Yes — the main opportunity is to make the transformation legible in 2 seconds. Right now the POS shows both modes, but the “why this is different” is still a bit too subtle; the shopper has to study it instead of instantly understanding “this stroller becomes a ride-on.”

• What feels weak in the current design:

  • The front has multiple floating images, so the eye doesn’t know what to read first.
  • The transform story is implied, not dramatized.
  • The small circular lifestyle photos add proof, but they compete with the product.
  • On the back, the transformation steps are there, but they’re too small to become the hero message.

• The clearest design move:

  • Make the front about one single idea: “Transforms from stroller to ride-on.”
  • Use the upper 60–70% of the front panel as one strong visual story.
  • Keep the lower front zone clean, since the actual product will occupy it.

• Best front layout direction:

  • Top: strong headline.
  • Upper-middle: large side-by-side or morphing visual of the two modes.
  • Between them: one bold directional transform cue.
  • Small support copy: “2 modes. 1 smart ride.” or “Stroll mode to ride-on mode in seconds.”
  • Bottom 30–40%: intentionally empty except maybe a very faint background shape.

• A stronger visual system than the current one:

  • Use a left-to-right story:
    • Left = stroller mode
    • Right = ride-on mode
    • Middle = bold arrow / motion path / “flip” cue
  • Or use one “hybrid” hero image:
    • stroller on one side
    • ride-on on the other
    • a seamless split or morph effect
  • This will communicate transformation faster than separate scattered images.

• Good headline options:

  • Stroll. Flip. Ride.
  • From Stroller to Ride-On
  • 2-in-1 That Grows With Them
  • One Ride, Two Modes
  • Transforms in Seconds
  • Stroller + Learning Bike in One

• Better copy than “World’s 1st patented stroller plus learning bike,” for retail readability:

  • A stroller that transforms into a ride-on
  • 2-in-1 stroller + ride-on
  • Transforms from parent-push to kid-ride
  • Grows from stroller mode to ride-on fun

• How I’d improve the front specifically:

  • Enlarge the product visuals significantly.
  • Remove at least one or two of the small round lifestyle images.
  • Replace the circles with one clean transformation diagram.
  • Use a bold cyan motion ribbon or arrow shape to connect the two states.
  • Add a very short benefit line:
    • Push mode for parents
    • Ride-on mode for kids
  • Keep the lower front area completely free of dense copy or key graphics.

• How I’d improve the back:

  • Turn it into a simple 3-step “How it transforms” panel.
  • Use numbered icons:
    • 1. Start in stroller mode
    • 2. Fold / adjust
    • 3. Ride-on mode
  • Add 2–3 benefit icons:
    • Grows with child
    • Easy transition
    • Compact / lightweight / fun
  • Include a QR code to a 10-second demo video.
  • Make the product sequence much larger than it is now.

• More innovative POS concepts:

  • Concept 1: “Split Identity”
    • Front panel split vertically.
    • Left half visually styled for stroller mode.
    • Right half visually styled for ride-on mode.
    • Product hero overlaps both halves.
    • Clean, premium, immediate.
  • Concept 2: “Motion Path”
    • A bold curved path sweeps across the upper panel showing the transformation journey.
    • The path can continue subtly onto the base platform.
    • Makes the POS feel dynamic.
  • Concept 3: “Before / After Hero”
    • Huge “before” and “after” silhouettes.
    • Very minimal text.
    • Strongest for quick retail communication.
  • Concept 4: “Parent to Child”
    • Show parent-push mode transitioning to child independence.
    • Emotionally stronger if your goal is “growth” and “development.”
  • Concept 5: “Interactive Retail”
    • Add a flip-panel, lenticular image, or pull-tab that changes the product from stroller to ride-on.
    • This is the most innovative if budget allows.

• Smart base-platform ideas:

  • Since the stroller sits on it, use the base as a support graphic, not a message-heavy area.
  • Good options:
    • subtle wheel tracks / motion line
    • a faint directional arrow leading to the front panel
    • a clean “2-in-1” badge
    • a parked silhouette outline showing intended product placement
  • Avoid too much type on the base, since the product will visually dominate it.

• Color and hierarchy suggestions:

  • Keep white space, but use it more intentionally.
  • Use one accent color for the transformation cue only.
  • Make stroller mode and ride-on mode slightly color-coded.
  • Use fewer background circles; they currently feel decorative rather than communicative.
  • Let the transformation graphic do the work.

