Listing video, version 4

Compliance memo: Schwartz awareness options (images 1 and 2)

To: Compliance. None of this is in the copy. The copy treats every claim as true and live.

Image 1

"Proven on millions of real Shopify orders" (1A subtitle) and "millions of orders behind them" (1B headline)

• Needs a substantiation file: which count (orders processed, orders touched by an offer, or orders analysed), the date, and the source. "Millions" must hold on the day the listing is published.
• "Proven" implies tested outcomes, not volume of processing. If the number is processed volume, "proven" is stronger than the evidence. "Behind them" in 1B implies the offers are derived from those orders. Confirm whether the engine learns from the pooled orders or only from each store's own orders.
• Whose orders? If pooled across merchants, check the merchant terms and privacy policy cover using their order data in aggregate. GDPR and CCPA apply to EU and California shoppers' data.
• Shopify's App Store listing rules prohibit unsubstantiated or misleading claims. Check the final wording against the current requirements for listing images.

"Chosen by order data" (1A headline)

• Does it hold for a store with few or no orders on day one? A new install has no history. If the engine falls back to defaults or catalog data, "chosen by order data" is overstated for those stores.
• "Order data" can be read as the merchant's own data. If some picks come from cross-store patterns, the headline is not exact.

"Data-backed" (1B subtitle)

• Only OK if every offer type carries a data reason. Check each offer type before using it.

"for Shopify" (both)

• Trademark use. The wording follows Shopify's partner brand guidelines as long as it doesn't suggest Shopify made or endorses the app. Confirm no "Built for Shopify" implication, since that badge is a separate program.

Image 2

"Bigger orders from product page to cart" (2B headline)

• A performance claim. It needs evidence that merchants see larger orders, ideally a measured average uplift, and the baseline and period. Without that, FTC and App Store claim rules apply. The claim is also universal ("bigger orders") rather than typical or average.
• Pair with merchant results or a stated measurement method in the substantiation file.

"work together" (2A headline)

• Implies bundle and cart offers never conflict. Confirm the engine prevents discount overlap in every case (stacking, Shopify discount combination limits). The brief's own source docs mention overlap warnings, which suggests conflicts can still occur and are surfaced rather than eliminated.

"first-party analytics" (2A subtitle)

• Define internally. It means the analytics are collected and owned by the app and not a third-party tool. Confirm that matches the architecture, and that the privacy policy and data-processing terms describe it.
• If it implies the merchant owns the data, confirm export and retention rights in the terms.

"measured and A/B tested" (2B subtitle) and "A/B tests" (2A subtitle)

• Applies to both bundles and the cart drawer. Confirm A/B testing is live for bundles as well as the cart, and not for the cart only.
• Test results carry statistical limits. Any results shown in other listing assets must not imply guaranteed uplift.

"Bundles and a cart drawer" in one app (2B subtitle)

• Confirm both are in the same plan tier. If bundles or the A/B tests sit in a higher tier, the listing and pricing section must be consistent.

General

• Superlatives and comparisons: none used. No "best", "#1" or named-competitor claims.
• Brand: no "CRO Cart" anywhere. The copy uses no brand name at all.
• Cross-check the final chosen lines against the other images for consistency of the "millions of orders" number and the analytics description.

Compliance memo: direct-response-craft options (images 1 and 2)

To: Compliance. Nothing in this memo is in the copy. The copy treats every claim as true and live.

Image 1

1A "Data picks every upsell and bundle"

• "Every" is an absolute. If any offer is merchant-set, defaulted, or overridden by hand, "picks every" is not literally accurate. Needs confirmation that data selects all offers the engine produces, or the headline should be treated as describing the engine's output only.
• "Data" is unspecified. Whose data: the merchant's own orders, or pooled data from other merchants? If pooled, there is a data-use and privacy disclosure question under the app's privacy policy and Shopify's partner data rules.
• "Battle-tested on millions of orders" (subtitle): needs a defensible source (count, date range, which stores). If the orders belong to merchants other than the listing's audience, say so internally. The number must remain true at listing time and at every future review.

