A Ridge wallet has a color, a texture, a personality. Put it into a conventional product catalog, however, and much of that personality can disappear. There is the object, sitting against a plain background, awaiting a click. According to Marpipe’s Ridge case study, the brand’s catalog ads were underperforming its static and video creative. The machinery could deliver the right product. The presentation gave the shopper little reason to care.
- The job: turn product feeds into branded catalog ads, images and videos.
- The buyer: ecommerce teams and agencies managing many products.
- The lesson: test a specific reason to buy before multiplying the design.
Marpipe addressed Ridge’s problem with Dynamic Color: a system that reads a product’s dominant color or pattern and matches its ad border accordingly. The company reports 53% higher return on ad spend than Ridge’s original catalog ads. A border sounds like a modest intervention. That is precisely what makes the example interesting. Advertising departments will debate a campaign’s grand idea for weeks; an ordinary product feed can quietly escape the conversation.
01 / The spreadsheet with selling ambitions
Catalog advertising starts with structured information: product identifiers, pictures, prices, descriptions and availability. Ad platforms use that feed to populate ads with relevant products. Retailers get scale without commissioning a separate advertisement for every item. They also inherit a temptation: once the feed works, leave it alone.
Marpipe sells control over what comes out of that system. Teams connect a catalog, build a branded design and populate it with product-specific information. Prices, offers and other attributes become parts of the creative. Feed management handles fields, categories and channel requirements. The output can support advertising on Meta, Google, TikTok, Pinterest and Snapchat. The ad platform still handles delivery; Marpipe works on the material being delivered.
That places it between ecommerce infrastructure and advertising creative. Its alternatives include native catalog tools, manual design work and platforms such as Hunch, Smartly, Celtra and Kargo. Marpipe’s distinctive proposition is the combination of catalog design and feed work in one place. A general design tool can make an attractive ad. Keeping thousands of versions consistent with changing product data is a different assignment.
02 / Three products, three missing arguments
Consider a mattress. Raymour & Flanigan tested a free next-day mattress delivery message in its dynamic product ads. Marpipe reports a 54% ROAS increase against the control, followed by testing and scaling designs across Meta, Snapchat and Pinterest. The creative improvement was an intelligible benefit. A mattress arriving tomorrow is a more useful proposition than a mattress merely existing.
DermStreet’s missing argument was price. Its Marpipe case study describes plain product images that failed to communicate the skincare retailer’s pricing advantage. Dynamic price and discount overlays put that information into the image across the catalog. The reported result was 30% higher ROAS and a 30% reduction in cost per acquisition. In this account, the first thing to fail was the explanation of the offer.
Ashley Canada asked a different question: would furniture sell better when shoppers could picture it in a home? Using Marpipe’s generative tools, it placed existing product images into lifestyle scenes and tested them against standard catalog creative on Meta. The published case reports 33% higher purchase ROAS and 19% lower cost per purchase. Its marketer, Leia Dufresne, described the practical appeal:
“The initial setup was fairly quick - roughly two minutes once I understood how the platform worked.”
Leia Dufresne / Ashley Canada
These are company-published outcomes. They make persuasive hypotheses, rather than promises. Ridge tested presentation, Raymour tested an offer, Ashley tested context. The copyable habit is to identify what information the shopper lacks, make that change visible and compare it with the existing version.
03 / Before catalogs, combinations
Founded in 2019 by Dan Pantelo, Marpipe originally made its public case around automating multivariate creative testing. It generated combinations of headlines, images and other elements, then helped marketers examine their performance. An $8 million Series A announced in September 2021 followed a $2 million seed round. Stage 2 Capital led the Series A, with investors including Commerce Ventures, Ripple Ventures and Samsung.
The underlying complaint was clerical as much as creative. Manual testing meant producing variants, naming assets, launching experiments and organizing results. Today’s catalog emphasis applies that same appetite for automation to a repeatable retail problem. Marpipe describes an intentionally small engineering and design team; its careers page emphasizes speed, candor and remote flexibility. Even its newsletter, The Catalog Cult, makes a specialist obsession sound like a social invitation.
04 / Motion has two meanings here
Product Level Video turns existing images into individual SKU videos through templates with motion, overlays and effects. Marpipe explicitly says this workflow need not use AI. Generative Catalogs is the separate AI offering for copy, backgrounds and product imagery. Its page still labels generative video “coming soon.” Buyers should distinguish templated animation from generated footage; both can move, but they require different judgments about accuracy.
That distinction matters when scale arrives. A misleading product detail reproduced across a catalog remains misleading. Feed quality matters too: stock status, descriptions and channel-specific fields are inputs to every attractive output. Marpipe’s September 2026 best-practices guide starts with the feed before moving to creative and campaign structure. A better frame cannot repair an unavailable product.
05 / The price of a better question
As of October 2026, Marpipe offers free exploration and a $199 monthly Startup plan with a 500-SKU cap, one channel, one market and three live designs. Enterprise displays $999 monthly, with SKU charges and optional additions affecting the bill. Managed feed service lists five cents per SKU with a $400 minimum. Media spend remains separate. The customer tests do not disclose their complete campaign costs.
The sensible starting point is therefore small: audit the feed, choose one missing benefit, keep a control and decide what success means. Teams with little catalog activity or too little test volume may have less to gain from mass variation. Marpipe’s most useful idea is quite ordinary: give the product a reason to be noticed, then let shoppers demonstrate whether the reason was any good.