hidden charges",²⁷ and the advertising code of the Advertising Standards Authority for Ireland (ASAI) says quoted prices should normally include VAT and other "inescapable costs to the consumer".²⁸ So "€30 plus €5 delivery", shown clearly from the start, is lawful. A mandatory "booking fee" or "service charge" added at checkout is not: it belongs in the headline price.

Worked example: €30 plus delivery against €35 delivered

Priya tests both on 5,000 visitors each. "€30 plus €5 delivery" converts at 4.8% (240 orders), "€35 delivered" at 4.4% (220 orders). The pooled rate is 4.6%, the standard error of the difference $\sqrt{0.046 \times 0.954 \times (2/5{,}000)} = 0.0042$, and $z = 0.004 / 0.0042 = 0.95$, a p value of about 0.34. No detectable difference, and the revenue per order is identical. In this case the choice comes down to what fits the brand, and "delivered" is simpler to read.

Prospect theory: losses weigh more than gains

Prospect theory, from Daniel Kahneman and Amos Tversky's 1979 paper, describes how people weigh outcomes relative to a reference point, with losses weighing more than gains of the same size.²⁹ Their 1992 version gives the value of a gain or loss $x$ as³⁰

$v(x) = \begin{cases} x^{\alpha} & \text{if } x \geq 0 \\ -\lambda(-x)^{\beta} & \text{if } x < 0 \end{cases}$

where $v(x)$ is how much the outcome is felt to be worth, $x$ is the gain (positive) or loss (negative) in money, $\alpha$ and $\beta$ (both estimated at 0.88) make each extra euro matter a little less, and $\lambda$ is loss aversion, the factor by which losses outweigh gains, estimated at 2.25 in that paper. Expected value, the rational benchmark this departs from, is in Foundations: expected value.

How well it holds up. Loss aversion is real, but 2.25 is not a constant of nature. A 2024 meta analysis of 607 estimates from 150 articles put the mean at about 1.96, with a 95% interval of 1.82 to 2.10.³¹ Losses weigh roughly twice as much, not exactly 2.25 times.

Worked example: a €50 gain and a €50 loss

With the 1992 estimates, a €50 gain is felt as $50^{0.88} \approx 31.3$. A €50 loss is felt as $-2.25 \times 50^{0.88} \approx -70.4$. The loss weighs 2.25 times the gain, by construction.

What this does not show is that "don't miss out on €50" sells better than "save €50". Loss aversion is about how outcomes are weighed, not about how a message is worded. The research on gain and loss framed messages, mostly from health communication, finds small and inconsistent differences: a meta analysis of 53 studies on disease detection found loss framed messages only slightly more persuasive, a correlation of 0.04.³² So I no longer recommend turning every benefit into a fear ("don't be left behind while competitors pull ahead"). Frame the offer clearly. Where something genuinely ends, say so honestly. And never invent the deadline, because that is now illegal.

Fake urgency is a banned practice

Section 55 of the Consumer Protection Act 2007 lists commercial practices that are prohibited outright. One of them is making a false representation that a product, or particular terms, will only be available for a limited time, to get an immediate decision and deprive people of the time to make an informed choice.³³ The British regulator's guidance on its version of the same ban gives the obvious example: a countdown clock that says the offer ends, after which the offer carries on and the clock restarts.³⁴ If the offer really ends when the clock says, that is fine. If a near identical offer then appears shortly afterwards, it may still mislead.³⁴

The same applies to "only two left" when there are 200, and to a "sale" that never ends. In the British Emma case, a countdown timer suggesting that prices were available only for a period that was always about to expire was among the concerns the parties had resolved before the court hearing.¹⁸

How to test any of this on your own shop

Every effect above is a hypothesis about your customers until you test it. Three rules make a pricing test worth having.

1. Decide the sample size first, from the effect you care about, and run until you reach it. Peeking daily and stopping on the first good day is the most common way to fool yourself.
2. Judge on revenue or margin per visitor, not conversion rate. A lower price almost always converts better.
3. Watch the long run. A tactic that lifts this month's orders and costs trust is a bad trade. Lifetime value matters more than any single order.