1B "Bigger orders, from offers chosen by data"

• This is an outcome claim about order value. Shopify's app listing requirements prohibit unsubstantiated performance claims and guaranteed results. "Bigger" is comparative, with no baseline stated. Needs evidence (A/B or before/after data) before use. It is the riskiest line in my set.
• "Proven on millions of orders" (subtitle): "proven" asserts a result, not just exposure. Stronger than "tested". Needs the same substantiation as above, and arguably more.
• Combined with the headline, a reader could infer a guarantee of higher order value on their store.

Image 2

2A "Sell more from product page to cart"

• "Sell more" is an unbaselined performance claim. Same substantiation concern as 1B, though softer.
• "Measured and A/B tested" is fine if all bundle and cart offers actually can be A/B tested. Confirm it covers bundles on the product page as well as the cart drawer. If A/B tests exist for cart offers only, the subtitle overstates.

2B "Prove every bundle and cart offer works"

• "Prove ... works" can be read as a guarantee the offers will work. The product gives the merchant the means to see whether they work, and the copy can be read as promising they do. Confirm the legal read of "prove" as "enable you to prove".
• "Every" assumes complete analytics and test coverage on both bundles and cart offers. Confirm there are no offer types (e.g. some bundle types, free gifts, rewards) without analytics or test support.
• A/B test "proof" depends on traffic. Stores with low volume may never reach a clear result, so "prove" is not reachable for every offer on every store. This is a customer-expectation risk, not a legal one.
• "First-party analytics": confirm that this term is used the same way in the product docs and the privacy policy, and that the data is in fact collected by the app on the merchant's store, not via third-party trackers.
• "Deep" is a subjective superlative-adjacent word. Low risk, but flagging it.

General

• No Shopify trademark or Shopify-branded phrasing in the copy ("Shopify" is not used), so no trademark-usage issue. If a later version adds "for Shopify", the app-name/endorsement rules apply.
• No competitor comparison is made.
• "Millions of orders" appears in both image 1 options. One source of truth is needed for that figure, and the same number must be used everywhere on the listing and in the video.
• No brand name is used; "Thunder" is not needed.

Compliance memo: conversion copywriting (Wiebe), v5 images 1 and 2

To: compliance team. From: copy option 03. None of this is in the copy. The copy treats every claim as true and live.

Image 1 options

1A "Upsells and bundles, proven before you launch" / "Battle-tested on millions of orders, chosen from yours."

• "Proven" is an objective performance claim. We need evidence that it holds for the engine's defaults, and a clear reading of what is proven (reliability, or results). The listing could be read as promising results.
• "Before you launch" implies validation has already happened for this merchant's store. That is not true of any single store. Check that nothing reads as a guarantee.
• "Millions of orders": needs a documented count, its date, and its source. It must be orders touched by the engine, not total merchant volume.
• "Battle-tested" is informal but still a reliability claim. The evidence should exist: uptime, incidents, peak-traffic handling.
• "Chosen from yours" says the engine picks offers from the merchant's own order data. This must match how the product works in practice, including the minimum order history. A new store has no history.

1B "Your orders decide which upsells run" / "Upsell and bundle engine, battle-tested on millions of orders."

• "Decide" suggests the data runs things automatically. Check against the product, where the merchant or agent approves and publishes. Overstating automation is a misrepresentation risk.
• Same millions-of-orders substantiation as above.
• Processing merchant order data to choose offers: confirm the privacy policy and Shopify data-use terms cover this, and that no shopper PII appears in screenshots.

Image 2 options

2A "Bigger orders from product page to checkout" / "One app: bundles, cart drawer, measured and A/B tested."

