At virtually every digital marketing conference I attend, I hear the low rumble of agency practitioners concerned about the growth of “smart” advertising platform technologies.
Google’s hyper-automated Universal App campaigns, Smart Bidding, Dynamic Search and Responsive Search Ad technologies are just the tip of a very large iceberg of automated advertising options that are getting an increasing foothold into search marketing. These automated systems move digital marketing work away from minute bidding, placement strategies (and hair-splitting keyword strategies), and increasingly relies on entrusting the advertising platforms’ proprietary data and machine learning technologies to put ad copy in front of likely-to-convert users at a time that is deemed most likely to contribute to a “conversion” or acquisition.
While advocates of these automated technologies boast about the time and/or money saved by deploying them, when it comes to actually using these systems, what is often left a little hazy is what the day-to-day management of “smart” campaigns actually looks like. When the critical details are left out of what managing "smart" campaigns entails, it can not only disenchant the workers who have spent years honing their mastery of more manual advertising efforts , but it can also call into question the relevance of agency management - and particularly the relevance of agencies that advertise for SMB companies.
If a small business owner can simply name their price, what’s the point of hiring someone to do it for them? If advertisers can rely on Google to automatically create keywords and ads based on their website content, why don’t they just create a killer website and allow Google to automatically generate ad campaigns based on landing page content? The truth is that “smart” campaigns can drive excellent traffic and conversions for advertisers, but they require as much attention to detail to run effectively.
In this article, I will focus on ways in which agencies and marketing professionals should harmonize these smart systems' ability to aggressively pursue conversion events with overall profits. Focusing on deployment of Google Ads’ Target CPA bidding strategy as a use study, we’ll go over some new best practices in the automation age. By understanding the infrastructure needed to run and nurture these campaigns, we will also see how agencies are especially well-placed to develop scalable strategies that can produce great results for their clients.
All smart bidding strategies rely on machine learning systems that recognize patterns of data that align with conversion events. Your smart platform’s ability to optimize for conversions is only as good as your ability to communicate high quality conversion events to your advertising platform.
Make sure that you are keeping track of every desirable conversion and are communicating it appropriately. If you are in the e-commerce business, there are a multitude of ways to set up conversion tracking systems that will allow you to connect. If you rely heavily on offline conversions, such as phone calls or store visits, there is no perfect solution for connecting every conversion with every ad impression or engagement; however, attribution technologies are getting better and better.
Using Google Tag Manager’s website Call Conversion tracking tags, we were able to increase our offline conversion tracking capabilities by 500%. Or, if you would prefer to have advanced transcription capabilities baked into your conversion strategy, look into how you might use call management services such as Twilio or CallRail to manage call conversions. If your account is large enough, use Data Driven Attribution to make sure ad engagement experiences and conversion attribution are most effectively weighted.
In addition to counting all conversions, make sure that you are carefully auditing these conversions. Remember, your advertising platform will not distinguish the difference between a “good” or "bad" conversion, so you need to make sure that you can limit the number of bad conversions that get counted as much as possible and take measures to prevent bad conversions from happening. Otherwise, you run the risk of encouraging your advertising platform to optimize for undesirable results. If you run into a situation where it takes a few days to audit conversions for quality, consider applying a delay in conversion importation so that you only report back the good conversions.
As you audit the conversion performance of your campaigns, it is imperative that you make a special effort to analyze the search terms that are correlating with conversions, and scrutinize how well they are working for you. In some cases, you might find that a keyword or search term that is leading to lots of unprofitable conversion events is in the mix, and without your intervention, the “smart” system will continue to devote more and more resources to making that event happen again and again.
For example, let's say you are working on a call-only campaign for a pizza shop called "Little Johns" that is generating a lot of phone calls from people checking in on their order from "Papa Johns." If you aren't auditing your search terms, or the calls being generated, those phone calls will be counted as conversions by Google, and, in turn, Google will optimize for those conversions and throw more resources towards clicks that will exhaust your budget and annoy your client.
Keeping an eye on search term performance is especially important for Google and Bing search advertisers because both platforms have become increasingly broad in their keyword matching strategies. The more restrictive “exact,” “phrase,” and “broad modified match” keywords are matching a wider array of queries than ever before. This shift in breadth exposes individual keywords to more users than they would show up for otherwise, but it also means that you need to be even more careful about semantic specifics in your search term audits. If you feel that the search terms aren’t matching up with your ad copy, change the ad copy, or use AdGroup level negatives to funnel your audience to the right ads.
Learning how to develop compelling ad copy for traditional advertising formats is a challenge that takes a lot of effort, tons of testing, and no small amount of poetic acumen; for that reason, it can be hard to embrace the idea of responsive search ads, which take the strategy of headline and description line juxtapositions out of our hands and instead automatically arrange these elements based on different queries and user signals.
But their ability to be organized based on user queries rooted in ever-broadening keyword matches also opens up the possibility that your ads will show up in front of users that you might not otherwise be able to reach. I will admit that I was highly resistant to adopting this ad format; however, their performance has swayed me. Upon examining the different combinations that Google was trying out, I found that their machine learning algorithm was trying out headline pairs that I never would have tried, as well as delivering conversions at a competitive rate.
However, responsive ads really only work well when they are created with the same kind of creative and strategic input as the traditional ad formats. Since any headline or description line created for the responsive can theoretically be paired with any other in the ad, you have to anticipate how a variety of juxtapositions could entice people who see your ads, and also work hard to prevent any one headline or description line from reading as redundant with another. For this reason, these ads pressure advertisers to make each and every unit say something bold and unique about their client or company and their product or service.
So, when using responsive search ads, give each description line a unique flavor, appealing to distinct facets of your product or service’s value, and write each one as if it could stand by itself. That way, if any two lines were to appear together, the ad would be coherent. If you think a particular pair of headlines would be redundant if they appeared next to one another, they probably are. If you are working on headlines, sketch out what would be the best 30 character phrases pushing each of these selling points: Products or Services, Product/Service Benefits, Branding Terms, Calls to Action, Information about Options, Prices, Promotions or Discounts.
If you’re finding yourself stuck, try out Google’s responsive search ad creator beta. In addition to presenting a clean interface for ad creation, the tool provides a “score” for your ad, based on the uniqueness and relevancy of the ad content. Moreover, the tool can provide helpful content suggestions based upon site content and extant ad copy and extensions.
Once you have created an excellent ad, give it some time to perform in the wild, and check out which combinations are showing the most. If you find that there is a combination that works especially well, create a traditional ad that mimics that combination and add it to the AdGroup so that you can compare data for that specific ad against your next one.
Though smart campaigns are becoming increasingly capable of delivering results for low-volume campaigns, it is simply not true that every campaign can be automated. In addition to circumstances where volume or budget are too low to feed Google a reliable signal, some campaigns face conversion tracking challenges that make it impossible to feed reliable data.
Even so, you can leverage insights from the automated campaigns to improve accounts that aren’t ready for smart bidding yet. Information that correlates audience data with conversion for one campaign, for example, might be worth trying for another. And you can also gather data about ad copy that works for your automated campaigns and apply them to others.