Here is the core of the analysis, in Python with no extra libraries:

import math

def two_proportion_test(orders_a, visitors_a, orders_b, visitors_b):
    """z statistic and two sided p value for a difference in conversion rates."""
    p_a, p_b = orders_a / visitors_a, orders_b / visitors_b
    pooled = (orders_a + orders_b) / (visitors_a + visitors_b)
    se = math.sqrt(pooled * (1 - pooled) * (1 / visitors_a + 1 / visitors_b))
    z = (p_a - p_b) / se
    p_value = math.erfc(abs(z) / math.sqrt(2))
    return z, p_value

def visitors_needed(p1, p2, z_alpha=1.96, z_beta=0.84):
    """Visitors per group to detect p1 against p2 at 5% significance and 80% power."""
    return math.ceil((z_alpha + z_beta) ** 2 * (p1 * (1 - p1) + p2 * (1 - p2)) / (p1 - p2) ** 2)

# The charm price test from above
z, p = two_proportion_test(420, 10_000, 380, 10_000)
print(f"z = {z:.2f}, p = {p:.2f}")                    # z = 1.44, p = 0.15
print(visitors_needed(0.042, 0.038))                  # 37628 per group

# Revenue per visitor, the number that decides
print(420 * 49.99 / 10_000, 380 * 50.00 / 10_000)     # 2.10 against 1.90

For revenue per visitor itself, test it directly: compare the per visitor revenue of the two groups (most visitors at €0, buyers at their order value) with a t test or a bootstrap, because a conversion test alone says nothing about basket size. The complete guide to testing methods covers the options.

Present your value honestly

Every pricing choice sends a signal about the product, the brand and how you see your customers. The effects in this article are real tools, but most of them are smaller and less certain than their reputations, and a few have become legal risks.

• Test everything. What worked in a study, or for another company, may not work for your customers.
• Measure revenue and margin, not just conversion.
• Think long term. Tactics that win this month and cost trust are a loss.
• Stay on the right side of the line. Genuine reference prices, total prices up front, deadlines that are true. There is a line between presenting your value well and pressuring people into purchases they regret, and Irish and EU consumer law now draws much of it for you.

I have been applying behavioural economics to pricing for over a decade, across eCommerce, SaaS and B2B. The most useful thing it has taught me is the least glamorous: run the test properly, and believe the result.

Footnotes

1. The Nobel Prize, The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel 2002, summary, https://www.nobelprize.org/prizes/economic-sciences/2002/summary/

2. Amos Tversky and Daniel Kahneman, Judgment under uncertainty: heuristics and biases, Science 185(4157), 1124 to 1131, 1974, https://sites.socsci.uci.edu/~bskyrms/bio/readings/tversky_k_heuristics_biases.pdf

3. Richard A. Klein and colleagues, Investigating variation in replicability: a "Many Labs" replication project, Social Psychology 45(3), 2014, abstract at https://research.tilburguniversity.edu/en/publications/investigating-variation-in-replicability-a-many-labs-replication-/ and the list of effects, including anchoring, in the data paper at https://openpsychologydata.metajnl.com/articles/10.5334/jopd.ad

4. Zacharias Maniadis, Fabio Tufano and John A. List, One swallow doesn't make a summer: new evidence on anchoring effects, American Economic Review 104(1), 277 to 290, 2014, https://www.aeaweb.org/articles?id=10.1257/aer.104.1.277

5. Joel Huber, John W. Payne and Christopher Puto, Adding asymmetrically dominated alternatives: violations of regularity and the similarity hypothesis, Journal of Consumer Research 9(1), 90 to 98, 1982, https://econpapers.repec.org/article/oupjconrs/v_3a9_3ay_3a1982_3ai_3a1_3ap_3a90-98.htm

6. Dan Ariely, Are we in control of our own decisions?, TED talk transcript, December 2008, https://www.ted.com/talks/dan_ariely_are_we_in_control_of_our_own_decisions/transcript

