Dan Pantelo

The through-line

Dan Pantelo's consistent argument is that the boring, mechanical parts of growth outperform the parts marketers enjoy talking about. In advertising, that means catalog ads, a format built on spreadsheets rather than storyboards, which he says gets "sidelined" in creative conversations because "a lot of people talk about image creative, you know, UGC video creative because it's fun to talk about" [1]. In go-to-market, it means picking up the phone, showing up at events week after week, and treating distribution as the scarce asset rather than the content itself [2]. The unifying claim is that a misunderstood, unsexy channel is usually where the arbitrage lives, and that the numbers, once you pull them, tend to embarrass the conventional wisdom.

The second recurring preoccupation is leverage through positioning: whether a company should sit at the top of the market or the bottom, whether it should own its audience or rent it, and which direction of movement is actually available to it later [2].

On why catalog ads stopped being an afterthought

Pantelo's baseline definition is deliberately deflationary. DPA and catalog ads are the same thing, no difference [1], and the mechanic is unglamorous: "you're not uploading a creative. You're actually like uploading essentially a spreadsheet of all of your SKs" [1]. Every business already has a feed management tool whether it knows it or not, and if it isn't paying for one it is "just probably on Shopify's default built-in like baby one" [1]. The platform reads the SKU data, understands it has "a brown leather shoe here uh between the 70 to $80 price range," and serves it to people it believes are shopping for exactly that [1]. His summary: "It's not sexy, but, uh, it it works really well" [1].

What has changed is scale. Two years ago catalog ads were a retargeting line item that never took up much real estate in an ad account, with a decent click-through rate because it was signal-based, but no path to scale [1]. Now, across the 650-plus ad accounts Marpipe is connected to, representing roughly 3% of global ad spend on Meta, 24% of spend is catalog ads, a number he says usually surprises people [1]. He frames the format through the most optimised page on the internet: the Amazon product page, where the "you may also like" module sits directly beneath the product, and argues catalog ads are "essentially like the ad version of that placement" because the platforms already know what you are shopping for [1]. Marpipe itself powers DPA creative across Meta, Snap, Pinterest and programmatic for retailers including Champion, Kate Spade, Bloomingdales and Paramount, alongside DTC brands such as True Classic, Open Store, Ridge and Simple Modern [1].

On enriched creative versus the white background

The default output is his central complaint. A stock catalog ad is "just a product image on a white background. It's the same image that appears as image number one in the PDP. That's not good ad creative" [1]. He describes the waste geographically: the product occupies maybe 20% of the frame and the rest is "oceanfront real estate around the image that's just white" [1]. Anything a brand would put into an image ad, testimonials, value props, colour, belongs in its catalog ads too [1]. Turn that same feed into seven design templates and the ads stop looking like catalog ads while still getting "that like express lane on the highway when it comes to those ads being served" [1].

He backs this with performance data he says was pulled the morning of the conversation. Average click-through rate across all ads is 2.1%; non-catalog ads sit at 2.01%; catalog ads at 2.3%; enriched catalog ads, meaning ones with creative variables introduced, at 2.4% [1]. On ROAS the gap widens sharply: 5.2 across all ads, 7.25 for catalog ads, 8.18 for enriched catalog ads [1]. He was candid that the size of the gap surprised him: "I was just surprised to see this like the scale of the difference this morning when we pulled the numbers" [1]. The pitch he uses with brands is not a deck but the enriched catalog library on Marpipe's own site, because the before-and-after does the arguing [1].

On the feed mistakes brands keep making

His sharpest tactical point concerns how learning accumulates. In Meta's commerce manager, when you run ads from a source feed, "learnings accumulate at that feed level," functioning much like a pixel [1]. The common error follows directly: a brand launches a fresh catalog containing only, say, the men's items on sale, which means "you have to retrain a new catalog," and in his experience "it'll never work" [1]. He calls uploading new source feeds "such a mistake" and notes that most people simply do not know the learnings live there [1].

On going down-market without giving up the enterprise business

Pantelo is openly sceptical of low-cost SaaS as an end state. Told that Judgeme is a wild success, his response was flat: "It's not successful," and on the revenue, "It's a $12 million ARR business. That's That's nothing. That's tiny" [2]. His objection is not the price point but the ambition: "My my kurfuffle with Judge Me is that from everything I can tell they are like extremely happy to provide a lowcost product with maximum penetration and to just like let that thing run and like just draw distributions basically" [2]. Pressed on the strategic logic, he conceded the argument entirely and said so: "I take everything back if what they do is reinvest the profit into building new products," because a bottom-up attack on Yachtpo would make the upside massive and "of course it's a brilliant strategy" [2]. He also credits the product on its merits, having been a Judge Me customer when he ran an e-commerce store, and notes the pattern where higher-priced tools carry fewer integrations, so choosing one expensive tool commits you to buying ten more [2].