• If you want Nano Banana to generate better concepts, your prompt should be very constraint-based.

  • Don’t just ask for “more innovative.”
  • Tell it:
    • the retail objective
    • the exact physical format
    • what must stay blank
    • what visual hierarchy matters most
    • the tone: premium, modern, family, clever, not cluttered

• Prompt you can use for Nano Banana:

  • Design a retail POS graphic for a stroller brand called Globber. The product transforms from stroller mode into a ride-on mode. The main objective is to communicate the transformation instantly from 6–10 feet away. Use a premium, modern, minimal visual style with strong hierarchy and clean white space. Keep the lower 35–40% of the front panel mostly empty because the real product will cover that area. The front panel should focus on one bold transformation story using large product visuals, a clear directional cue, and very short headline copy. Avoid clutter, too many small photos, or scattered elements. Show both modes clearly and make the “before to after” transformation the hero. Use Globber’s cyan accent color sparingly for arrows, motion graphics, or key labels. The back panel should show a simple 3-step transformation sequence and 2–3 benefit icons. Create 3 different design directions: 1) premium split-screen concept, 2) dynamic motion-path concept, 3) minimal before-and-after concept.

• Even better prompt if you want stronger output:

  • Create 3 innovative POS concepts for a stroller display. The stroller sits on a base platform in front of a vertical backboard. The product transforms from a stroller into a ride-on. The design must make this transformability obvious in under 2 seconds to shoppers walking by. Front panel constraints: keep the bottom section clear of important graphics because the product blocks that area; use only the upper area for messaging. Prioritize one large hero visual over many small images. Show stroller mode and ride-on mode with a bold visual transition. Style should feel premium, clean, clever, and retail-ready, not childish or busy. Back panel should explain how it works with a clean 3-step sequence, lifestyle reinforcement, and a QR code area. Include short headline ideas and placement recommendations.

• If you want Nano Banana to stay close to your current format:

  • Redesign this existing POS format for better communication of transformability. Keep the same general structure: vertical back panel plus rectangular base platform. Improve the front by simplifying the composition, enlarging the product imagery, reducing small lifestyle circles, and making the stroller-to-ride-on transformation the main story. Keep the bottom front area free of key design elements because the product covers it. Make it feel more innovative and retail impactful while staying clean and premium.

• My strongest recommendation:

  • Front = one big transformation visual, one headline, one support line.
  • Back = three steps + benefits + QR.
  • Remove decorative clutter and make the “change from one mode to another” dramatically obvious.

• A very practical front-wireframe idea:

  • Top:
    • GLOBBER logo
    • “Stroll. Flip. Ride.”
  • Middle upper:
    • large stroller visual on left
    • bold cyan arrow / morph graphic
    • large ride-on visual on right
  • Small line below:
    • “2-in-1 stroller + ride-on”
  • Lower area:
    • mostly blank for product overlap
    • maybe only a faint oversized brand graphic in the background

If you want, I can do either of these next:

• write 5 polished Nano Banana prompts in different creative directions, or
• sketch a recommended front/back layout in words so your designer can build it immediately.

She’s partly right, but not fully. Raw AI images often do fall apart when enlarged for print, especially on faces, hands, wheels, fabric, and product edges, but that does not mean AI can’t be used for print-ready POS. It means the workflow has to be set up correctly.

• Where she is right:

  • Many AI-generated images look good on screen but break at print size.
  • Common failures are:
    • waxy or overly smooth skin
    • fake-looking hands
    • mushy textures
    • distorted product geometry
    • soft edges after scaling
  • Circular “bubble” crops can hide those flaws, which is probably why she is keeping them small.

• Where she is not right:

  • Other brands are absolutely using AI in print.
  • They usually are not just taking one AI image and dropping it into a full-size print file.
  • They use a production workflow:
    • generate
    • upscale
    • retouch
    • composite with real photography
    • sharpen selectively
    • color-correct for print

• The key misunderstanding is usually resolution requirements.

  • For print, not everything needs 300 dpi at full-size.
  • For retail POS viewed from a few feet away:
    • 150 ppi at final size is often very good
    • 100 ppi is often acceptable for larger-format displays
    • sometimes even lower works if the viewing distance is longer
  • 300 ppi is more important for close-held print like brochures, not always for a floor POS panel.