7. Shane Frederick, Leonard Lee and Ernest Baskin, The limits of attraction, Journal of Marketing Research 51(4), 487 to 507, 2014, https://marketing.wharton.upenn.edu/wp-content/uploads/2016/10/The-Limits-of-Attraction.pdf

8. Joel Huber, John W. Payne and Christopher P. Puto, Let's be honest about the attraction effect, Journal of Marketing Research 51(4), 520 to 525, 2014, https://people.duke.edu/~jch8/bio/Papers/HuberPaynePutoJMR%202014.pdf

9. Manoj Thomas and Vicki Morwitz, Penny wise and pound foolish: the left digit effect in price cognition, Journal of Consumer Research 32(1), 54 to 64, 2005, https://econpapers.repec.org/article/oupjconrs/v_3a32_3ay_3a2005_3ai_3a1_3ap_3a54-64.htm

10. Eric T. Anderson and Duncan I. Simester, Effects of $9 price endings on retail sales: evidence from field experiments, Quantitative Marketing and Economics 1(1), 93 to 110, 2003, https://econpapers.repec.org/article/kapqmktec/v_3a1_3ay_3a2003_3ai_3a1_3ap_3a93-110.htm

11. FORRT Replication Atlas, This number just feels right (Wadhwa and Zhang, Journal of Consumer Research, 2015), recorded replication by Harms and colleagues, Royal Society Open Science, 2018, https://forrt.org/flora-replication-atlas/doi/10.1086/678484/

12. Hilke Plassmann, John O'Doherty, Baba Shiv and Antonio Rangel, Marketing actions can modulate neural representations of experienced pleasantness, PNAS 105(3), 1050 to 1054, 2008, https://rnl.caltech.edu/publications/pdf/plassmann2008.pdf

13. Itamar Simonson, Choice based on reasons: the case of attraction and compromise effects, Journal of Consumer Research 16(2), 158 to 174, 1989, https://econpapers.repec.org/article/oupjconrs/v_3a16_3ay_3a1989_3ai_3a2_3ap_3a158-74.htm

14. Law Reform Commission, Consumer Protection Act 2007, revised, section 43 (Misleading: false, misleading or deceptive information), subsections (1) to (4) and (6)(a), updated to 24 November 2025, Revised Acts, https://revisedacts.lawreform.ie/eli/2007/act/19/revised/en/html

15. Law Reform Commission, Consumer Protection Act 2007, revised, sections 47 (Offence: misleading commercial practices) and 79 (Fines and penalties), updated to 24 November 2025, Revised Acts, https://revisedacts.lawreform.ie/eli/2007/act/19/revised/en/html

16. Government of Ireland, S.I. No. 597 of 2022, European Union (Requirements to Indicate Product Prices) (Amendment) Regulations 2022, regulation 4 (new Regulation 5A, prior price), Irish Statute Book, 2022, https://www.irishstatutebook.ie/eli/2022/si/597/made/en/html

17. Competition and Consumer Protection Commission, Price reductions, CCPC, accessed November 2026, https://www.ccpc.ie/information-for-businesses/selling-goods-and-services/pricing-and-price-reductions/price-reductions

18. Competition and Markets Authority v Emma Matratzen GmbH and others, [2026] EWHC 1995 (Ch), High Court of England and Wales, The National Archives (United Kingdom), https://caselaw.nationalarchives.gov.uk/ewhc/ch/2026/1995

19. Drazen Prelec, The pains and pleasures of pricing, lecture slides for 15.821 Listening to the Customer, MIT OpenCourseWare, 2002, https://ocw.mit.edu/courses/15-821-listening-to-the-customer-fall-2002/5cc52570803b6e1bbea64585acdce779_redblack20copy.pdf

20. Drazen Prelec and Duncan Simester, Always leave home without it: a further investigation of the credit card effect on willingness to pay, Marketing Letters 12, 5 to 12, 2001, https://link.springer.com/article/10.1023/A:1008196717017

21. Brian Knutson, Scott Rick, G. Elliott Wimmer, Drazen Prelec and George Loewenstein, Neural predictors of purchases, Neuron 53(1), 147 to 156, 2007, https://web.stanford.edu/~knutson/bad/bk07n.pdf