• "Bigger orders" is an outcome claim. It could read as a promise that order value rises. We need data to support it, or a plain reading as the aim of the features.
• "To checkout": the cart drawer and product-page bundles do not sit in checkout itself. Confirm the phrase does not imply changes to Shopify checkout, which carries restrictions (Plus-only extensibility, and so on).
• "Measured": the analytics must be first-party and complete, as in the owner's image 4. Keep the wording consistent.
• "A/B tested": confirm that A/B tests run on both bundles and the cart drawer.
• "One app": confirm both products are in a single install and a single bill.

2B "See which bundle or cart offer wins" / "One app, first-party analytics, A/B tests on both."

• "Wins" implies a result is always available. Statistical power depends on store traffic. A test may run without a clear winner.
• "First-party analytics" must be true: collected by the app, not by a third party, and access to it must be full. Check data residency and consent-mode handling.
• Same single-app confirmation as 2A.

Cross-cutting concerns

1. Shopify App Store listing rules. Check the rules on image and listing text: superlatives, unverifiable performance claims, no pricing in images, no ratings or review quotes unless real, and no mention of other apps. None of my options uses a ranking or "best" claim.
2. Competitor reviews as research. VOC came from public 1★ and 2★ reviews of other apps. No competitor is named and none of the quotes is used in copy. The reviews contain store names and developer replies that dispute some claims. Treat the quotes as perception, not fact. None of it is used to disparage anyone.
3. Implied comparison. "Proven", "arrives proven" and "one app" are implicit contrasts with other apps. Check that this does not amount to comparative advertising that needs substantiation.
4. "Data-backed" family. Both image 1 options rely on data language. Our data sources must be described accurately: order history, plus any aggregate learning across stores. If cross-store data is used, check merchant consent and contract terms.
5. Agent and AI wording. I deliberately kept AI out of the image 1 and 2 headlines. Agent claims belong to image 3 and the AI-native image, where the owner has already approved wording.
6. Brand. "CRO Cart" is not used. No brand name appears in the copy.
7. Translation and markets. Claims should survive translation: "proven" and "battle-tested" may carry stronger legal weight in some jurisdictions (for example, EU consumer-protection reading of "proven").

Compliance memo: copy v5 / 04 Positioning (Dunford), images 1 and 2

To: compliance team. From: copy lead. Re: the four headline options in 04-positioning-dunford.md.

None of these concerns appear in the copy. Every claim was written as true and live at launch. No Shopify trademark is used and the brand name is not in the copy.

Rules cited (Shopify App Store requirements):

• 1.1.4: factual information only
• 4.3.3 and 4.3.4: no statistics or data in listing content or images; no "first/best/only"; no guarantees
• 4.4.4: images show the actual UI

Severity: High = likely flagged or needs substantiation; Medium = depends on reviewer or launch scope; Low = minor wording risk.

Compliance memo: 05 JTBD / switching forces (copy v5, images 1 and 2)

To: compliance team. Nothing here is in the copy. The copy treats every claim as true and live. These are the points to check before publishing.