• So if she is saying:

  • “AI cannot be printed large” → not correct
  • “AI cannot be enlarged carelessly without artifacts” → correct

• The real issue is probably one of these:

  • She’s using a tool that outputs too small.
  • She’s enlarging with a weak upscaling method.
  • She’s relying on fully AI-generated product/lifestyle shots instead of hybrid compositing.
  • She’s judging quality at 100% zoom on screen instead of at final viewing distance.
  • She doesn’t have a retouch workflow after generation.

• Best-practice workflow for print POS:

  • Use real product photography for the stroller itself.
  • Use AI for:
    • background scene
    • mood/lifestyle environment
    • supporting secondary imagery
    • concept comps
  • If people are shown, retouch faces/hands manually.
  • Upscale using a dedicated tool, not just standard resize.
  • Add selective grain/texture back in so it doesn’t look airbrushed or fake.
  • Output test crops and print them at actual size before approving.

• If you want high-res AI lifestyle imagery that feels more real:

  • Start with a wider composition so you are not over-cropping.
  • Generate multiple variations and choose the most natural anatomy and textures.
  • Use image-to-image or inpainting to fix weak areas.
  • Upscale in stages rather than one giant jump.
  • Retouch details like eyes, hands, wheels, seams, shadows.
  • Add realistic print sharpening at the end, not before.

• Important distinction:

  • AI is weakest when it must render:
    • your exact stroller product accurately
    • branded hardware details
    • exact mechanics of transformation
  • AI is stronger when it renders:
    • a family scene
    • outdoor setting
    • mood image
    • background context
  • For your POS, the hero stroller should probably be real/product-render accurate, not pure AI.

• For your current design specifically:

  • Those circular photos are probably acting as “safe zones” because small crops make flaws less noticeable.
  • If you want larger lifestyle visuals, you likely need:
    • better source generation
    • proper upscaling
    • manual retouch
    • or real photography instead of fully generated images

• A practical quality rule:

  • Ask your lead to test a 10" x 10" or 12" x 12" section at final print size.
  • Print:
    • the original
    • the upscaled version
    • a retouched version
  • Then judge from intended retail distance, not from 6 inches away on a monitor.

• What I would recommend strategically:

  • Don’t make the success of the POS depend on giant AI lifestyle photos.
  • Make the hero communication graphic-based:
    • large real product image
    • strong transform arrow/motion system
    • short headline
  • Then use one or two lifestyle images only as support.
  • That reduces risk and usually looks more premium.

• A good middle-ground position to take with your design lead:

  • “You’re right that low-res AI gets fake fast in print. But we don’t need to avoid AI entirely — we need a better production workflow and we should use AI where it’s strongest.”

• If you want a simple decision rule:

  • Use real photography or accurate render for:
    • hero product
    • transformation steps
    • close-up details
  • Use AI for:
    • supporting lifestyle moments
    • background scenes
    • mood-building imagery
  • Use vector/graphic design for:
    • arrows
    • transitions
    • structure
    • premium visual hierarchy

My short answer: she is right about the risk, but wrong if she’s treating that risk as a hard limit. High-quality print use of AI is possible; it just usually requires hybrid production, not one-click generation.

If you want, I can help you with either:

• a print-ready AI image workflow you can hand to your design lead, or
• exact specs for what image resolution you actually need for your POS dimensions.

She probably does not need 300 ppi for the entire photographic area, even for close-view POS. For a shopper-facing display that people may approach closely, 300 ppi is a good ideal for small text and critical product detail, but for large-format lifestyle photography on a POS panel, 150–200 ppi at final size is often the practical target.

• The first important distinction:

  • What she means is probably 300 ppi, not 300 dpi.
  • In design files, you’re usually talking about image resolution in pixels per inch.
  • The printer has its own dpi behavior, but that is not the same thing as the placed image needing to be 300 everywhere.

