Skincare marketing agency

Skincare search is ingredient-led, so the Shopping title has to be too. A title that reads Brand Serum 30ml loses to one that reads Brand 10% Niacinamide Serum 30ml on every query that contains the active, because Google matches the query against the title first. We rebuild titles in that order: brand, active and strength, product type, size, and skin concern where it fits.

Search campaigns split by intent. Ingredient queries such as vitamin C serum or salicylic acid cleanser are cold and comparative; they get a tight ad group with the page that lists the strength and pairing advice. Concern queries such as serum for dehydrated skin go to a page built around that concern. Brand queries get their own campaign so you can see what is actually driven by awareness elsewhere.

Ad copy stays inside the cosmetic line. Describe what the formula does to the appearance of skin: hydrates, smooths, evens tone. Copy that says treats, cures or heals a named condition triggers a healthcare review and usually a disapproval, and the product stays out of the auction until it is fixed.

Skincare creative on Meta and TikTok works when it shows the product being used in a real routine: the drop count, the texture on the back of a hand, where it goes in the order, what it looks like under moisturiser. Shoppers who have never touched the formula need that visual, and it does not need a claim attached to it.

Two policies shape what we make. Meta does not allow ads to assert or imply something about the viewer, so copy like your acne-prone skin needs this is rewritten as for skin that is prone to breakouts. Before-and-after images that suggest a health result are restricted, and skin close-ups can be flagged as well. We keep results claims off the creative and let reviews and ingredient facts carry them on the page.

Audiences follow the routine, not the demographic. People who bought a cleanser see the serum that follows it. People who bought the sample see the full size. Creators are briefed on the honest use case, with the layering advice included, because the comments under the video are where the second wave of shoppers decides.

A skincare feed goes wrong in predictable places. The 30ml and 50ml of the same serum are submitted as one product with one price, so the cheaper query lands on the wrong price and the click wastes money. We split sizes into separate offers grouped by item_group_id, with size stated in the size attribute and in the title, and unit pricing measures filled where the market shows price per millilitre.

Identifiers cause the second wave of problems. A brand that owns its formulas usually has GTINs from its own barcode range; those go in. A brand that does not have one for a product should mark identifier_exists as false rather than leave the field blank or borrow a code, which gets the offer disapproved for a mismatched identifier.

Then the descriptive fields do the work. Product highlights carry the active, the strength, the skin type it suits and the fragrance-free or vegan facts. Product detail pairs carry skin concern and usage time. Images are the bottle on a plain background with no overlaid text, because a promotional overlay is a disapproval reason on its own. The description repeats the how-to-use, because Google reads it for matching and shoppers read it in the Shopping panel.

Skincare search has three layers and a store needs all of them. Product pages rank for the exact product. Collection pages rank for the concern: dehydrated skin, dullness, uneven texture, redness. Ingredient pages rank for the active: what niacinamide does, what strength to start with, what not to layer it with. Most skincare stores have the first layer only, so the concern and ingredient traffic goes to publishers.

Shoppers now ask AI assistants the same questions in full sentences: can I use vitamin C and retinol together, what strength of retinol should a beginner use, is this serum fine for sensitive skin. Assistants answer from pages that state facts plainly and cite usage. We write product and ingredient pages as direct answers with clear headings, list the pairing rules and cautions in plain text, and keep the specifications in the visible HTML rather than in an image.

Structured data does the rest: product markup with each size as an offer, FAQ markup on the layering questions, and review markup that reflects reviews actually on the page. Internal links run from the ingredient page to the products that contain it, and from the concern page to the routine that addresses it.

A skincare store's email calendar is a set of run-out dates. A 30ml serum used morning and night lasts roughly six to eight weeks. A 50ml moisturiser lasts longer, a 15ml eye cream longer still. A flow that sends the reorder reminder at thirty days for everything is early for some products and late for others, and a late reminder is a customer who has already bought elsewhere.

We map each product to an expected usage window from its size and directions, then set replenishment flows per product. The first message lands about a week before the estimated empty date, with a one-click reorder and the option to switch to a subscription at a small saving. The second message, for people who did not act, asks whether the product suited them and offers the alternative size or the neighbouring product.

The sample-to-full-size flow is separate. Someone who bought a discovery set gets a sequence timed to when each sample would be used up, with the full size of each one and the routine order explained. SMS is used for the reorder reminder only, because that is the one message a skincare customer actually wants on their phone.