Price Matching Policies: When to Match and When to Hold

Target and Newegg Now Match Only Themselves

A mass retailer and a large electronics specialist ran the competitor price-matching experiment at scale, and both have quietly concluded it was not worth it.

Target's guarantee now covers Target.com, Target stores and Target Circle offers only, inside a 14-day window. Competitor matching is gone, dropped effective 28 July 2025, and Target will not even cross-match between Target and its own Target Plus marketplace (Target, policy retrieved August 2026). Newegg has landed in the same place from a different direction: it matches only items sold and shipped by Newegg, on a 7-day window, and pays the difference as store credit valid for 365 days (Newegg, retrieved August 2026). Amazon, Best Buy, Walmart and Target are named explicitly as sites Newegg will not match.

The short answer to when you should match: on products a buyer genuinely comparison-shops, where the rival listing is the same item, in stock, and priced without a bundle or rebate attached. Hold everywhere else, and hold especially when the guarantee would apply to a catalogue far wider than the handful of items anyone actually cross-checks.

This post is the deep dive behind one line in the competitive pricing strategy guide, which places matching as one of five approaches. What follows is the part that guide does not have room for: what a published guarantee does to your competitors, what the research actually found, where competition law draws its line, and the exclusion clauses that decide whether the policy costs you margin or earns you retention.

A Guarantee Is Addressed to Customers and Read by Competitors

Here is the result that makes price matching more interesting than it looks: a price-match guarantee can raise the equilibrium price rather than lower it.

The mechanism is a pre-commitment. If I undercut you and you have a standing guarantee, you automatically match me for every customer who notices, so I buy no share from you. All I have done is lower the price I charge my own existing customers. Knowing that in advance, I do not cut at all. The promise is made to consumers, but it functions as a commitment to retaliate instantly, and your competitor is the one who reads it that way. The higher price is sustained with no agreement of any kind, which is precisely what makes it hard to police.

The empirical anchor is Hess and Gerstner, "Price-Matching Policies: An Empirical Case", Managerial and Decision Economics 12(4):305-315 (doi:10.1002/mde.4090120405, 1991), who looked at supermarket data and found prices rose after the policy was adopted. Corts, "On the competitive effects of price-matching policies", International Journal of Industrial Organization 15(3) (doi:10.1016/S0167-7187(96)01035-1, 1997), works the competitive effects through formally. Edlin carried the argument into law with a title that states the problem outright: "Do Guaranteed-Low-Price Policies Guarantee High Prices, and Can Antitrust Rise to the Challenge?", Harvard Law Review 111 (doi:10.2307/1342058, 1997).

There is a more recent field observation pointing the same way. After Shell Germany launched a card-holder price-matching guarantee in May 2015, stations in affected markets began coordinated midday price increases of roughly 3 cents, and the authors model the guarantee as a tacit-collusion device (Cabral, Duerr, Schober and Woll, SSRN working paper, 2021). Treat that one as suggestive rather than settled: it is a working paper and has not been peer reviewed.

State the mechanism precisely, because a loose version of it is wrong. Collusive effects require rivals to monitor and match in turn, and require consumers to actually be able to claim. Take away either condition and the story collapses.

The Literature Disagrees With Itself, and the Disagreement Is the Useful Part

Anyone who tells you price matching is a collusion device, full stop, has read one half of the field.

Arbatskaya, Hviid and Shaffer collected 515 low-price guarantees from newspaper advertising across 37 US metropolitan areas and found that most of them lacked the features a collusive guarantee needs: they typically covered only advertised competitor prices, and they imposed substantial hassle costs on anyone trying to claim (Journal of Law and Economics 47(1):307-332, doi:10.1086/386275, 2004). A guarantee that only applies to a price you can produce in print, and that takes an afternoon to claim, is not a credible instant-retaliation threat.

Hviid and Shaffer had already shown why that matters, in a paper whose title does the work: "Hassle Costs: The Achilles' Heel of Price-Matching Guarantees", Journal of Economics & Management Strategy 8(4):489-521 (doi:10.1111/j.1430-9134.1999.00489.x, 1999). Even small frictions in the claim process unravel the collusive equilibrium. The rival can no longer be sure you will match every defection, so undercutting becomes profitable again.

