More than half of all marketers plan to adopt AI over the new two-year period. As they begin to experiment with ways to put AI to use, the top applications, per eMarketer, are largely display advertising-oriented. Programmatic is probably the best example. But what about search, one of the largest investment areas in digital marketing today? How might AI be incorporated there?
Search is one of the largest investments in digital advertising, with roughly $30 billion spent annually. And, globally, search spend is now more than $90 billion – and climbing. It’s also the online channel used by most marketers (86%) for customer acquisition. As spend rises, marketers must optimize investments and costs, while improving ROI. For this to occur, advertisers need “smarter” and more informed search strategies. They need to adopt newly available technologies to do this or will get left behind. AI and machine learning can deliver that in a number of ways, yet adoption is still in its infancy. That needs to change.
Here are just three ways AI can help optimize search for advertisers.
1. More auction transparency.
Google AdWords is a blind auction. As a result, it’s virtually impossible to know what competitors are doing and to understand the root causes of fluctuations, such as increases in cost-per-click (CPC) or falling click-through rates (CTR). For marketers, this “black box” is a big challenge. They need to optimize and streamline operations to gain market share. Optimizing campaigns means monitoring competition performance closely. Without knowing why a competitor’s campaign performance might change over time, the temptation is to increase spend – with no guarantee of success.
AI and machine learning have the potential to illuminate AdWords’ inner mechanics, shedding light on and predicting competitor behavior. AI can be used, for example, to continuously track and capture paid and organic search data through Google, without direct access to AdWords auctions. The technology can home in on search engine results pages (SERP), indexing them for data on competitor preferred keywords and phrases. Then, machine learning can collate and interpret the findings to identify and predict budgets, buying strategies, uncover profitable keywords, and more.
This kind of intelligence can’t be developed without AI and machine learning capabilities today. It’s impossible to process, interpret and action the data, given its volume and scale, without an AI-driven solution. So more search marketers are investing in them to plug the AdWords gap.
2. Protecting brands and reducing costs.
Online and offline, brand infringement is a costly risk facing all brands today. Traditionally, it occurs when competitors, affiliates or other third-party advertisers abuse a brand (e.g., trademarks, keywords, ideas, and products) by co-opting aspects of that brand as their own. In search, brand infringement means searching for branded or trademarked keywords, only to be presented with ads from the brand’s competitor.
Competitors bid on these terms, “piggybacking” off of their value. (This is similar to “conquesting” in display.) And the risk here is serious. Reputation and brand integrity is only part of it. For some companies, brand infringements can mean hundreds of millions in revenue lost annually, lost campaign clicks and higher CPCs.
Unfortunately, traditional search tools provide few -- if any -- safeguards for brand infringement. However, AI and machine learning technology is poised to dramatically improve infringement identification and evaluation.
These tools allow for more real-time tracking and analysis of competitor bidding strategies. AI can tell a marketer what the infringements are, with machine learning estimating the impact on performance and optimization. Studies indicate that this could cut costs by more than 50%.
3. True ad copy optimization.
Using AI in search to understand a competitor’s bidding strategies is not enough to be successful. Equally important are deeper insights on ad copy and campaign language -- understanding the best-performing ads in the market, as well as the worst. What calls-to-action worked best in an ad, for example? Or what themes are being over or underutilized? How is overall ad copy evolving based on the season? Changing even a single word in an ad’s copy can have a huge impact on end-user performance.
The problem, however, is knowing where to start. Being able to analyze and identify broader trends across paid search ads to benchmark campaigns and fine-tune performance is unmanageable without a tool that can index, track and evaluate. Human analysis isn’t enough here.
AI and machine learning deliver unique and unmatched opportunity on market and campaign optimization. The natural language processing (NLP) capabilities of these tools can enable analysis of ad copy at scale in a way that wasn’t available previously, while tying identified patterns to performance and attribution. This is crucial intelligence. It allows search marketers and brands to be more data-driven in their approach to campaign testing. And even the most modest improvements in copy can deliver higher CTRs and lower CPCs.
AI and machine learning have been conversation centerpieces in digital advertising for years now. However, most discussions and applications have focused on the display side.
As search spend rises, the category is ripe for change and innovation through artificial intelligence. Candidly, we have only scratched the surface here. This is largely because the technology has needed time to develop.
But, from auction transparency to brand protection or ad copy optimization, the technology is now at the point to provide impressive impact in both the immediate and the long-term. Those who fail to adopt new search technologies that use machine learning will be outpaced by their competitors who have.