How should we use samples and trial sizes as offers?
Our answer
Give each sample one job. Add trial sizes of repeat-driving products to existing orders or above a spend threshold, and sell trial sizes to new customers with credit toward full size. Compare against a no-sample group.
- Confidence
- Medium
- Strongest proof
- 1 official rule
- 1 dataset, 4 practitioners, 8 opinions or claims
- Sources
- 12 sources
- 14 statements
R
BAJH
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentA trial size given above a spend threshold can replace a launch discount. RetentionX (retention analytics software maker) advises judging it on the sample's cost against the extra margin, not retail value against a discount percentage.
Giving samples to anyone who requests one draws people who never buy. ATTN Agency (paid media agency) cites 200,000 requests converting at 4% and calls samples inside other orders the most capital-efficient use. It puts a $2.50 sample's delivered cost at $9.40.
Counting sample recipients who later bought overstates a sample's results. RetentionX advises counting only first-time buyers of the full size, within a set window, against a holdout.
Our judgment
Start with trial sizes in existing orders or above a spend threshold, because it is the cheapest route and keeps the price intact. That goes by RetentionX's threshold example and ATTN Agency's view of cross-sampling. To win new customers, sell a trial size or discovery set with a credit toward the full size, or qualify requests, rather than giving samples to anyone who asks. That follows ATTN Agency's open-sampling example, the free-trial result reported on Scalability School and the discovery set with a voucher at Juliette Has a Gun. Judge samples against a group that got none over 90 and 180 days, following RetentionX's measurement rules and Brandmovers' advice to judge gifts on incremental margin against a control group.
Our judgment, built from RetentionX, ATTN Agency, Scalability School, Juliette Has a Gun and Brandmovers: our reading of these sources, not a finding from any one of them.
What the sources say
14 statements from 12 sourcesJuliette Has a Gun sells a $34 discovery set of seven scents that comes with a $34 voucher toward a full-size bottle, valid for two months. No results are given.
“comes with a $34 coupon to redeem on your next purchase of a full sized bottle!”
The FTC's guides say ads may promise money back or satisfaction only if the full price is refunded on request, and must clearly disclose any material conditions. US only.
“refunds the full purchase price of the advertised product at the purchaser's request”
GlobalPETS reports shoppers favor larger value packs, with small packs used as trials. An Attest UK survey found 42% lose trust and 42% stop buying when brands shrink sizes without cutting prices.
“four in five consumers react negatively when brands reduce product sizes without lowering prices”
A Flywheel report on 1.5 million TikTok Shop samples found only 47% led to a video and 14% of creators drove sales; creators posting 20+ videos averaged $74.07 a video versus $7.19 for a single post.
“only 47% of the time do those creators actually make a video, and only 14% of creators generate sales”
Brandmovers, which sells promotions, says a gift with purchase beats a discount when protecting price matters and the gift costs far less than it seems worth. Set the spend threshold above the median basket.
“gifts on orders that would have qualified anyway are pure cost.”
CTC, which sells TikTok Shop management, wants at least three videos back per sample and post rates of 50% to 70% from creators, and says finding the top 10 takes hundreds of samples.
“So if we send a sample, we want to get at least three videos back per that sample.”
Format Finder, a subscription app, moved from one $97 plan to $29, $59 and $97 tiers. Turning annual billing on by default moved AOV most, and free trials failed because Meta trial costs never paid back.
“the single change that moved AOV the most was toggling annual billing on by default”
RetentionX, which sells retention analytics, says a trial-size sample as a gift with purchase can replace discounting. Judge it on sample cost against incremental margin, ideally with a holdout group.
“The comparison should be sample COGS vs incremental CM1”
RetentionX argues that when results take about 90 days, a 30-day first unit runs out before proof arrives, and suggests a starter pack covering the evaluation window. A framework with no data.
“The first solution is to introduce a starter pack that covers the realistic evaluation window.”
Brand leaders in Digiday favor bundles and gifts with purchase over discounts at holiday, when people buy multiples. In its survey of 90 professionals, gift guide use fell 19 points from 2025.
“The cost of acquisition is high during holiday, so being able to tap an existing audience can be cost-effective”
Retail Brew reports CVS's beauty mini section, in 2,700 stores since last October, saw sales climb 140% year over year and is expanding to 1,200 more stores. One retailer's growth as reported.
“Since then, its beauty mini category sales have climbed 140% YoY”
Adwave suggests starter kits or discovery sets for first-time buyers and replenishment or volume bundles for repeat buyers, built from order data, starting with two or three bundles and checking margin after discount.
“Good bundles come from order data, not brainstorms.”
ATTN Agency, which sells paid media and email, says a $2.50 sample can cost $9.40 delivered and cites 5 to 30% conversion to first purchase by category. These are its client rules of thumb, with no method.
“$2.50 product sample actually costs $9.40 to get into someone's hands”
Goodwin, a law firm, says US law (ROSCA) still requires clear terms, express consent and a simple way to cancel subscriptions. The FTC restarted its cancellation rulemaking in 2026, and state renewal laws still apply.
“If a company’s cancellation pathway is significantly more difficult than its enrollment path, the FTC may view that imbalance as inherently unfair.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- The strongest disagreementThe best case against the answer, from a cited source
- What the research saysWhat academic research adds, kept apart from the answer
- How sure we areWhat its confidence rests on, and what would change the answer