Ipsy, the beauty subscription service with over 2.5 million active subscribers, has launched a marketing services arm that sells brands direct access to its member base, according to Modern Retail. The company now charges beauty brands for custom sampling campaigns, influencer activations through its creator network, and data packages on product performance—converting its distribution model into a two-sided revenue engine.

The mechanics are straightforward. Brands pay Ipsy to design and execute test campaigns inside the subscription box ecosystem, then receive performance data on open rates, repurchase intent, and demographic breakdowns. Ipsy bundles this with optional paid media placements across its owned channels and creator partnerships, effectively selling the same customer file three times: as a subscriber, as a data subject, and as an audience segment.

This works because Ipsy controls the choke point. A beauty brand launching a new serum has two paths: spend six figures on Instagram ads hoping to reach the right cohort, or pay Ipsy a fraction of that to place 10,000 samples in bags already going to verified beauty enthusiasts, then buy the performance data showing which zip codes reordered. The brand gets attribution. Ipsy gets paid twice. The subscriber gets product. No one loses.

The underlying pattern is audience arbitrage. Ipsy built a list by subsidizing discovery—selling $12 bags of $50 worth of samples. That list became valuable not because of its size, but because it's self-selected, behaviorally verified, and tied to purchase data. By opening a services desk, Ipsy monetizes the margin between what a brand would pay Meta for a lookalike audience and what it costs Ipsy to activate its own file. The company captures the difference without adding inventory risk.

A small physical-product brand runs this same play at micro scale. You build the list first, even at a loss. Sell a sampler box, a trial set, a welcome bundle—anything that gets a buyer to raise their hand and declare interest in your category. Once you have 500 verified customers, you approach adjacent brands in your niche and offer to co-pack their sample into your next shipment for a flat fee of $250. You're selling access, not media. The brand gets 500 warm intros. You net $250 to offset your fulfillment cost. Your customer gets an extra sample. If the adjacent brand converts 50 customers at a $30 average order, they just paid $5 per acquisition. You've built a service line from your existing operation without adding overhead.

The steal tightens when you add data. After the co-pack ships, send the partner brand a one-page PDF: how many boxes included their sample, how many customers clicked the QR code you insisted they include, how many used the discount code. Now you're not just a co-packer—you're a performance channel. Next quarter, you charge $400 for the same 500 samples because you can prove conversion. The list becomes an asset that appreciates as you demonstrate reliability.

The broader move here is recognizing when your customer list is more valuable as a media property than as a buyer pool. Ipsy didn't invent this—trade publishers have sold access to their subscriber bases for decades. But Ipsy applied it to physical product by making the media vehicle itself a tangible good: the box. A one-person brand does the same by treating every fulfilled order as a billboard someone paid you to deliver. You already have the shipping cost sunk. The incremental cost of including a partner sample is near zero. The incremental revenue is pure margin. You've turned your COGS into a two-sided marketplace without changing your fulfillment process.