As PPC management becomes more automated, there are some tricks for setting your campaigns apart from everyone else who has access to the same automation tools. Everyone can now get average results thanks to machine learning, but if you’re aiming for stellar results, that requires knowing how to collaborate with the machines.
One way to help the machines is by teaching them more about your business — be the PPC teacher to the learning models behind today’s most popular automations. There are two factors that you may not have optimized before that have now become critically important: how you communicate your goals and how you report conversion values. In this post, we’ll review how to get started with this new way of optimizing.
A shift in how PPC is optimized
Optimizing goals and conversion tracking may not be the PPC optimization tasks you’ve been doing for years. But in this new world, they’re taking on the same importance as what advertisers are more used to, like selecting keywords, making bid adjustments, and doing ad optimizations.
Think about the machine as a key new hire on your team. To get new team members to perform at their peak, you teach them the ins and outs of your business. What is it you’re trying to achieve, and how do you measure the outcomes? Machines, like humans, benefit from this training and makes it more likely they will deliver the sort of results you’re after.
Value-focused PPC optimization
Ad platforms like Google Ads provide advertisers with ways to control goals and value reporting. Let’s review how to use these to teach the machines better what they should do to make you happy.
Say there’s a retailer with a catalog comprising thousands of products, and their goal is to increase profits for their company. We’ll encounter their ROAS goal in a moment, but keep in mind that the target ROAS should be just a setting the advertiser uses to achieve their true goal of profitable PPC.
There are two basic ways to direct the machines towards meeting this profitability goal, and both have to do with helping the machines understand the advertiser’s break-even point:
- In the first scenario, the advertiser has many campaigns, each with a different tROAS target based on profit margins. They also use conversion tracking to report the order value of each sale.
- In the second scenario, they maintain just one campaign with a single tROAS. They report profit value through conversion tracking.
Scenario 1: optimize the goal to reach profitability
Scenario 1 feels more natural to many advertisers because it is built on the existing account structures they’ve had for years. There is a lot of structural granularity, and performance is optimized by setting different targets for different elements, e.g., by setting a different tROAS for every campaign. This is very similar to what advertisers used to do in the days of manual bidding, where each ad group or keyword had its own bid.
The tROAS goal of each campaign is a factor of the average profit margin for the products in the campaign. A campaign with higher-margin products can break even at a lower ROAS than a campaign with lower-margin products.
Because conversion tracking reports the value of the order, the margin-based tROAS steers the campaigns towards profitability. For example, the campaign selling products with a 50% margin can break even at a 200% ROAS. So a $100 order in that campaign equates to $50 of profit. Because of the 200% ROAS target, the campaign should return roughly $2 in sales for every advertising dollar spent, or it can spend $50 to get the $100 sale. So $50 was spent on ads to get $50 in profits from a $100 sale. The campaign breaks even. Goal achieved!
Scenario 2: optimize conversion values to reach profitability
Scenario 2 is more in line with what Google advocates in the age of automation: a flatter structure. But for this to work well, it assumes advertisers do a good job communicating the correct value they receive from conversions.
When advertisers use conversion tracking to tell Google how much profit they get from a conversion, then any performance at or above the 100% ROAS target will be profitable. If, on the other hand, the advertiser reports the order value rather than the profit of that order, a 100% tROAS won’t deliver good results because it will treat sales of high and low margin products the same.
Continuing with the example from scenario 1, if the $100 sale resulted in $50 of profit and that $50 was reported as the conversion value, then the 100% tROAS of the campaign means that the system can spend $50 to get that sale. It’s the same result as in scenario 1, but using different settings to achieve it.
Reporting profitability through the conversion value field has the added benefit that it optimizes for scenarios where a user buys something else than what was in the ad they clicked on. It handles the issue of users jumping between products of different margins once they reach your site from an ad.
Optimize both goals and value reporting
Both scenarios above are used to achieve an advertiser’s true business goal: delivering profitable PPC clicks. But either implementation is imperfect because it doesn’t take full advantage of all the ways you can teach the machines what you truly want.
That brings us to the third scenario where advertisers leverage both controls: the target will reflect a true goal, and the value will reflect a true measure of contribution.
Advertisers with granular account structure and optimized value reporting are in the best position to teach the machines about their business goals.
For example, they can set different profitability goals for different business lines. An advertiser might notice that patio furniture is nearing the end of the season and that it’s better to forego profitability in exchange for not being stuck with unsold inventory in winter. Granular account structure allows them to treat different items with different goals.
But because they also report the true business contribution value of a sale of a deck chair, when they say they are okay with breaking even rather than making a profit on that sale, the automated bidding process can set a good bid, meanwhile continuing to optimize other parts of the account towards higher profits.
Now let’s take a look at implementing a strategy that optimizes PPC with an account structure that supports setting correct business goals and reporting conversion values that reflect the true business value.
Create PPC campaigns faster
Setting up many campaigns to support a variety of business goals is something well understood by PPC advertisers. There’s no magic to it, just lots of hard and repetitive work unless you have tools like those from Optmyzr at your disposal. For example, the Campaign Automator can take business data like a product feed and turn it into thousands of neatly organized ad groups in campaigns focused around common elements like product category, brand, and sales priority. The Optmyzr Shopping Campaign Builder and Refresher can do the same for shopping campaigns and product groups. These tools can also restructure campaigns with ease if the PPC team decides to pivot their strategy on different attributes that require a different structure.
Once advertisers have built out the perfect structure to support their business goals, they can move on to optimize how they report conversions to Google.
Optimize value reporting
Optimizing conversion values is new territory for many advertisers. The value data already flows through conversion tracking, and using that as an optimization lever has traditionally not been top of mind for many advertisers.
But now that it is a priority, there are three options for advertisers to report better conversion data to Google when they can’t send the correct value in real-time with the conversion tracking pixel. Those options are:
- Offline conversion import (OCI) lets an advertiser create additional conversion signals using the Google click ID (gclid).
- Conversion adjustments let advertisers restate the value of a conversion with a transaction ID.
- Conversion value rules (beta) lets advertisers build rules to modify the reported value based on criteria like audience, location or device.
Conclusion
To optimize PPC in an automated world, it’s increasingly important to know how to set the high-level inputs that teach the machines about our businesses. By better communicating true business goals and the value a business derives from a conversion, the machines can be taught to prioritize the right types of conversions at the right cost. And to achieve this, you can deploy a mix of solutions like Optmyzr’s tools for creating advanced account structures that align with business goals and free tools from Google for fixing conversion values once you know more about the quality of a conversion.