Claims in the copy that need substantiation

1. "Proven at scale" (1A), "proven on millions of orders" (1B), "battle-tested on millions of orders" (1A subtitle). These are performance and scale claims. Confirm the order count and its source. Say whose orders they are: orders processed by the app, or orders analysed. Confirm the figure is still "millions" at publish date. "Proven" can read as a results guarantee, so check that the evidence shows the engine works and not only that it ran.
2. "Data-backed" (1A subtitle) and "chosen by data" (1B). Confirm that offers are in fact selected from order data in the default flow, not only when the merchant asks for it. Check that merchants can see the data basis, if "data-backed" is read literally.
3. "Raise order value" (2A). This is an outcome claim. No result is promised or quantified, but a reviewer may read it as an implied earnings claim. Check that we have evidence for typical uplift, and that the wording is accurate for stores where an offer loses an A/B test.
4. "Before and after add-to-cart" (2A). Bundles are on the product page and the drawer is after add-to-cart. Confirm there are no bundle placements that break this (for example bundles inside the drawer), so it is not misleading about where each lives.
5. "Proven before launch" (2B). Risk: an A/B test runs on live traffic, so the offer is already live to some shoppers while it is being tested. "Before launch" may overstate. If compliance is strict, this is the weaker option on accuracy. Suggested fix if needed: "proven before rollout", or drop to 2A.
6. "Measured" and "first-party analytics" (2A and 2B subtitles). Confirm the definition: data collected by the app itself, owned and accessible to the merchant. Confirm the "every offer" scope covers both bundle and cart offers.
7. "A/B tests on every offer" (2B). Confirm that every offer type, bundle and cart, can be put in an A/B test. If some types can't, "every" is false.
8. "One app" (2A, 2B). Check that bundles and the cart drawer are one install and one billing relationship.

Shopify App Store listing rules to check

• No guarantees of results, no unsubstantiated superlatives or comparisons. "Proven" and "battle-tested" are the nearest to this line. We have no comparative claim against another named app.
• No pricing, ratings or reviews on listing images or in the title text.
• Screenshot and title text must match the shipped product. This includes the A/B and analytics claims in the subtitles.
• "Add-to-cart" is a generic term. "Shopify" does not appear in the copy. If it is added later, check Shopify trademark usage rules.
• "Thunder" is the only brand name allowed. It is not used in these options.

Privacy and data

• "Chosen by data" and "data-backed" imply the app processes merchant order data. The listing's privacy policy and data-use disclosure should say this, including any use for AI or recommendations.
• "First-party analytics" means shopper behaviour data collected on the merchant's store. GDPR and ePrivacy apply for merchants selling to EU/UK shoppers (consent, cookies or storage). Confirm that how the analytics collects data is covered by the privacy policy, and that the merchant's consent banner covers it.
• If "millions of orders" is built on data pooled across merchants, check that the merchants' terms allow this use and that the data is anonymised.

Advertising and consumer law

• The phrases "proven", "battle-tested", and "data-backed" are objective claims. Under EU/UK/US rules we should hold the evidence file before using them, and keep it current.
• A buyer could reasonably take "Raise order value" as meaning their own order value will rise. Consider where the listing shows the qualifier, for example in the description. The qualifier is not in the image copy.

Internal consistency

• 1A and 1B use the same scale claim. If the number changes, both must change. They also line up with image 4 ("Know what every offer earns"): the analytics wording in 2A and 2B has to match that image's subtitle (first-party, complete user-behaviour analytics).
• The headline claims for images 1 and 2 must agree with the AI-native image: the agent can run bundle and cart offers and A/B experiments.
• Brief says never "CRO Cart". This memo and the options use no product name.

Compliance memo: 06 StoryBrand options (images 1 and 2)

Nothing here is in the copy. The copy treats every claim as true and live.

Claims to substantiate

1. "Battle-tested on millions of orders" (1A, 1B subtitles). We need a documented count, a definition of "orders" (orders the app touched, orders analysed, or orders with an offer shown), and a date. The claim must stay true as numbers change. Check whether the figure includes test or development stores, and whether it is "millions" in the plural across the whole base or per merchant.
2. "Proven" / "arrive proven" (1A, 2A). This can read as a guaranteed outcome or as pre-validation on the reader's own store. The evidence is aggregate order data and not a result for each merchant. The Shopify App Store prohibits unsubstantiated performance claims. Decide whether "proven" needs a support document, or whether we should fall back to "data-backed".
3. "Data-backed" (1B headline). Define what the data is (the merchant's own orders, aggregate orders across merchants, or both). If it includes cross-merchant data, check privacy-policy coverage and merchant consent.
4. "Lift you can prove" (2A). This implies the analytics and A/B tests demonstrate incremental lift. The tests must support that: attribution method, holdout design, significance handling. A merchant who reads this could expect positive lift. The copy does not promise lift, but the headline names it as the benefit.
5. "Deep first-party analytics" (2A subtitle). "Deep" is a subjective superlative. "First-party" needs a consistent definition: data we collect directly from storefront events and orders, not a third-party tool.
6. "Never clash" (2B). This is an absolute. Shopify discount combination rules and overlapping offers can still conflict. Earlier materials mention overlap warnings, which means clashes can be detected, not that they are impossible. If 2B is chosen, check what the engine guarantees.
7. "Your busiest days run on..." (1B). This implies reliability and performance at peak traffic (BFCM and similar). We would need load or incident data. Also check the SLA wording.
8. "Shopify" in the 1B subtitle. Trademark and branding: use must follow Shopify's brand guidelines and App Store listing rules (for example, no implying endorsement, and the app name or description rules).
9. "Engine" / "upsell and bundle engine". This is descriptive. Check that it matches the category the app is listed under.