• For your type of POS, a better rule is:

  • 300 ppi: small text, logos, icons, fine line art, barcodes, QR codes
  • 200–250 ppi: hero product photography if the shopper will inspect it closely
  • 150–200 ppi: large lifestyle imagery on a rigid POS panel
  • Below 150 ppi: can work for background-only areas, but riskier for close-view use

• So if the shopper can walk right up to it:

  • I would not be casual and say “100 ppi is fine.”
  • But I also would not insist the whole thing must be 300 ppi.
  • A smart target is:
    • keep all vector elements as vector
    • keep the real stroller/product image as high-res as possible
    • aim for at least 150–200 ppi at final size for the AI-generated lifestyle portions
    • use 200+ ppi if a face or child is large and meant to be scrutinized

• Why 300 ppi is often over-applied:

  • It comes from brochure/magazine standards.
  • POS and large-format print are different.
  • Once the panel gets larger, the required ppi often drops because the image is not read like a handheld piece.
  • Even when people stand close, they usually inspect the product first, not pore over the lifestyle background at nose distance.

• In your workflow, this matters a lot:

  • Since you start with a real product photo and generate the environment around it, you do not need every part of the final image to have equal sharpness.
  • In fact, it can look more realistic if:
    • the stroller remains crisp
    • the background has natural depth falloff
    • less-critical areas are slightly softer
  • Real photography is not uniformly razor-sharp edge to edge either.

• A more realistic quality standard for your team:

  • “Does it look convincing at intended viewing distance?”
  • Not:
    • “Does it hold up at 200% zoom on a Retina screen?”

• Practical example:

  • If the visible photo area is 24" wide x 36" high:
    • at 300 ppi you’d need 7200 x 10800 px
    • at 200 ppi you’d need 4800 x 7200 px
    • at 150 ppi you’d need 3600 x 5400 px
  • 200 ppi is much more achievable while still being strong for retail.

• Where she may still be right:

  • If the faces are large
  • If the child’s hands are prominent
  • If the AI-generated area includes fine stroller-adjacent edges or fabric transitions
  • If the print substrate is very smooth and high-quality, which can reveal flaws more clearly
  • In those cases, weak AI will absolutely start to look fake when enlarged

• But that is a content-quality problem, not only a ppi problem.

  • A bad AI face at 300 ppi still looks bad.
  • A well-composited, well-retouched image at 180–220 ppi can look very good.

• What I’d recommend as your internal standard:

  • Product: real photo, highest res available
  • Text/logo/arrows/icons: vector only
  • Lifestyle/support image:
    • 150–200 ppi minimum at final placed size
    • 200–250 ppi if faces are large or in the shopper’s main eyeline
  • Then test print before final approval

• Best proof method:

  • Print a 1:1 crop of the most critical area, not a scaled-down letter-size proof.
  • Test the actual problem zones:
    • child’s face
    • hands
    • stroller edge where the AI environment meets the real product photo
    • any fabric texture
  • Pin that crop up and view it from:
    • 12 inches
    • 24 inches
    • 36 inches
  • That will settle the argument quickly.

• If you want a strong compromise to propose to your lead:

  • “Let’s stop debating 300 ppi as a blanket rule. Let’s define critical zones and non-critical zones.”
  • For example:
    • critical zones = product, faces, top headline area
    • non-critical zones = soft background, environmental context, edge fade areas

• For your POS specifically, I’d design around this reality:

  • Don’t depend on one huge AI lifestyle image as the hero.
  • Let the hero be:
    • the real stroller/product image
    • a big transform visual
    • short copy
  • Then use lifestyle imagery as a secondary proof point, where a little softness is less damaging.

• One more useful production note:

  • If the image still feels “too smooth and fake,” the fix is not always more resolution.
  • Often the fix is:
    • restore micro-contrast
    • add natural texture/grain
    • fix edge halos
    • avoid over-smoothing skin
    • keep realistic shadow detail
  • Over-retouching is often what makes AI composites look synthetic.

My direct answer: for close-view POS, 300 ppi everywhere is usually stricter than necessary. A better print standard is 150–200 ppi for large photographic areas, 200+ ppi for important faces/product-adjacent detail, and vector for all text and graphics.

If you want, I can do one of two useful next steps:

• calculate the exact minimum pixel dimensions you need once you give me the front-panel size, or
• help you build a simple “critical zone resolution spec” you can hand to your design lead.

Don’t respond by arguing about whether it “looks bad.” Respond by changing the decision standard. The real issue is not fuzziness at 300% zoom; it’s that she is rejecting options before testing them in the actual use case.

• Based on what you showed:

  • At 300%, yes, it looks softer.
  • That is not surprising.
  • It does not look obviously unusable to me.
  • A 300% zoom is not the right approval standard for shopper-facing POS.

