Snapchat Ads

The Audience

Snapchat skews heavily toward under-35s, with very high penetration among teenagers and young adults in several markets. Usage is frequent, short-session and camera-led.

For brands targeting that demographic it offers reach that is genuinely hard to replicate. For brands targeting older audiences it is largely irrelevant, and this should be established before any planning.

Ad Formats

Snap Ads — full-screen vertical video up to three minutes, skippable, with swipe-up actions.

Story Ads — a branded tile in the Discover feed opening to a series of snaps.

Collection Ads — product tiles beneath a video for direct shopping.

AR Lenses — branded augmented reality experiences users apply to their own camera. The distinctive Snapchat format.

Filters — branded overlays available in a location or to an audience.

AR Lenses

Lenses are the format Snapchat is genuinely differentiated on. Users actively choose to apply them, often share the result, and spend substantially longer with the brand than any passive format achieves.

They suit try-on categories — cosmetics, eyewear, footwear — and entertainment or game tie-ins. Production cost is higher than standard video and the engagement profile is different enough that it should be measured differently.

Creative Requirements

Vertical, full-screen, sound-on, and fast. Snapchat viewing is rapid and the first moment decides everything, as on TikTok.

Snapchat creative also rewards authenticity over polish. Content that looks like a snap rather than an advertisement performs better, which again means creator and user-generated content is frequently the strongest source.

Where It Fits

Strongest for youth-focused consumer brands, entertainment, gaming, fashion, beauty and app install campaigns targeting young users.

Weakest for B2B, older demographics, and high-consideration purchases where the platform's session behaviour does not support deliberation.

Treat it as a specific tool for a specific audience rather than a general performance channel.

What to Test First, and With What Budget

The most common way a Snapchat test fails is that it was never given the conditions to succeed: too little budget, too many audiences, and creative borrowed from another platform.

A defensible first test. One objective, chosen for what you actually need to learn. Two or three audiences at most, each with enough budget to exit the learning phase — splitting a small budget across six audiences produces six inconclusive results. Native creative, shot vertically and sound-on, not a repurposed asset. A run long enough to clear the learning period and then produce readable data, which is usually a fortnight rather than a few days.

Instrument it properly before launching: the pixel or SDK firing correctly, the conversion events you care about defined, and — where available — a server-side connection, because the browser-side signal on mobile is the weakest part of the measurement chain.

Lenses: the Differentiator and Its Real Cost

AR Lenses are the platform's genuine differentiator, and the reason they under-deliver for most advertisers is a budgeting error rather than a creative one.

A Lens has three costs, and most plans carry one. The media to distribute it. The production — a designer or studio working in the Lens toolset, which is a specialist skill and a real timeline. And the concept, which is the part that decides whether it works: a Lens people want to use for its own sake, rather than an advert wearing a filter.

The ones that perform tend to do something functional — a try-on that answers a real question about fit, colour or scale — or something genuinely playful that people want to send to a friend. Sharing is the whole mechanism: an unshared Lens is expensive display advertising.

So the metric to judge it on is not impressions but shares and play time. And the decision to make before commissioning one is whether you have a concept that survives the question why would anyone send this to a friend — because if the answer is thin, the media budget is better spent on formats that do not depend on it.

Sources

What each claim on this page rests on. Entries are typed so you can see which are primary.