Listing policy checks

• No superlatives or comparisons ("best", "#1", "replaces X") appear in these options. Keep it that way in any edit.
• No competitor names, pricing, or free-trial language.
• Headlines and subtitles stay within the length limits (headline ≤ 7 words, subtitle ≤ 10).
• Brand: copy uses neither "CRO Cart" nor "Thunder" in the headlines.

Other concerns

• 1B uses "Your busiest days" to imply stakes. This is mild loss framing and not a scare claim, but a reviewer may read it as an implied outcome.
• 2B's "clash" frames other cart and bundle setups as a problem. It names no competitor, but check it against the App Store rule on disparaging comparisons.
• Any later edit that adds a number (percent lift, order count) needs its own substantiation.

Compliance memo: 07 concrete-falsifiable-dry (images 1 and 2)

Nothing here is in the copy. These are the concerns I had while writing.

Image 1

1A "Upsells picked by millions of orders" / "Data-backed upsell and bundle engine for Shopify"

• "Millions of orders" needs a documented, current count and a definition: orders processed, orders analyzed, or orders that carried an upsell. "Picked by" implies the orders influenced the choice of offers, so the substantiation should be about that, not just volume.
• If the orders include other merchants' data used to choose or rank offers, check the merchant data-processing terms, privacy policy and Shopify's protected customer data requirements for cross-merchant use. "Picked by millions of orders" could read as cross-merchant learning.
• "Data-backed" should be traceable to a described method (what data, which model or rule).
• "Shopify" appears in the subtitle in a descriptive sense. Check it is consistent with Shopify's app listing and trademark guidance, and that nothing suggests endorsement.
• "Upsells" alone in the headline leaves "bundles" to the subtitle; confirm the App Store listing name and category are consistent with "upsell and bundle engine".

1B "Upsells that report what they earn" / "Data-backed offers, A/B tested, proven on millions of orders"

• "Report what they earn" is a revenue-attribution claim. The attribution method (last-touch, A/B holdout, assisted) should be documented, and the merchant should be able to see it. Order-level revenue may overstate incremental revenue if no holdout is used.
• "Proven" is an absolute word. It sits beside "A/B tested" and "millions of orders"; confirm both are literally true for the product as shipped, not only for some features.
• "A/B tested" claims the product runs tests, not that every offer was tested. Make sure the screen it appears on does not suggest the second.

Image 2

2A "Bigger orders, from product page to cart" / "Bundles and cart drawer, measured and A/B tested"

• This is an outcome claim about merchants' revenue. Shopify's app listing rules and general advertising-substantiation rules expect support for performance claims. We need data on order value lift, with sample size and comparison method, or a basis that it is a typical result.
• Wording is comparative ("bigger") without a stated baseline. A baseline (versus no upsell or bundle) should be documented.
• "Measured" assumes first-party analytics cover both bundle and cart in the same view. Check this holds for every plan and market.