• What you want to say, in substance:

  • I appreciate the quality concern.
  • But we cannot kill concepts based on enlarged screen inspection alone.
  • We need to review them at final placed size and, if needed, with a print test.
  • I need options brought forward, not pre-rejected.

• A good response you can send:

  • “I understand the concern about softness, and I agree we should protect print quality. But 300% zoom is not the standard we should use to reject an image for POS. Please don’t eliminate concepts before showing them in context. I’d like us to evaluate this based on final placed size, intended viewing distance, and a 1:1 test crop print of the most sensitive area. If it fails that test, we move on. If it passes, we use it. Going forward, please bring AI options forward with risks noted rather than ruling them out before review.”

• If you want it a little firmer:

  • “I need us to stop using extreme zoom as the decision-maker. For this POS, the approval standard is final-use performance, not whether a face gets a little soft at 300% on screen. Please show me the option in layout and, if needed, with a print test before deciding it’s unusable.”

• If the pattern keeps repeating, set a process:

  • No AI route gets rejected without:
    • final placed-size mockup
    • 1:1 crop of the most critical area
    • print proof if it’s close
  • Review standard:
    • actual size in layout
    • 18–36 inch viewing distance
    • only critical zones judged tightly
  • Decision rule:
    • if acceptable in print/context, it stays in consideration
    • if not, it gets replaced

• This is the key management point:

  • She is treating “possible imperfection” as “automatic no.”
  • What you need from her is:
    • informed experimentation
    • evidence-based recommendation
    • not unilateral filtering

• I would also say this once, clearly:

  • “Your job here is not just to protect against defects; it’s to help us find workable AI solutions. I need you to test and present, not default to no.”

• If you want to reduce defensiveness, avoid:

  • “You’re being resistant.”
  • “This is overkill.”
  • “Other people can do it.”
  • Those are true-feeling statements, but they usually make people dig in harder.

• Better framing:

  • “Let’s use a shared approval method.”
  • “Let’s test before rejecting.”
  • “Let’s distinguish critical quality issues from zoom-level softness.”

• A very practical next message could be:

  • “Please place this image at final size in the POS layout and send me:
    • the full layout view,
    • a 100% crop at final placed size,
    • and a 1:1 print crop of the child’s face. Then we’ll decide based on real use, not 300% zoom.”

• If you need one sentence that is both calm and firm:

  • “I’m aligned on quality, but I’m not aligned on rejecting AI imagery based on 300% screen zoom without a final-size layout review and print test.”

Honestly, the strongest move here is to stop debating taste and create a rule. Once the rule is “test in context before rejecting,” her resistance has less room to control the outcome.

If you want, I can draft:

• a polished Slack message,
• a firmer email,
• or a short AI-review SOP you can put in place for your team.

Here’s a nicer, more collaborative version that still keeps the point clear:

As we’ve discussed over the past several months, there’s a growing expectation — including from Pierre — that we build real capability around printable AI. I know we’re still refining that process, but I don’t think we should rule options out based on extreme zoom views alone, especially when other brands are clearly finding ways to make this work.

For this POS, I’d like us to evaluate imagery based on how it performs in its actual end use: in layout, at final size, and if needed, with a print test. A face looking a little soft at 300% on screen shouldn’t automatically disqualify an option. Please show me the image in context before deciding it isn’t usable.

If you want, I can also give you:

• a softer version
• a firmer manager version
• or a super-short Slack version.

Absolutely — here’s a version that adds that point but keeps it constructive and professional:

As we’ve discussed over the past several months, there’s a growing expectation — including from Pierre — that we build real capability around printable AI. I know we’re still refining that process, but I don’t think we should rule options out based on extreme zoom views alone, especially when other brands are clearly finding ways to make this work.

For this POS, I’d like us to evaluate imagery based on how it performs in its actual end use: in layout, at final size, and, if needed, with a print test. A face looking a little soft at 300% on screen shouldn’t automatically disqualify an option. Please show me the image in context before deciding it isn’t usable.

I’ll be honest: it’s starting to feel frustrating that we’ve had this conversation a number of times, but the approach to AI still seems very cautious and quick to default to “not usable.” I do understand that you’re trying different things and that the timelines are tight. But at some point, we need to move beyond defaulting to no and instead figure out how to make it work when the quality is close. I really need you to lean into AI more fully and help lead that adoption for the team.

