A beauty brand's starter guide to Instagram advertising

Measurement comes before the ad itself. You can't optimize a campaign you can't measure — and not being able to answer 'is this working?' in the first weeks is what makes most people quit.

• Install the Meta Pixel together with the server-side Conversions API; test that product views, add-to-carts and purchases actually fire. Meta describes the Conversions API as a connection that sends marketing data from your server to its systems to optimize ad targeting, decrease cost per result and measure outcomes; server events link to the same dataset ID and are processed like Pixel events.
• Switch your Instagram account to a professional account and organize it under a Business Manager / business portfolio; don't run ads off a personal profile.
• If you sell online, connect your product catalog: Meta lists adding products to a catalog and connecting a data source (the Pixel or Conversions API) as prerequisites for Advantage+ catalogue ads. Retargeting is one of that format's two audience options.
• Prepare your profile for ad traffic: your bio, highlights and recent posts are the first storefront a visitor from an ad will see.

In cosmetics, the purchase decision rarely happens in one touch: a user sees the product, meets it a few more times, checks reviews, then buys. Your campaign structure should mirror that journey.

Cold audience: introduce

For people who've never heard of you: video that shows the problem and the result and hooks in the first three seconds. The goal here is qualified traffic and engagement before sales.

Warm audience: convince

For people who watched your video, visited your profile or landed on your site: trust-building content — social proof, how-to-use, ingredient stories.

Hot audience: remind and close

Dynamic retargeting for cart abandoners and product-page visitors. At this stage the clarity of the offer matters more than the creative: shipping, easy returns, bundle value.

On a new account with a small budget you don't have to build all three layers separately from day one — and Meta gives its own reason to simplify. It recommends avoiding high ad volumes: when you create many ads and ad sets, the delivery system learns less about each one, and combining similar ad sets also combines their learnings (Meta Business Help Center). What matters is the logic: a different message for people at different stages — delivered through as few ad sets as your budget can actually support.

In beauty, creative is more decisive than targeting. But be skeptical of anything sold as a 'proven formula' — different product, price and audience means different results. The right approach: race a few distinct angles against each other on a small budget.

• Real usage footage: the product's texture, application, finish — the product itself, not a stock image.
• UGC-style delivery: user language, phone-shot feel, honest tone. (If it's creator content, don't pass it off as an organic customer — let it be what it is.)
• Problem → solution framing: open with a real annoyance like 'midday shine' and present the product as the fix.
• Short copy, single message: one benefit per ad; an ad that lists five benefits communicates none.

A category-specific warning: cosmetics and personal care sit in a sensitive zone of ad policy, and the rules are written down in Meta's Transparency Center. The personal attributes policy prohibits ads that directly or indirectly assert or imply attributes such as a person's race, religion, age, sexual orientation, disability, or physical or mental health including medical conditions (Meta Transparency Center). An opener like 'embarrassed by your dark spots?' hovers right on that line.

On before/after imagery there's a widespread misconception: the restriction is scoped, not blanket. Meta's Health and Wellness standard prohibits side-by-side comparisons in two places — weight-loss transformations, and wrinkle or anti-aging treatments such as Botox and dermal fillers. The same standard, however, permits ads targeting people 18 and older to show general cosmetic products, procedures and surgeries depicting a before-and-after transformation, provided they don't employ negative self-perception tactics. For anti-aging products, zoomed-in or close-up images are allowed, but they must reflect realistic outcomes over time and must not be side-by-side (Meta Transparency Center). A makeup before/after and a cosmetic-procedure before/after are not the same case — but the permission depends on getting your targeting and your language right.

The body-image clause applies directly to this category too: the same standard prohibits messaging that could make people feel negatively about the way they look, exploits insecurities to conform to certain beauty standards, reinforces negative body images or features body-shaming. Practical translation: sell the product on a concrete benefit, not on a deficiency the viewer is supposed to feel about themselves.

There's no universal answer to 'how much should I start with' — it depends on your price point, margin and goal. But a few principles are universal:

• Treat the first weeks as a learning period — it's a defined phase at Meta. Ad sets usually exit the learning phase after about 50 results in the week following the ad set's last significant edit, and during that phase ad sets are less stable and usually have a higher CPA (Meta Business Help Center).
• Don't edit while you're in it: Meta states plainly that editing an ad, ad set or campaign during the learning phase resets learning and delays the delivery system's ability to optimize. Panic-toggling campaigns day to day triggers exactly that reset.
• Keep the structure as simple as your budget: one campaign, a few creative variations; a small budget split across ten campaigns learns nothing anywhere. Meta likewise recommends realistic budgets — a very small or inflated budget gives the delivery system an inaccurate indicator of whom it should optimize for.
• Pick your decision metric up front: for e-commerce that's usually ROAS and cost per purchase (CPA); clicks and likes are not decision metrics.
• Work backwards from margin: without knowing the CPA at which you're profitable, you can't answer whether the ads 'work'.

• Running ads before the pixel and conversion events are set up — you spend money and learn nothing.
• Settling for 'Boost post' — you get engagement but never use the optimization power of a sales campaign.
• Checking results daily and toggling campaigns in a panic — you keep resetting the learning phase.
• Launching with a single creative — when it fatigues (frequency climbs), you have no alternative ready.
• Great ad, weak product page: sending paid traffic to a slow, trust-poor page.