2B "Bundles and cart upsells that never clash" / "One app. Every offer measured and A/B tested."

• "Never" is absolute. If the product warns about overlaps rather than preventing them (the earlier listing copy used "overlap warnings"), "never clash" overpromises. Define what "clash" means (stacked discounts, duplicate offers, conflicting Shopify discount combinations) and confirm the product prevents it in every case, including with other apps' discounts and Shopify native discounts. Third-party apps and manually created discount codes are the likely gap.
• "Every offer measured and A/B tested" could read as "every offer is automatically tested". Confirm it means "can be measured and tested". The subtitle is a claim about capability, so the screenshot or video must not imply automation if there is none.
• "One app" is a competitive claim by implication (against two-app setups). It is not naming a competitor.

General

• No hedges were put in the copy per the brief; every claim above is therefore stated flat, and each needs substantiation on file.
• Neither "CRO Cart" nor "Thunder" appears in these lines. If the brand is added later, check the listing name rules for the App Store.
• "Millions of orders" appears in both image 1 options; the figure and its date should be the same everywhere it is used (listing text, video, images).
• Image 4 ("Know what every offer earns", first-party analytics) is already chosen. 1B repeats part of that idea (revenue per offer); if that overlap is unwanted, 1A avoids it.

Compliance memo: Ogilvy headline craft, images 1 and 2

None of this is in the copy. Everything in the copy is written as true and live. These are the points to check before publishing.

Image 1

1A "Upsell and bundle engine for 8-figure stores" / "Every offer data-backed, proven on millions of orders"

• "8-figure stores" is an audience qualifier. It implies a revenue threshold. Confirm that stores below that threshold can install without friction, and that nothing reads as a restriction.
• "Every offer data-backed": confirm that every offer the product creates, including offers a merchant or agent writes by hand, is backed by data. If some offers are manual, "every" overstates.
• "Proven on millions of orders": confirm the total, what counts as an order, whether the orders come from many stores or one, and that the merchants whose orders were counted allowed their use as aggregate proof. Keep dated evidence for the number.
• "Proven" can read as a guarantee of results. Check Shopify App Store rules on performance claims and on proof wording.

1B "Thunder raises order value with data-backed offers" / "The upsell and bundle engine proven on millions of orders"

• "Raises order value" is an outcome claim. It needs substantiation: before-and-after or A/B evidence, and a clear basis such as average across stores. Check whether Shopify or advertising rules require a stated basis for a performance claim.
• Present tense and universal ("raises") may be read as applying to every store. Check that this holds for the stores we measured, including stores with few orders.
• "Data-backed": confirm what data. Merchant's own orders, pooled data across stores, or both. If pooled data crosses stores, check the privacy terms and the merchant data-use terms.
• Same "proven on millions of orders" points as above.
• The brand "Thunder" appears in the headline. Confirm trademark clearance and that the listing name and the headline name match.

Image 2

2A "Sell more from product page to cart" / "One app: bundles and cart drawer, measured and A/B tested"

• "Sell more" is a comparative outcome claim with no baseline. Check whether it needs a basis or evidence on file.
• "Measured": confirm analytics cover both bundles and cart drawer with the same definitions, so "measured" is not read as "attributed revenue is exact".
• "A/B tested": confirm A/B testing covers both bundles and cart offers today. If testing is only available for part of the cart, this overstates.
• "One app" is true only if no extra app, embed or theme edit is required for either piece. Check install steps.
• Merchant analytics use shopper behaviour data. "First-party" is used in the Option 2B subtitle. Check consent, cookie and privacy-policy wording for the markets we sell in (GDPR/ePrivacy, CCPA). Check that first-party is described accurately.