If you want, I can also tighten this into:

• a more diplomatic version
• a firmer manager-to-direct-report version
• or a shorter Slack message.

Yes — the best prompt is one that tells Nano Banana exactly what problem to solve: make the transformability obvious fast, keep the front bottom clear, and create a premium retail-ready concept rather than a generic ad.

• Here’s a strong master prompt you can use:

Create a premium retail POS design concept for Globber’s stroller-to-ride-on product. The display consists of a vertical backboard and a base platform. The key message is that the product transforms from stroller mode into ride-on mode, and that transformation must be obvious to shoppers in under 2 seconds from several feet away.

Design the front of the POS so that the upper portion clearly communicates the 2-in-1 transformation using large product visuals, a bold directional cue, and minimal copy. Keep the bottom 35–40% of the front panel mostly free of important graphics or text because the real product will physically cover that area. The design should feel premium, modern, minimal, and retail-ready — not cluttered, childish, or overly decorative.

Use one strong hero transformation story instead of many small images. Show stroller mode and ride-on mode clearly, ideally with a left-to-right transition, split-screen concept, or dynamic motion path. Use clean white space and restrained use of Globber’s cyan accent color for arrows, labels, or motion graphics. Avoid small floating circles, busy layouts, or too many separate callouts.

For the back of the POS, create a simple 3-step transformation explanation with clean visual sequencing and 2–3 supporting benefit callouts. Include space for a QR code to a demo video.

Create 3 distinct creative directions:

1. Premium split-screen transformation concept
2. Dynamic motion-path concept
3. Minimal before-and-after concept

Make the output look like a real retail display design presentation, with polished composition, strong hierarchy, and believable print-ready graphics.

• If you want a version that stays closer to your current format, use this:

Redesign this existing Globber POS format to make the stroller-to-ride-on transformation more innovative and immediately understandable. Keep the overall structure of a vertical back panel with a rectangular base platform. Improve the front by simplifying the composition, enlarging the product imagery, reducing small lifestyle elements, and making the transformation the main visual story. Keep the lower portion of the front panel clear because the physical product blocks that area. The style should feel clean, premium, modern, and bold, with strong retail visibility from a distance. On the back panel, show a simple and elegant 3-step transformation sequence with minimal copy and clear benefit communication.

• If you want Nano Banana to generate more visually ambitious concepts, use this:

Design an innovative retail POS for a transformable stroller that becomes a ride-on. The display should feel surprising, smart, and highly legible in-store. Focus on one big transformation idea, not multiple small images. Show the two modes in a dramatic and elegant way using either a morph visual, split identity, or a bold motion graphic. Keep the bottom of the front panel intentionally open because the actual product covers that area. Use premium typography, modern white space, subtle brand graphics, and a restrained cyan accent. The design should instantly communicate: stroller mode to ride-on mode. Make it feel like a high-end juvenile product display in a major retailer.

• If you also want it to generate supporting lifestyle imagery around the real product photo, add this block:

Use the actual product photo as the hero and build a realistic lifestyle environment around it. Keep the stroller geometry accurate and preserve the real product details. The surrounding scene should feel natural, premium, and believable for print. Avoid overly smoothed skin, fake anatomy, distorted hands, or artificial textures. Prioritize realistic lighting, natural depth of field, and clean retouchable edges.

• If you want headline ideas included in the output, add:

Include short retail headline options such as:

• Stroll. Flip. Ride.

• From Stroller to Ride-On

• 2-in-1 for Growing Kids

• One Ride, Two Modes

• Transforms in Seconds

• Best extra instruction to improve results:

  • Ask for concept boards, not just one final image.
  • Example:
    • Show 3 different front-panel concepts and 2 back-panel concepts.
    • Include layout hierarchy, headline placement, product placement, and motion-graphic recommendations.

• If you want one short version you can paste quickly, use this:

Create 3 innovative retail POS concepts for Globber’s stroller-to-ride-on product. The display has a vertical backboard and a base platform. The front must instantly communicate that the product transforms from stroller mode to ride-on mode. Use a premium, minimal, retail-ready style with strong hierarchy and clean white space. Keep the bottom 35–40% of the front panel mostly empty because the real product covers that area. Make the upper panel the hero area with large product visuals, a bold transformation cue, and very short copy. Avoid clutter, scattered images, or too many small lifestyle bubbles. The back should show a clean 3-step transformation sequence plus 2–3 benefits and a QR code area. Create 3 distinct directions: split-screen, motion-path, and minimal before/after.

