Snapchat ads A fifth channel has to earn its place.

That belief is out of date, and it is also not the question worth asking. The platform skews younger than the other large social networks, but it has aged with its original users and its audience in Australia is wider than the stereotype suggests. What actually decides whether it is worth your money is not the average age of the platform — it is whether the specific people who buy your service are there, and whether your purchase can survive being interrupted. Those are answerable questions about your category and we answer them before recommending a test.

Usually not yet, and this is the most common honest answer we give about this channel. Adding a platform adds a creative bill, a measurement surface and a reporting line, and it takes attention away from the channels already producing counted work. The case for adding it strengthens when the existing channels are saturated at your target cost, when your buyers genuinely are present here, and when you have the creative capacity to keep a full-screen format supplied. If any of those three is missing, the money is better spent deepening what already works.

As a starting point, not as a plan. This is a full-screen, sound-on, vertical environment, and an asset cropped out of a feed placement arrives looking exactly like an asset cropped out of a feed placement. The parts that transfer are the idea, the claim and the offer; the parts that do not are the framing, the opening and the sound design. Re-cropping is the single most common reason a test on this platform reads as a failure — the channel gets blamed for a decision made in the edit.

The same way we insist everything is measured: a browser-side signal and a server-side path together, outcomes defined as things that happen in your business rather than as page views, and one reporting window agreed across every channel before anything runs. That last point matters more here than people expect, because every platform sets its own period for claiming credit and those periods differ. Without one agreed window and one agreed source of truth, you get a quarter of arguing about whose number is right.

Because they are counting different things over different periods, and both can be correct at once. A platform reports outcomes it believes it caused, within a window it sets, and revises that figure for days as delayed reporting arrives. Your own records only know about people who identified themselves — they made contact, booked, or paid. The work is not to make the two agree, because they will not. It is to decide which one the business runs on, name the size of the gap and explain the reason for it in writing.

Anything expensive, slow and heavily researched, and anything with no visual dimension at all. A full-screen social ad interrupts; it does not answer a question somebody was already asking. Where the purchase takes months and involves several people, the interruption can start a relationship but will not close one, and judging it on a fortnight of enquiries will produce a wrong answer. For those businesses the first money belongs in demand that already exists — search — and in the follow-up that catches the people who were interested but not ready.

They are genuinely native to this platform and they are also the part most often bought for the wrong reason. A branded camera experience is a brand instrument: it earns its cost when there is an occasion, an audience already inclined toward you, and a reason for somebody to want to use it. It is not a lead-generation mechanism and we will not sell it as one. For a service business testing this channel for the first time, straightforward vertical video is the sensible starting point.

Yes, and the list is not identical to the other platforms', which is exactly why it gets missed. Advertising policy differs by platform and by market, some categories are restricted or prohibited outright, and claim rules vary too — wording that clears one platform's review can be refused by another. We read the policy position for your category and your market before anything is produced, because discovering a restriction after a shoot is an entirely avoidable cost that businesses pay all the time.

Enough that the result can be read, which is a question about your own numbers rather than a figure we can put on a page. It depends on what a customer is worth to you, how often they buy, and how large a difference you would need to see before you would act on it — and we work that out with you before anything is switched on rather than after. What we will not do is quote a confident minimum, because a number invented for a web page is exactly the sort of thing this site does not publish.

Expect the first fortnight to understate the result and to be read that way — delayed reporting means a period that ends before the delay expires is always incomplete. A month is the earliest point at which the first reconciliation against your own records is meaningful, and a fair test usually needs longer than that plus enough creative variety that you are testing the channel rather than one video. Anybody quoting a confident cost per result in week one is quoting a number that is still moving.

Yes, in writing, and this is the part of the arrangement worth checking with any agency. If after a fair test the channel is not producing counted work against your own records, the recommendation is to stop and move the money to where it is. A channel that keeps running quietly at a low level because nobody wanted to have the conversation is a cost with no owner, and it is one of the most common things we find when we take over an account.

Not one we publish, and we would rather write that here than let a page imply otherwise. There is no Snapchat case study, no cost per result and no account screenshot for this channel anywhere on this site. What we can show is the same discipline applied in depth on other channels, with the figures dated: outcomes counted rather than assumed, reconciled monthly against the client's own records, and reported as one number with its working shown. If proof on this specific platform is what you need before you commit, that is a fair requirement and you should hold us to it.