A commute-specific message or a broader morning routine? Two matched concepts and a proposed test isolate the use occasion as the creative question.
Explore the proposed creative
Marianne Jaen
Creative Strategy & Media Buying
Creative diagnosis, hypotheses, copy, concept layouts, production brief and proposed test.
Two proposed static ads, a production brief and a testing framework. The campaign has not launched.
The historical morning-labelled ad (H3) had the higher link click-through rate (CTR). The commute-labelled ad (H1) recorded lower cost per purchase (CPA) and higher return on ad spend (ROAS), making message relevance a question worth testing.
Historical simulation · September 15–28, 2026 · H1 and H3 testing only. Bars start at zero. New concepts have no performance results.
H1 provides the stronger commercial reference: 15 purchases, $30 CPA and 2.50× ROAS. Its supplied message centres on the commute.
H3 recorded 3.00% link CTR and 1.50× ROAS. With only six purchases, the sample is too sparse for a reliable winner or loser decision.
| Observed evidence | Interpretation | Creative implication | Limitation |
|---|---|---|---|
| H1: 15 purchases, $30 CPA and 2.50× ROAS. Its supplied hook references the commute. | The commute-labelled execution provides a promising reference for a new message hypothesis. | Extend: develop a commute-context headline without assuming the original execution caused the result. | No finished H1 asset inspected; 15 purchases and incomplete delivery context do not establish a winner. |
| H3: 3.00% link CTR, six purchases and 1.50× ROAS. Its supplied hook references the whole morning. | Attention alone did not translate into stronger observed purchase efficiency. | Test relevance: compare a broader morning message with a specific commute message. | No click or impression counts; purchase conversion cannot be calculated. Technical or landing-page issues remain possible. |
| H1 and H3 use different supplied messages; other execution differences are unknown. | Historical differences cannot be attributed to the message alone. | Explore: create a new matched pair, changing only the context word. | The new pair will not explain the historical performance gap. |
| Product performance substantiation and approved final assets are unavailable. | Leak-proof and heat-retention claims require product substantiation. | Keep claims supportable: use neutral copy and an illustrative product layout. | Any benefit-led variation requires verified claims and approved product assets. |
| Historical ad | Spend (CAD) | Purchases | CPA (CAD) | ROAS | AOV (CAD) | Link CTR |
|---|---|---|---|---|---|---|
| H1 testing | $450 | 15 | $30.00 | 2.50× | $75 | 1.50% |
| H3 testing | $220 | 6 | $36.67 | 1.50× | $55 | 3.00% |
CPA (cost per purchase) = spend ÷ purchases. ROAS (return on ad spend) = attributed revenue ÷ spend. AOV (average order value) = attributed revenue ÷ purchases. H1 and H3 imply approximately $1,125 and $330 revenue from rounded scenario AOV. Attributed revenue is not profit or incremental revenue.
H2’s $180 spend, two purchases and reported 4.00× ROAS imply approximately $360 AOV; order values need verification. H1’s scaling records conflict: the analysis reports 1.20× ROAS and an earlier project record reports 1.50×. Without matching spend, purchases and revenue, scaling remains excluded. Click and impression counts are unavailable, so no aggregate CTR or click-to-purchase rate is calculated. Small purchase counts and unmatched delivery conditions limit interpretation.
Both proposed ads use the same product image, layout, supporting line and call to action (CTA). Only “commute” versus “morning” changes. Concept A is the commute version; Concept B is the morning reference.
A mug for your daily routine.
Explore Cozy CupFor people taking coffee to work, naming the commute may make the mug feel more immediately relevant.
A mug for your daily routine.
Explore Cozy CupFor people whose routine does not involve commuting, a morning message may offer a more relevant starting point.
Proposed concepts use an AI-generated product illustration. Final production requires approved product photography and confirmation of the green-and-cream brand treatment.