• One smart tip:
  • If Nano Banana tends to get too generic, add:
    • “Do not design this like a poster ad. Design it like an in-store structural POS graphic system with clear hierarchy and product obstruction considered.”

If you want, I can also give you:

• a prompt specifically for front only,
• a prompt specifically for the back panel,
• or a prompt specifically for generating larger, more believable lifestyle images around your real product photo.

AI generated image

AI generated image

Here’s a more casual version that still gets the point across clearly:

For printable AI, I just want to reiterate something we’ve talked about a lot over the past few months: there’s a growing expectation — including from Pierre — that we get stronger at using AI for print. I know we’re still figuring out the best workflow, but I don’t think we should rule things out based only on extreme zoom views, especially when other brands are clearly finding ways to make it work.

For this POS, I’d really like us to judge the imagery based on how it works in the real use case: in layout, at final size, and, if needed, with a print test. If a face gets a little soft at 300% on screen, that alone shouldn’t automatically make it unusable. I’d just like to see it in context before we decide it won’t work.

I’ll be honest — it’s getting a little frustrating that we’ve talked about this a number of times, but the mindset around AI still feels pretty cautious and quick to go to “not usable.” I do get that you’re trying different things and that timing is tight. But at some point, we can’t keep defaulting to no when something is close — we need to figure out how to make it work. I really need you to lean into AI more and help lead that adoption for the team.

And on the brainstorming side, I think we really need to be using AI more to help with concepting and ideation too. I’m happy to do more of that upfront as well, so we can go in with a clearer brief and stronger direction instead of you having to design from scratch or too blindly.

If you want, I can also make this:

• softer and more collaborative
• shorter for Slack
• or a little firmer without sounding harsh

Here’s a shorter, softer Teams version that still makes the point:

For printable AI, I just want to reiterate that there’s a growing expectation — including from Pierre — that we build stronger capability here. I know we’re still figuring out the right workflow, but I don’t think we should rule options out based only on extreme zoom views, especially when other brands are making this work.

For this POS, I’d like us to judge the imagery based on how it performs in the real use case: in layout, at final size, and, if needed, with a print test. A face looking a little soft at 300% on screen shouldn’t automatically make it unusable — I’d like to see it in context first.

I know you’re being thoughtful about quality, and I appreciate that. At the same time, I’d really like us to lean in more on AI and focus on finding workable solutions when something is close, rather than defaulting to no too quickly. I need your help leading that for the team.

For brainstorming too, I think we should be using AI more to help with concepting and ideation. I’m happy to help more upfront so we have a clearer brief and direction to work from.

If you want, I can also make this:

• even shorter and more casual
• warmer/more encouraging
• or more manager-like but still gentle

Here’s a revised Teams version with that added in, while keeping it gentle:

For printable AI, I just want to reiterate that there’s a growing expectation — including from Pierre — that we build stronger capability here. I know we’re still figuring out the right workflow, but I don’t think we should rule options out based only on extreme zoom views, especially when other brands are making this work.

For this POS, I’d like us to judge the imagery based on how it performs in the real use case: in layout, at final size, and, if needed, with a print test. A face looking a little soft at 300% on screen shouldn’t automatically make it unusable — I’d like to see it in context first.

I know we’ve talked about this before, but it’s becoming more urgent now and more expected of us. I really appreciate that you care about quality, and that’s important. At the same time, I’d like us to lean in more on AI and focus on finding workable solutions when something is close, rather than defaulting to no too quickly. I need your help leading that for the team.

For brainstorming too, I think we should be using AI more to help with concepting and ideation. I’m happy to help more upfront so we have a clearer brief and direction to work from.

If you want, I can also make paragraph 3:

• a little firmer
• a little warmer
• or even shorter for Teams.

Sure, here’s a shorter, more casual version:

I’m just using AI to brainstorm based on Stoney’s feedback, but there are really two separate things here: using AI to create printable images, and using AI to generate ideas. We need to be able to do both.

On the printable side, I just want to reiterate that there’s a growing expectation, especially from Pierre, that we get stronger here. I know we’re still working out the right process, but I don’t think we should rule things out just based on extreme zoom views.

