Marketing teams are always tasked with finding ways to grow the share of ‘Organic’ or ‘Direct’ traffic. After-all, in the long-run, the only way online businesses can become sustainable, is if most of the traffic they get, comes to them organically.
However, in the short run, paid channel marketing teams are often incentivized to lower the Cost per Acquisition (CPA). It is in this context, teams often look to grow the share of retargeting conversions, because retargeting vendors typically surpass other channels, in terms of delivering the lowest CPA’s.
But marketers, who look at their data a little more closely, quickly come to realize that typical ‘black-box’ retargeting vendors deliver that superior performance, at the cost of organic traffic. To illustrate why this happens it is important to look at how retargeting works and some of the factors that affect retargeting conversions. Let’s begin by look how two key factors – ‘funnel position’ and ‘recency’ affect retargeting.
As is depicted in the figure (Fig.1) below, let us assume, if we were to retarget consumers with ads, who had just visited the home page of an e-commerce app, the resulting conversion rate were ~X.
Now if we were to retarget users who were abandoning cart, given that users in this lower funnel segment have spent more time engaging with the product they wanted to buy, they would have a higher propensity to convert. Let us therefore assume the conversion rate would be ~5X.
As is depicted (Fig.2) below. In a typical e-commerce site, 80% of the conversions happen within the first 48 hours of the user having engaged with the product.
The next 5 days typically, account for 14% of the conversions, and the remaining 21 days account for 6% of the conversions.
This behavior varies by vertical, but is indicative of the broad trends we see. Thus, if we were to retarget users who had visited in the last 48 hours the likelihood of conversion would be much higher than for the next 5 days.
The conversion rates we would see for the following 21 days would be a fraction of the rates we see for the first 7 days.
Effect of Recency and Funnel Position on eCPM
Having independently looked at the effect of ‘Funnel Position’ and ‘Recency’ on retargeting conversion rates – it would be useful to start looking at how these two factors start affecting the effective cost of buying media or eCPM, in the context of retargeting.
Whatever pricing model your retargeting vendor is offering you, (CPC, CPT, CPA, CPS – which is % of Revenue) all these pricing models effectively result in the retargeting player taking a CPM bet on your users. Irrespective of pricing model offered to marketers, eventually your retargeting partner is buying media on a CPM basis. So when your retargeting vendor is offering you a CPA model for pricing they are simply doing the following calculation:
Avg. CPA = Avg. CPM /(Avg. CTR * Avg. CTC)
The Avg. CPM or more specifically the Avg. eCPM (where e stands for “effective”) is just a derivative of the above equation
Avg. eCPM = Avg. CPA * Avg. CTR * Avg. CTC
This avg. eCPM (that is derived from the Avg. CPA) price for the client is a flat line a, depicted in (Fig. 3)
The actual CPM price available to DSP’s just buying media directly on exchanges is depicted as a flat line b, in the same (Fig. 3)
However, there is a substantial difference between the eCPM of specific cohorts of users and the average eCPM illustrated by taking five specific cohort of users as defined below and depicted in (Fig. 3).
1. Cart Abandoners + Recency Day 2
2. Cart Abandoners + Recency Day 7
3. Product Viewers + Recency Day 2
4. Product Viewers + Recency Day 7
5. Category Viewers + Recency Day 2
Cohort 1, has a highest propensity to convert given that they were furthest along in the conversions funnel at the cart position and had recently abandoned their cart. Thus the eCPM of this Cohort is the lowest
Similarly the eCPM of Cohort 5, is the highest because these users are at the top end of the conversion funnel and thus have the least likelihood of converting with typically lower CTR’s and CTC’s
The eCPM curve across these 5 Cohorts of users typically has an exponential curve c in (Fig. 3).
How retargeters make money?
The area between line a and line c is the gross margin retargeting players make. Naturally, the gross margin made on users who have a higher propensity to convert is much greater. Thus retargeting vendors have a perverse incentive to chase after proverbial ‘lowest hanging fruit’, which is also typically your traffic that would have come-back and converted organically.
This is the reason why most retargeting players, run a black box model and do not show you the audience cohorts they are targeting and the margins they are making on each of these cohorts.
So what should marketers do?
As the client paying the monies to the vendor, and being responsible to the P&L head to spend their marketing budget most judiciously, the marketing manager has the full rights and the fiduciary duty to ask for where these monies are being spent.
As the world of advertising moves increasingly towards increased transparency and flexibility, it is imperative, that retargeting vendors will have to reveal, the audience cohorts they are targeting and where they are serving the ads.
As marketers, you have the right to ask your retargeting partners to block specific cohorts, you believe, have a high propensity to convert like ‘cart abandoners in the last 24 hours’.
Doing such an exercise, will immediately lead to a rise in your share of organic convertors.
So go ahead and try the above. You may be honestly surprised by the results!
RevX is an app retargeting platform that powers growth for mobile businesses through dynamic retargeting. The platform is built on integrated and transparent technology combining four key pillars - audience intelligence, programmatic media, personalized ads, and ROI optimization. Mobile marketers across verticals like e-Commerce, travel, lifestyle, hyperlocal and gaming use RevX to enhance user engagement by activating new users, converting existing users and re-activating lapsed users.