22. Government of Ireland, Consumer Rights Act 2022, Schedule 3 (Information to be provided: off premises contract or distance contract), clauses (h) and (t), Irish Statute Book, 2022, https://www.irishstatutebook.ie/eli/2022/act/37/schedule/3/enacted/en/html

23. Government of Ireland, Consumer Rights Act 2022, Schedule 5, Part 1 (Contract terms presumed unfair), term 9, Irish Statute Book, 2022, https://www.irishstatutebook.ie/eli/2022/act/37/schedule/5/enacted/en/html

24. Richard Thaler, Mental accounting and consumer choice, Marketing Science 4(3), 199 to 214, 1985, https://econpapers.repec.org/article/inmormksc/v_3a4_3ay_3a1985_3ai_3a3_3ap_3a199-214.htm

25. Eric A. Greenleaf, Eric J. Johnson, Vicki G. Morwitz and Edith Shalev, The price does not include additional taxes, fees, and surcharges: a review of research on partitioned pricing (including Morwitz, Greenleaf and Johnson, Divide and prosper, Journal of Marketing Research, 1998), Journal of Consumer Psychology, 2016, preprint at https://business.columbia.edu/sites/default/files-efs/pubfiles/26031/Morwitz%20Greenleaf%20Shalev%20Johnson%20JCP%2015%2000002%20R1%20April%2019%202015%20.pdf

26. Law Reform Commission, Consumer Protection Act 2007, revised, section 46 (Misleading: withholding, omitting or concealing material information), subsection (3)(c) and (d), updated to 24 November 2025, Revised Acts, https://revisedacts.lawreform.ie/eli/2007/act/19/revised/en/html

27. Competition and Consumer Protection Commission, Product pricing obligations for Irish businesses, CCPC, accessed November 2026, https://www.ccpc.ie/information-for-businesses/selling-goods-and-services/pricing-and-price-reductions

28. Advertising Standards Authority for Ireland, ASA Code, section 4, Misleading advertising, rule 4.23, adstandards.ie, accessed November 2026, https://adstandards.ie/code/misleading-advertising/

29. Daniel Kahneman and Amos Tversky, Prospect theory: an analysis of decision under risk, Econometrica 47(2), 263 to 291, 1979, https://econpapers.repec.org/article/ecmemetrp/v_3a47_3ay_3a1979_3ai_3a2_3ap_3a263-91.htm

30. Amos Tversky and Daniel Kahneman, Advances in prospect theory: cumulative representation of uncertainty, Journal of Risk and Uncertainty 5(4), 297 to 323, 1992, https://econpapers.repec.org/article/kapjrisku/v_3a5_3ay_3a1992_3ai_3a4_3ap_3a297-323.htm

31. Alexander L. Brown, Taisuke Imai, Ferdinand M. Vieider and Colin F. Camerer, Meta analysis of empirical estimates of loss aversion, Journal of Economic Literature 62(2), 485 to 516, 2024, https://www.aeaweb.org/articles?id=10.1257/jel.20221698

32. Daniel J. O'Keefe and Jakob D. Jensen, The relative persuasiveness of gain framed and loss framed messages for encouraging disease detection behaviours: a meta analytic review, Journal of Communication 59(2), 296 to 316, 2009, https://www.cbsm.com/articles/33455-the-relative-persuasiveness-of-gain-framed-and-loss-framed-messages-for-encouraging-disease-detectio

33. Law Reform Commission, Consumer Protection Act 2007, revised, section 55 (Prohibited commercial practices), subsection (1)(n), updated to 24 November 2025, Revised Acts, https://revisedacts.lawreform.ie/eli/2007/act/19/revised/en/html

34. Competition and Markets Authority (United Kingdom), Unfair commercial practices (CMA207), banned practice 7, GOV.UK, 18 November 2025, https://assets.publishing.service.gov.uk/media/691b9bd821ef5aaa6543ee6f/Unfair_commercial_practices_CMA207_18_Nov_2025__2_.pdf