If a face looks a little soft at 300% on screen, that doesn’t automatically make it unusable. For this POS, we want the images as large as possible, like we discussed, so I’d like us to review them in context first, in layout and at final size, before we decide no.

I know we’ve talked about this before, but it’s becoming more urgent now and more expected of us. I really appreciate that you care about quality, but I need us to lean in more on AI and focus on finding workable solutions when something is close, instead of defaulting to no too quickly. I need your help leading that for the team.

And for brainstorming too, I think we should be using AI more for concepting and ideation. I’m happy to help more upfront so we have a clearer brief and direction to work from.

If you want, I can make this even shorter and more Teams-y.

Here’s a cleaner, more natural version that keeps your tone:

Well, I’m just using AI to brainstorm based on Stoney’s feedback, but there are really two separate things here: using AI to create printable images, and using AI to generate ideas. We need to be able to do both.

On the printable side, I just want to reiterate what we’ve talked about a few times now: there’s a growing expectation, especially from Pierre, that we get stronger here. I know we’re still figuring out the right process, but I don’t think we should rule things out based only on extreme zoom views. If a face looks a little soft at 300% on screen, that shouldn’t automatically make it unusable, if you know what I mean.

I really appreciate that you care about quality, but as we’ve discussed, I need us to loosen up a bit on the hesitation and not default to no too quickly. I need your help leading that for the team. Let’s make sure we’re always reviewing images in context first, in layout and at final size, before deciding something won’t work.

For this POS, we wanted the images to be as large and impactful as possible, so I do understand Stoney’s feedback, to be honest.

And on the brainstorming side too, I think we should be using AI more for concepting and ideation. I’m happy to help more upfront so we have a clearer brief and direction to work from.

If you want, I can also make this:

• a tiny bit softer
• a bit more manager-like
• or even shorter for Teams chat.

Here’s a cleaner prompt you can paste into Nano Banana:

Redesign the existing Globber POS format shown in the reference images. Keep the same overall structure: one vertical back panel and one rectangular base platform.

The goal is to make the stroller-to-ride-on transformation feel innovative, premium, and instantly understandable from a distance. The design should communicate the transformation in a very simple, visual way, with strong retail impact and minimal clutter.

Use the attached lifestyle images for the stroller mode and rider mode. The front panel should use large, premium lifestyle imagery rather than small circular cutouts or isolated product shots on white. Make the transformation the hero story by clearly showing stroller mode transitioning into ride-on mode. Use a bold but elegant visual cue such as a directional arrow, split composition, or motion path to connect the two modes.

Keep the lower half of the front panel mostly clean because the physical product will block that area in-store. Only allow a very light background graphic or subtle brand element in that lower area. Concentrate the main communication in the upper half of the front panel.

The design style should feel clean, modern, premium, bold, and retail-ready. It should be easy to understand in 2 seconds from several feet away. Use strong hierarchy, restrained typography, and simple benefit messaging. Avoid clutter, too many callouts, busy layouts, small bubbles, or a white-background catalog look.

For the back panel, create a simple and elegant 2-step transformation story with minimal copy. Show how the product changes from stroller to ride-on in a very clear visual sequence. Include concise benefit communication such as 2-in-1, transforms in seconds, or grows with your child. Keep the back clean and premium.

Create one presentation image that shows both the front and rear panel designs of the POS structure in the same view or as a polished retail concept board. Make it look like a realistic, professional in-store display proposal.

If you want, add this at the end to get better output:

Use the Globber brand look and colors from the reference images, especially the clean white base and cyan accent. Preserve the product’s real proportions and geometry. Make the design feel like a real printed retail POS, not a digital ad poster.

You can also use this slightly firmer version if Nano Banana tends to get too loose:

Important constraints:

• Keep the POS structure exactly as shown in the reference
• Front panel upper half = main communication zone
• Front panel lower half = mostly empty because product blocks it
• Use attached lifestyle images, enlarged and integrated cleanly
• Make stroller-to-ride-on transformation the main visual story
• No small circular lifestyle bubbles
• No busy collage layout
• No childish or overly playful design
• Output should show both front and back panel concepts

If you want, I can also turn this into:

• a shorter prompt optimized for image models
• a more art-directed prompt with exact headline suggestions
• or 3 separate prompts for 3 different design directions.

AI generated image

AI generated image

AI generated image