A commute-specific message may make the mug more relevant to shoppers who take coffee to work. H1 informs this hypothesis; it does not prove the message caused its results.
Challenge: more clicks without better purchase outcomes would weaken the case.
For people with varied morning routines, a broader message may strengthen relevance because it does not depend on commuting.
Challenge: attention without viable purchase efficiency, or stronger comparable evidence for the commute version, would weaken the case.
| Copy element | Concept A · Commute | Concept B · Morning routine |
|---|---|---|
| Headline | Your coffee. Your commute. | Your coffee. Your morning. |
| Supporting copy | A mug for your daily routine. | A mug for your daily routine. |
| CTA | Explore Cozy Cup. | Explore Cozy Cup. |
| Proposed platform text | Meet Cozy Cup. Explore the mug for your daily routine. | Meet Cozy Cup. Explore the mug for your daily routine. |
| Proposed platform headline | Explore Cozy Cup. | Explore Cozy Cup. |
Main message: a mug for an everyday coffee routine. The visual presents the product clearly; it is not evidence of performance. Both use “Explore Cozy Cup” to invite product consideration without adding an offer.
This new pair tests context specificity under matched conditions. It cannot explain the historical H1/H3 performance gap. Images, offers and benefit claims would be separate test variables.
Build a matched pair for the same eligible coffee-mug shoppers. Preserve the message comparison from concept through final export.
Production requirements: approved product photography, confirmed brand guidance and a verified destination. The two concept layouts are complete; final exports depend on those inputs.
| Brief element | Direction |
|---|---|
| Objective / audience | Develop a matched static pair to test commute-specific versus general morning relevance among the same eligible coffee-mug shoppers. Audience eligibility is a launch assumption, not a verified customer profile. |
| Format / placement | Proposed 1080 × 1350 px (4:5) feed static. Confirm actual placements and safe areas. Produce identical adaptations if other placements are included. |
| Assets needed | Approved product photography or packshot, actual product details, brand color/type guidance, destination URL and confirmed availability. The concept rendering is illustrative; final imagery must match the actual product. |
| Opening / hierarchy | Headline at top → dominant mug image → shared supporting line → CTA. Keep identical font, line breaks, scale, crop, margins and contrast. |
| Message / copy | Use the exact pair above. Both versions use identical platform text and headline. Do not introduce a sale badge, performance claim or new benefit. |
| Variations to produce | Concept A: “Your coffee. Your commute.” Concept B: “Your coffee. Your morning.” Replace the illustration with the same approved image in both files. |
| Brand requirements | Approve Cozy Cup’s palette and typography, and verify that the depicted mug matches the actual product. Apply the same treatment to both versions. |
| Preserve | Image, background, supporting copy, CTA, dimensions, offer terms, destination and all other design choices. Document any unavoidable differences. |
| Required approvals | Product representation, copy, brand treatment, placement specifications and landing-page continuity. If the final assets change the proposition, review the hypothesis again. |
| Scope of the completed deliverable | Concept layouts, production brief and test framework are complete. Live production would require approved assets, launch ownership and test sign-off. |
Compare Concept A (commute) with Concept B (morning) in a new concurrent test. Historical H1 and H3 inform the hypothesis and are excluded from the test comparison.
Commute versus morning. One context-word change.
Audience, offer, page, objective, attribution and placements where feasible.
Purchase ROAS, interpreted with CPA, purchases and AOV.
Develop a supported message; keep sparse or unreliable evidence inconclusive.
