Choosing the right metric to optimise your campaign against is essential in harnessing the power of real-time advertising (RTA): the beauty of RTA is that it allows the advertiser to reach the ‘right’ users – those that will help you achieve your ultimate goal – and since the optimisation metric plays a huge part in dictating which users the campaign reaches, it’s important to pick the right one.
Most of the campaigns I’ve managed are prospecting campaigns, looking to drive new acquisitions, and ultimately aiming to drive the lowest CPA (cost per acquisition) possible. The obvious choice of optimisation metric is CPA, since this will drive the campaign to reach those users who are most likely to yield a low CPA.
But, as with most choices in life, the obvious choice isn’t always the right one. Prospecting for new acquisitions is tough: in fact, it’s very tough to acquire a user as a customer through the course of one campaign, when that user, as a prospect, is likely to be relatively unengaged with the brand or may have even never engaged with it at all. Because of this, prospecting campaigns often won’t get more than a few acquisitions, particularly at the beginning. But a few acquisitions are not enough for the campaign to predict the ‘right’ users with any precision, and the power of RTA is somewhat lost.
However, optimising against a CPUL (cost per unique landing) in a case such as this can harness the power of RTA in a way that optimising against a CPA cannot. It is much easier to drive users to the landing page at the top of the acquisition funnel, than it is to drive them down the whole acquisition funnel in a single campaign, and indeed there is little doubt that a prospecting campaign will drive many more landings than acquisitions. By optimising against a CPUL metric, the campaign has far more data with which to predict the ‘right’ users, and will be able to do so with far more precision. As a result, the campaign will drive many unique landings; more unique landings than the same campaign optimised against a CPA.
Of course, if your goal is to ultimately drive acquisitions, it’s not clear that driving landings achieves this in the most efficient way. But what it does do is drive engagement in unengaged users, thus initiating the user into the acquisition funnel. This is an important achievement, because users at the top of the funnel are potential acquisitions, and can be easily turned into actual acquisitions through an additional retargeting campaign. In this way, tackling the acquisition funnel in parts rather than in one go often leads to more user engagement and ultimately more acquisitions. The overall result can be very effective.
There are, unsurprisingly, a couple of things to be wary of when choosing to optimise against a CPUL. Perhaps the most important pitfall of the metric is that it is susceptible to inauthenticity: accidental clicks or click robots will often show up as landings, though they do not represent user engagement at all. And, if a significant proportion of the campaign’s landings are inauthentic, the campaign will struggle to drive acquisitions from them. But, this problem can be easily by-passed, by blocking any low quality inventory from the campaign. Though it is likely to push up CPUL figures, it will drive good quality users into the funnel, users who are ultimately far more likely to become acquisitions.
So CPUL may not be the obvious optimisation metric, and it is certainly not always the right one. But, when running a prospecting campaign, it’s worth bearing in mind how efficiently the CPUL metric can drive user engagement and how, in the right context, it can ultimately drive those all-important acquisitions.
Kruti Shrotri, Account Executive, Infectious Media