The reason he was interrogating it so closely is that Marpipe is running the same play, building a lower-cost, self-serve style version of the product alongside the enterprise offering, and he was explicitly asking what it feels like operationally, whether it behaves like a separate business, and how retention and growth compare [2]. Related pricing experiments include requiring a credit card upfront for free trials, and startup-level plans on the DPA product [1][2].

On why email outbound plateaus and the phone does not

Pantelo built the early business on cold email, saying it produced Marpipe's first 100 customers about four years ago, and it still runs [2]. But he is precise about its ceiling: "we're running cold email outbound. It's booking a set number of meetings every week, but when we try to push it further, you just can't like it's like a retargeting campaign, you know, like it works great, but you can't scale it" [2]. Concretely that is five to ten meetings a week from one person allocating maybe 20% of their effort [2]. His explanation is about attention economics rather than copy: "That top of inbox real estate is so fragile. But if you call someone, they pick up, you've got them. If you send them an email and they don't check it within, I don't know, an hour, two hours, it's on the second page of their inbox. They're not going back to the second page" [2]. He reports being genuinely surprised that cold calling is working as well as it currently is, given that marketers are picking up the phone, and Marpipe is scaling it out [2]. He remains restlessly experimental about the channel, resolving on the spot to have his email operator test a lowercase, phone-number-in-the-body follow-up format he had just heard described [2].

On events and social content as one system

The go-to-market insight he has most conviction in is that events and social content compound into an impression of omnipresence. He has seen this from the inside as a partner in an events business in the space called ecom founders, and says building that up over the past year or two showed him firsthand "how crazy effective events actually are" [2]. Marpipe is now constructing its own version of that engine: how it selects events, how it shows up at all of them [2]. He treats the pieces as interlocking rather than parallel, describing events as part of a broader go-to-market play alongside a sales team and everything else working to get people across the finish line [2].

On owning content versus partnering for distribution

On the content side, Marpipe is hiring a full-time in-house creator, and Pantelo is specific about the bar, wanting someone who makes content like Orin John or Ashwin, while admitting such people are "Very hard to find" [2]. But he is clear-eyed about the tradeoff. Creators of that calibre give you distribution, and "assuming the creator is good, like really good, like John, right? Somebody who's been able to build an audience, then they don't need you" [2]. The honest framing is partnership: "you transact with them as a partner and you get the distribution from them. I think partners are awesome" [2]. Marpipe has worked with Orurin before and tapped his audience [2]. Build it in-house and "you get to own it for some time and for as long as you retain that employee, but uh it will slip from you and you won't have a um reliable system for working with creators for distribution as well as content" [2]. His proof that distribution beats product differentiation is Death to stock, which he describes as "literally a a commodity SAS product" tearing up its industry purely because it has cracked distribution and no other stock company knows how to [2].

Takeaways

• Catalog ads and DPA are the same thing, and they now account for 24% of ad spend across the 650-plus Meta accounts Marpipe tracks, roughly 3% of global Meta spend [1].
• Enriched catalog ads outperform on both metrics Marpipe measures: 2.4% CTR and 8.18 ROAS, versus 2.01% and 5.2 for non-catalog image and video ads [1].
• Meta accumulates learnings at the source feed level, like a pixel, so uploading a filtered new catalog forces retraining from scratch and generally fails [1].
• Default catalog creative wastes the frame; testimonials, value props and colour belong in catalog ads exactly as they would in an image ad [1].
• Cold email behaves like a retargeting campaign: reliable at five to ten meetings a week from 20% of one person's time, but it will not scale past that, whereas cold calling currently converts because marketers answer [2].
• A low-cost, high-penetration product is only a great strategy if the profits fund a move up-market; otherwise it is a distribution asset left idle [2].
• Hiring an in-house creator buys ownership only while the employee stays, so treat established creators as distribution partners rather than trying to replace them [2].
• Events plus social content, run relentlessly, are what make a company feel ubiquitous, an insight Pantelo drew from co-running the ecom founders events business [2].