Then there is the reading that most sellers will recognise from their own motives. Moorthy and Winter, "Price-matching guarantees", RAND Journal of Economics 37(2):449-465 (doi:10.1111/j.1756-2171.2006.tb00025.x, 2006), conclude that available evidence supports the signalling theory: a guarantee is a credible signal to uninformed shoppers that you are already cheap. A retailer who expects to be undercut regularly cannot afford to offer one. Earlier still, Png and Hirshleifer, "Price Discrimination Through Offers to Match Price", Journal of Business 60(3) (doi:10.1086/296402, 1987), read matching as a way to separate informed shoppers from uninformed ones and charge them differently.

Three readings, three implied policies. Collusion says the guarantee stabilises a high price. Signalling says it buys trust from shoppers who will not check. Discrimination says it hands a discount to the minority who do. Which one dominates in your case depends on things you can observe: how easy the claim is, whether your rivals match too, and whether the price you are guaranteeing is one anybody can verify.

Where the Law Draws Its Line: Agreement, Not Matching

What follows is general information, not legal advice. Pricing policy is jurisdiction-specific and a competition lawyer should see anything you publish.

The line is not the matching. It is the agreement.

In the United States, the reference point is E.I. du Pont de Nemours & Co. v. FTC, better known as "Ethyl", 729 F.2d 128 (2d Cir. 1984). The FTC had challenged a set of facilitating practices, including advance price announcements and most-favoured-nation and delivered-pricing clauses, under Section 5 of the FTC Act. The Second Circuit set the order aside, holding that conscious parallelism combined with facilitating practices is not enough absent evidence of anticompetitive intent or collusion. Unilaterally publishing a match policy is not, on that authority, unlawful in itself. It does sit in a category regulators watch, which is a different statement from being safe by default.

The clearest teaching case on the other side of the line comes from the UK. In case 50223, the Competition and Markets Authority found that two online sellers had agreed not to undercut each other on Amazon UK, an agreement implemented through automated repricing software and running from March 2011 to July 2015. Trod Ltd was fined £163,371 and its managing director was disqualified for five years (Competition and Markets Authority, decision August 2016). Nothing about repricing software was the problem. The agreement was.

Three practical rules follow, and they are cheap to observe:

• Publish unilaterally. Your policy is addressed to your customers. It is a public commitment about what you will do, not a proposal to anyone else.
• Never discuss it with a competitor. Not at a trade show, not in a supplier group chat, not "just to understand where you are on this".
• Never accept an invitation to hold together. If a rival suggests you both stop discounting, that is the invitation the Trod decision punishes. Decline it in writing and keep the writing.

The general legality of algorithmic repricing is covered in the dynamic pricing explainer. The angle specific to matching is narrower: a guarantee is a facilitating practice by design, so the documentation trail showing you set it unilaterally is worth more here than it is for an ordinary repricing rule.

The Decision: Match Where It Is Checked, Hold Where It Is Not

Strip the theory away and the operating decision comes down to whether the shopper in front of you is actually comparing.

Match where demand is elastic and the product is comparison-shopped. Hold where demand is inelastic. The elasticity half of that sentence is worked out properly in the price elasticity guide; the second half is the one sellers skip. An identical, widely stocked SKU with a model number a buyer can paste into a search box is comparison-shopped. Your own bundle, your own configuration, a product where you are the only seller within two days' delivery: not comparison-shopped, whatever its elasticity. Matching there gives away margin to customers who were never going to leave.

Then price the worst case rather than the average. A guarantee costs nothing in a quiet month and everything during a rival's deepest promotional week, which is exactly the week you have least room to absorb it. Peak season is where a broad guarantee turns expensive fastest, and the phased approach in the Black Friday monitoring playbook is worth reading alongside this decision, because a policy written in March gets tested in November.

Three situations argue for holding even on a comparison-shopped product. When the rival's lower price is attached to stock they do not have, matching is pure margin loss against a listing nobody can buy. When the lower price is a clearance or open-box unit, you are not matching the same offer. And when you sell products under manufacturer pricing floors, a match rule can walk you straight into a violation, which is why MAP monitoring and a match policy need to be written together rather than in sequence.

The Exclusion Checklist You Can Lift Into Your Policy

Everything above is argument. This is the part you can copy. Each of these eight classes appears in at least one of the two published policies cited at the top of this post, and together they define the difference between a guarantee that costs margin and one that does not.

