Color cosmetics is the most visually rewarding category in beauty to advertise and the least forgiving to scale. The creative works, the swatches stop the scroll, the cost per click looks enviable next to skincare. Then the contribution never quite arrives, and the founder is left explaining a healthy cost per purchase to an investor looking at a flat bank balance. The gap is almost always in the same place, and it is not the media.
Makeup sells a choice, skincare sells a routine
That one sentence explains most of the difference. A serum is bought once and reordered on a schedule the product itself creates, so the second purchase is designed in. A foundation is bought after a decision between shades, and the wrong decision comes back. Skincare's repeat behaviour is a feature of the category. Color's repeat behaviour has to be built, one reason at a time.
The practical consequence is that a color brand cannot run a skincare growth plan, which is exactly what most of them inherit, usually from an agency whose other accounts are skincare. The plan will look sensible and will quietly assume a reorder that is not coming. How differently the two categories behave inside the same media buy is covered in our comparison of Meta and TikTok for beauty acquisition.
The return rate belongs in your acquisition cost
This is the correction that changes the most numbers in color cosmetics. In most brands, returns sit in a customer service report and never touch the marketing sheet. In color they are not a service failure, they are a shade mismatch, which means they are a direct output of the creative and the product page. A return is an acquisition you paid for and did not keep.
Move the number. Take your contribution per order, reduce it by your return rate, and recalculate how many orders a customer has to place before they pay for themselves. The threshold moves further than founders expect, and it moves for a reason you can actually fix.
Worked example, substitute your own inputs
A color brand runs a $28 average order value at a 68% gross margin, so a first order contributes $19.04. Acquisition cost is $52. Break even therefore needs 2.73 orders from that customer.
Now put returns in. At a 12% return rate, effective contribution per order drops to $16.76, and the break even threshold rises to 3.10 orders. Nothing about the media changed. The business just got 14% harder.
Reducing returns from 12% to 6% recovers more contribution than a 10% cut in media cost would, and it is usually the easier of the two to achieve.
Shade is a conversion problem, not a merchandising one
Every color brand knows shade range matters. Fewer treat it as a paid media variable. If your creative shows one hero look on one skin tone, you are asking a large share of your audience to imagine the product on themselves, and imagination converts badly and returns worse. Creative built across the range costs more to produce and pays for itself twice, once in conversion rate and once in the returns you avoid.
The same logic runs onto the product page. Shade finders, swatches on multiple undertones, and honest wear photography are conversion rate optimisation, not brand polish. Judge them on return rate, not on how the page looks.
TikTok Shop changed where the decision happens
Color cosmetics is the category where social commerce landed hardest, because the product demonstrates itself in a way a serum cannot. The consequence for growth is structural: a meaningful share of the purchase decision now happens inside an app, on someone else's content, with no visit to your site at all.
That is an opportunity and an attribution problem in the same motion. The opportunity is obvious. The problem is that a channel converting outside your site will look worse in your own analytics than it is, and a brand judging it on last click will cut the thing that is working. Read it on blended efficiency rather than platform attribution, and sanity check the media price against our Meta advertising cost benchmarks for beauty brands.
On search, stop competing with the retailer
Broad makeup search terms belong to Sephora, Ulta and a handful of publishers with twenty years of domain authority, and a growing brand will not take them. The winnable ground is narrower and better: shade matching, undertone, wear time, layering, formulation for specific conditions, and honest comparison against products the buyer already owns. That traffic is smaller and closer to a decision.
It is also the traffic the AI answer engines increasingly quote, because specific beats generic when a machine is choosing what to cite. A brand that publishes genuinely specific answers earns mentions a retailer's category page cannot.
The four numbers a color brand has to know
- Contribution per first order, which is average order value multiplied by gross margin.
- Return rate, measured on units rather than on tickets raised.
- Effective contribution, which is contribution reduced by that return rate.
- Break even order count, which is acquisition cost divided by effective contribution.
Drops make media planning harder than it looks
Color runs on launches in a way skincare does not, and a drop calendar quietly breaks the assumptions most media plans are built on. Budget arrives in spikes, creative has a shelf life measured in weeks, and the retargeting pool built during a launch is gone before the next one. Brands that treat each drop as a separate campaign relearn the same lessons every quarter.
The fix is to run an always on prospecting layer underneath the drops, sized to keep the audience pool warm between launches, and judge it on its contribution to launch week rather than on its own immediate return. Cutting it between drops is the most common and most expensive mistake in the category.
What to do this quarter
Three things, in order. Put your return rate into your acquisition maths and recalculate your break even order count, which will take an afternoon and will probably change your view of which channels are working. Rebuild one core creative concept across your shade range and measure it on returns as well as on conversion rate. Then pick two narrow search questions your product genuinely answers better than a retailer does, and publish real answers to them. Everything else, including channel mix and bidding strategy, is downstream of those three. Our beauty growth practice runs this order at the start of every color engagement, and more of the working is in our insights library.