2B "Only winning bundle and cart offers stay" / "One app runs both, with first-party analytics and A/B tests"

• "Only winning offers stay" implies the product removes losers automatically. In practice the merchant decides. Check that the product does what the headline says, or that merchants cannot read it as an automatic promise.
• "Winning" needs a definition: which metric (revenue per visitor, conversion, or another), what sample size, and what significance. Check that the product does not declare winners on thin data.
• "First-party analytics": confirm the data is collected on the merchant's own domain, and that the wording is accurate for cart events and for orders.
• Same "One app" install check as 2A.

General

• Word counts treat hyphenated compounds (data-backed, 8-figure, first-party) as one word. If the listing tool counts them as two, 1A and 1B move to 8 headline words and the 1A subtitle to 9.
• The copy uses "Shopify" nowhere, so there is no third-party trademark issue in these lines. Option 1A names stores generally, not Shopify.
• No testimonials, ratings or third-party names appear in any of the four headlines or subtitles.
• The headlines avoid "best", "#1", "guaranteed" and "revolutionary". Check whether "proven" and "winning" need to be treated at the same level.
• Check with whoever owns the App Store listing review: some listing reviews treat "millions of orders" as a claim that needs a source at submission.

Compliance memo: behavioral-psychology options (images 1 and 2)

To: compliance. None of this is in the copy, because the copy is written as if every claim is true and live. These are the points I'd want you to check before it ships.

Image 1

1. "Millions of orders" (1A headline, 1B subtitle). Needs a defensible source: orders that passed through Thunder offers, not total store GMV or orders on stores that merely installed it. State which count it is, over what period, and who counted.
2. "Behind millions of orders" (1A). "Behind" implies Thunder produced or influenced those orders. If the number covers orders where no Thunder offer was shown, the wording overstates.
3. "Data-backed" (1A, 1B). Implies offers are selected from actual data. Confirm that holds for default setups and new stores with little order history, not only mature ones.
4. "Battle-tested" (1B subtitle). Puffery in normal use, but it sits next to a number, so it will be read as a factual claim. Same substantiation as point 1.
5. "How shoppers actually buy" (1B). Implies behavioural data on shoppers is processed. Check the privacy policy and the Shopify data-processing disclosures (GDPR/UK GDPR, CCPA/CPRA) cover order-level and session-level analysis, and that this is the same data the listing says is first-party.
6. "Upsell engine" / "for Shopify stores". Nothing here suggests Shopify endorses or certifies the app. Keep it that way in the final layout (no Shopify logo marks beyond the permitted badge).
7. Overlap with prior rounds. 1A and 1B are new constructions, but they carry the same "millions of orders" message as the v4 H slot, so any substantiation issue found there applies here too.

Image 2

1. "Raise order value" (2A). An outcome claim. Even phrased as a goal, it can be read as a promise of uplift. Check there is evidence for an average uplift, and whether the app-store reviewers and ad regulators (FTC, ASA, Shopify App Store requirements on performance claims) would want a "results vary" qualifier outside the image. No figure is stated, which helps.
2. "Product page to cart" (2A). Check the bundle widget works on all supported theme types (Online Store 2.0, headless) as the listing implies. Cart drawer coverage should match.
3. "A/B tested" (2A subtitle). Confirm the A/B testing works for both bundles and the cart drawer, not just the drawer.
4. "Keep winners" (2B). Implies a winner always emerges. Some tests are inconclusive, and the product may declare no winner. Same for "only offers that earn stay": if the product does not auto-remove losers, "keep winners" is merchant action, not automatic.
5. "Test bundles and cart offers" (2B). Check experiment scope: if bundle experiments are not available, this overstates.
6. "Deep first-party analytics" (2B subtitle). "First-party" means data the merchant collects through their own storefront. Confirm Thunder's data collection matches that definition, and that cookie/consent behaviour is described correctly (EU and UK consent rules).
7. "In one app" (2A, 2B subtitle). Fine as a fact if both modules ship in one install. Avoid pairing it with anything that says "replaces your other apps", which G already rejected and which would also be a competitor-comparison claim.