#	Exclusion class	What it keeps out	Evidenced by
1	Third-party and marketplace listings	Prices set by sellers the competitor does not control, including your own marketplace arm	Target excludes cross-matching with Target Plus; Newegg excludes marketplace sellers
2	Named flash and holiday events	Doorbusters and time-boxed event pricing you cannot sustain	Newegg names Black Friday, Cyber Monday, Doorbusters and Shell Shocker
3	Clearance, closeout, open-box, refurbished, used	Stock that is not the same offer as your new unit	Both policies
4	Bundles and minimum-purchase offers	Prices that only exist attached to something else	Both policies
5	Competitor pricing errors	Mispriced listings, usually worded as "typographical errors"	Both policies
6	Out-of-stock and limited-quantity items	A price on something nobody can actually buy	Newegg
7	Rebates, coupons, gift-card offers, financing, add-on services	Effective prices assembled from promotions rather than the listed price	Target excludes rebates and gift-card offers
8	Claim window and refund form	Everything outside a short window, and cash refunds where credit will do	Target 14 days; Newegg 7 days, refunded as store credit valid 365 days

Item 8 is the underrated one, and it is worth separating from the rest.

A short claim window plus a refund issued as store credit converts a margin giveaway into a retention mechanic. The customer who claims gets their difference back and now has a balance they can only spend with you. Newegg's version, 7 days and credit valid for 365, is close to the maximum defensible form of that: short enough that few claims land, long enough on the credit side that the money almost certainly comes back through the same checkout. Compare that with an open-ended guarantee refunded to the original card, which is a straightforward transfer out of your margin with no second transaction attached.

The exclusions also produce the effect the literature identified. Hassle costs are the reason the guarantee can be offered at all, and the reason it fails as the instant-retaliation threat the collusion model assumes.

Two Numbers Nobody Can Give You

Be suspicious of anyone quoting a current figure for what share of retailers operate a price-match guarantee, or for how often shoppers actually claim one. There is no reliable, current, publicly verifiable source for either, and the numbers that circulate are mostly repeated between marketing blogs without a primary source underneath.

What can be observed is the direction of travel. The two policies cited at the top of this post are both primary, both current, and both moved the same way: inward. That is a mass retailer and a large specialist telling you what they concluded after running the programme, which beats a survey statistic with no methodology attached.

The claim-rate gap has an explanation, too, and it is the hassle-cost finding again. Guarantees are cheap to offer precisely because they are awkward to claim. Any figure suggesting high redemption should make you ask what counted as a redemption.

Watching the Prices You Have Promised to Match

A match policy is a standing commitment against prices you do not control, which makes it the one pricing decision that genuinely requires monitoring rather than merely benefiting from it. If your policy names a competitor, you want to know they have moved before a customer arrives holding a screenshot.

Respot handles the mechanical half of that. You paste a competitor's product URL and it detects the product details automatically, then tracks price and stock changes per variation and alerts you when either moves. Per-variation tracking matters more here than in general monitoring, because a match claim is always about one specific size, colour or configuration, and a parent-level price tells you nothing about whether the claim is valid. Stock alerts cover exclusion 6 directly: a rival price attached to an out-of-stock listing is not a price you owe anyone.

Checks run on an adaptive cadence with a floor set by plan, so contested listings get looked at more often than quiet ones. The free plan covers 5 trackers with 7 days of price history and no credit card, which is enough to watch the handful of listings your policy is really exposed to. Paid plans run to 100, 400 or 2,000 trackers with longer history, which is the version you want if the policy covers a catalogue rather than a shortlist.

Write the Exclusions, Then Decide

The decision this post asks you to make is not whether price matching works. It is which of your products are genuinely being compared, because that set is almost always smaller than the set your guarantee would cover.

Draft the eight exclusions first and see what the promise looks like afterwards. If a policy narrowed that way still seems worth publishing, publish it, unilaterally, and price the worst week rather than the average one. If it does not, you have learned something at no cost, and you are in the company of two retailers who reached the same answer with considerably more data than either of us has.

Either way, the policy is only as good as your knowledge of the prices it points at. Start tracking the listings you have promised to match before you publish the promise.