CTR and cost per click (CPC) describe attention and traffic efficiency; purchase outcomes determine the commercial assessment. ROAS requires reliable purchase values and is not profit. If multiple creative elements change, the comparison becomes a concept test.
| Test | Hypothesis | Control or reference | Variation | Primary metric | Diagnostic metrics | Decision rule |
|---|---|---|---|---|---|---|
| Test 1 · Proposed headline comparison | A concrete commute context may improve purchase relevance versus a broader morning context. | New Concept B “Your coffee. Your morning.” Historical H1/H3 are context only. | New Concept A “Your coffee. Your commute.” Only the context word changes. | Purchase ROAS, subject to verified revenue. | CPA, purchases, AOV; link CTR, CPC and landing-page views. Purchase/landing-page-view rate only with valid counts. | Develop the stronger message only if mature, sufficiently informative purchase evidence supports a commercially meaningful advantage under comparable delivery. Otherwise retain uncertainty and investigate or continue within the agreed risk limit. |
| Interpretation issue | Practical treatment |
|---|---|
| Low delivery / few purchases | Mark inconclusive. Do not declare a loser from a sparse sample or force additional spend beyond the agreed loss limit. |
| Unequal spend / delivery | Compare exposure and delivery mix. Shared-ad-set allocation is screening evidence, not automatically a controlled comparison. Avoid assuming equal budgets create equivalent samples. |
| Delayed conversions | Agree an attribution-maturity window before evaluation; allow for conversion lag and use the same reporting cutoff. |
| Uncertain order values | Validate revenue, currency, duplicate events and refunds. Do not decide on ROAS while purchase values remain unreliable. H2 remains outside a scaling recommendation. |
| Budget and duration | No amount or fixed timeline is proposed without economics. Agree total risk, expected CPA, meaningful detectable difference, traffic requirements and conversion lag before launch. |
| Scaling | A promising test earns further validation. Higher-budget delivery requires separate evidence and profitability checks. |
Use a concurrent randomized split where feasible. Shared-ad-set allocation can screen concepts but unequal delivery does not automatically provide a controlled comparison. Agree risk limits, useful evidence, attribution maturity and evaluation timing before launch; no universal budget or duration is assumed.
Before launch: validate purchase events and values; inspect mobile landing page and checkout; confirm stock, price and claims; agree acquisition economics and test budget; approve final exports and launch ownership.
The completed work translates the simulated performance pattern into two matched concepts, a production brief and decision criteria. The concepts have not launched, so no performance improvement is established.
Purchase outcomes give the commute-labelled execution priority for further investigation. Changing one context word creates a focused comparison without assuming what drove the historical results.
Confirm product assets, check the purchase journey and tracking, then agree test economics with the account owner before preparing the final pair.
Preserve a message supported by mature purchase evidence.
Fix an evidenced weakness in the message or purchase journey.
Resolve low delivery, conversion lag or unreliable values.
Respect agreed loss limits and unsupported-claim boundaries.
| Question | Current evidence | Test status | What would change the decision | Next action |
|---|---|---|---|---|
| Does a specific commute context help? | Historical H1 is promising, but the cause is unknown. | Test 1 · not launched | Mature purchase evidence favors A with viable CPA and purchase value. | Retain A’s message and validate before expanding the creative or budget. |
| Does broader morning relevance work better? | H3 attracted attention; historical purchase outcomes were weaker. | Test 1 · not launched | Comparable purchase evidence favors B. | Retain B and develop the next single-variable variation around it. |
| Is attention failing to convert? | H3 has higher CTR, but click totals, landing-page arrivals and journey diagnostics are unavailable. | Pre-launch verification | Validated technical, landing-page or message-continuity issues emerge. | Revise the affected journey, document changes and establish a new comparable reference. |
| Is the result too uncertain? | No new test delivery exists. | Test 1 · not launched | Low delivery, unequal exposure, late conversions or unstable purchase values. | Gather more evidence within the agreed budget; remain inconclusive if uncertainty cannot be resolved. |
| Can we add a stronger product benefit? | Heat-retention and leak-proof substantiation is unavailable. | Deferred | Approved substantiation and product imagery support a specific benefit. | Develop a separate benefit test; do not add it midway through Test 1. |
Related work: Cozy Cup Creative Analysis covers the broader historical findings. This study develops the next creative response. All new concepts remain unlaunched.