Psychology ethics

1. No scarcity, urgency, countdown or fear-of-missing-out devices are used in either image. I used loss and regret aversion only in the sense of "removing the fear of a bad offer" (2B). If anyone strengthens that into "stop losing sales", it would need a basis.
2. Social proof is a number, not a testimonial or a named logo. No customer endorsements or logos are implied.
3. Brand: "Thunder" is not used in any headline or subtitle. "CRO Cart" appears nowhere.

Word counts

All headlines are 7 words. Subtitles: 1A 8, 1B 9, 2A 9 ("A/B tested" counted as two words), 2B 8.

Compliance memo: 10 video-hook-tactics, listing images 1 and 2

To: Compliance team. None of these concerns appear in the copy. Each line below needs a yes/no from you before publishing.

Image 1

1A "The upsell engine that proves every offer"

• "Proves every offer": implies each offer has measured results available. Confirm that analytics and A/B testing cover every offer type (upsells, bundles, cart-drawer offers), not only A/B-tested ones. Merchants with low traffic may see "not enough data" and read "proves" as a guarantee.
• "Proves" can be read as "proves it works" (a performance guarantee) rather than "shows the evidence". Confirm the meaning we can stand behind is "shows what each offer earns".
• Subtitle "battle-tested on millions of orders": need a source for the order count (what exactly is counted, which period, across how many shops). Confirm "millions" is true on the date of publication and that it counts orders processed with Thunder offers, not all orders on installed stores.
• "Upsells and bundles": confirm Shopify App Store rules allow a category claim ("engine") without a rating or review reference.

1B "Upsell and bundle engine for 8-figure stores"

• "8-figure stores" implies annual revenue of at least 10 million (in what currency?). Confirm we have merchants at that tier and that the claim is "built for" and not "used by". Check that it is not read as a restriction (smaller stores excluded) or an endorsement by those stores.
• Subtitle "data-backed": the strength of the data claim depends on what the data is (merchant's own orders, a pooled dataset, or research). Confirm which one, and whether pooled data across merchants is covered by our terms and privacy policy.
• "Proven across millions of real orders": same sourcing requirement as 1A. "Proven" is a stronger word than "tested" and may be read as a results guarantee; confirm we are comfortable.

Image 2

2A "Bigger orders, product page to cart drawer"

• This is an outcome claim (higher order value). Confirm we have evidence that offers increase order value, for what share of merchants, and over what period. If not universal, a regulator may read it as a typical-results claim.
• It compares implicitly with the merchant's current orders; there is no baseline stated. Confirm that is acceptable.
• "Product page to cart drawer": confirm both placements are available on all plans and all themes, including the online-store 2.0 requirement for product-page blocks.
• Subtitle "A/B tested and measured": confirm A/B testing is available to every merchant who installs, not only paid tiers, and that "measured" refers to first-party analytics owned by Thunder (not Shopify analytics).

2B "Bundles and cart offers that work together"

• "Work together" implies no discount conflicts. Confirm what the discount engine actually guarantees: does it prevent stacking with other apps' discounts, or only with Thunder's own offers? A merchant who runs another discount app may read this as a guarantee across apps.
• Subtitle "one discount engine": confirm technically accurate, given Shopify discount functions limit on stacking and the combinations rules.
• "Every offer measured, A/B tested": same availability check as 2A.

General (all four options)

• Brand: "Thunder" is not used in these options. If it is added later, confirm trademark clearance for use alongside Shopify marks.
• Numbers, "millions", "8-figure": ensure the evidence file is kept so the claim can be substantiated on request (App Store review, advertising-standards complaint).
• "First-party analytics": confirm the data-collection and consent position (cookies, GDPR/ePrivacy for EU shoppers) behind the analytics promise is complete, since two of the four subtitles lean on measurement.
• None of the copy mentions AI, agents or guaranteed revenue; if G later merges these with image 3 or 4 copy, re-check those claims separately.