Why Google’s New Selective Optimization Will Change the Game for SMBs
Published: November 20, 2018
Author: Tom Leonard
At this point, if you’re not using or at least testing Google’s Target CPA bidding (tCPA), you’re missing out. It continues to improve, and it allows marketers to achieve better results and have time for more strategic initiatives instead of slingin’ bids.
As great as tCPA is, though, there are a couple of limitations keeping advertisers from diving in headfirst: conversion volume and daily budget. Many smaller advertisers simply can’t afford to take advantage of tCPA, for reasons I’ll dig into below, and this has left tons of untapped potential in the GDN and YouTube.
Enter Selective Optimization.
If you’re anything like me, you’ve been envious of every social account manager who casually says, “We’re not seeing enough purchase volume to effectively optimize; let’s shift that up the funnel to an add-to-cart optimization.” Google’s solution over the years has been to roll your micro conversions into the conversion column to give the algorithm more data. This, however, makes it unclear what exactly you’re optimizing for and what your target CPA should be. Today, the best workaround is creating a separate Google Ads account for each conversion event you want to optimize towards – but this solution is less than ideal because you can’t leverage MCC-level tracking (Google still makes you choose the conversion opted in the conversion column at the MCC level).
The reason tCPA is so important on GDN/YouTube is simple: the reach is immense. There are so many impressions up for grabs that, left to your own manual lever-pulling, you would need months of testing to find success on a prospecting audience and even more if your conversion event is low-funnel like a purchase. And even if you do see plenty of purchase volume, smaller advertisers still may struggle with tCPA for one reason: daily budget.
Google suggests setting your daily budget to 20 times your CPA goal to give the algorithm enough volume. If you’re operating with a CPA goal of $100, you’d need to start your daily budget at $2,000. Then, you’d want to give Google two weeks to let the algorithm learn. Because this is just the learning period, you can’t judge success based on this time period; you need to give it another couple weeks to see if you’re hitting your CPA goal using a well-educated algorithm. Four weeks later, and you’ve spent $56k to test out tCPA. Hope it worked…. ¯\_(ツ)_/¯
This is a steep price to pay for most advertisers. If this is where you’ve given up hope before, keep reading.
As a smart, data-driven advertiser, you know a lot about your purchase funnel. Say 7 out of 10 people who add to cart ultimately purchase. You’d be willing to pay $70 per add to cart to hit your CPA goal. So if you could optimize towards someone clicking add to cart, you’d still need to invest $39,200 into your test. In this case, Selective Optimization isn’t much help. But take it a couple steps further. You may know that about 20 out of every 100 people you drive to your landing page actually view a product page, and 30% of people who view your product page add to cart.
I’ll let you do the math and meet me in the next paragraph.
You’d be willing to pay $21 for someone to complete your “PDP view” micro-conversion. Now you can set your daily budget at $420 for an all-in cost of $11,760 for the same 4-week test. This 79% reduction in cost is much more digestible for most advertisers. Sure, you’re assuming your rates stay the same with an influx of traffic, but I’d argue you’re much better off taking this calculated approach than shooting in the dark with manual bidding on the GDN or YouTube.
With responsive ads and TrueView for action making historically upper-funnel channels more DR focused, an ever-improving tCPA, and now Selective Optimization, lower-volume/spend accounts can now tap into GDN and YouTube with a strong chance of success. This will be a game-changer.
*Selective Optimization is currently in Alpha for SEM/GDN campaigns. YouTube compatibility is slated for H1 2019.