The Age of Automation allows us to decrease workload through the successful use of data. However, a fair amount of time and effort is required upfront to create the optimal automation process and to understand the data necessary to run processes smoothly and effectively. Here, we’re not just talking about supply chain or product development automation, but rather, the ever-growing realm of digital advertising and how machine learning impacts your campaigns.
So, what is machine learning? Google defines it as the time it takes “for Google Ads to gather the performance data it needs to optimize your bids.” Similarly, Facebook defines it as the time during which their “system starts learning who to show ads to.”
Machine Learning in Digital Advertising
Machine learning sounds like something pulled from a Philip K. Dick science fiction novel, but in reality, it’s something we work with every day at Market Mentors. From scanning a website to tracking leads and optimizing your digital campaigns, we utilize AI (artificial intelligence—or machine learning) and automation to achieve your marketing goals and enhance your ROI (return on investment).
Without machine learning, our strategists would be stuck reading through stacks of data sets to find the perfect keyword to bid for in your search campaign, and they could very well end up creating the next Wuphf.com. (We love the opportunity to work in a little humor from The Office!)
Instead, we do the work in the beginning. We use historical data, keyword forecasting from a variety of tools, and our strategists’ experience with prior campaigns to develop strategies to best meet your objectives. After we’ve built out your campaigns, ad groups, ads and landing pages, and tagged your campaigns for conversion tracking, we pass the baton to the platforms that employ machine learning to help us find the users who will find your product or service relevant.
Within approximately seven days, the machine (the ad delivery system, like Google or Facebook) learns about your campaign and the users who fit into your targeting. During the process, it is vital to keep your campaign changes to a minimum, especially when it comes to bid strategy, keywords and campaign structure. The more edits you make, the longer you extend the learning phase.
Relearning Also Affects Your Campaigns
This is also the same with relearning, which happens when the algorithm needs to readjust according to any significant changes you make to your campaigns. During this time, just like in the learning phase, you can expect your campaigns to deliver less efficiently than they did in the automated phase.
Changes that trigger the relearning period include:
- Changing a bid strategy or adjusting settings within the bid strategy.
- Revising conversion actions rather than adding more conversion actions.
- Changing your audience targeting beyond standard optimization.
- Significant changes to your creative.
- Pausing and reactivating campaigns after a week.
This isn’t meant to scare you from making changes when they are necessary, but it underscores the importance of goal setting and strategy building at the beginning of a campaign.
Preventing or Limiting the Relearning Period
Your campaign is less effective during relearning periods, so we want to limit the changes we make once the campaign is live. To do this, we recommend that you:
- Review your overall advertising goals upfront. If your goals and objectives have changed, we may need to change your strategy and campaigns.
- Avoid pausing campaigns for more than a week, especially ones that you want to reuse in the future. Instead of pausing campaigns, we recommend you keep them running at a smaller budget.
- Run campaigns based on product or service lines and change creative and keywords based on offers and timely topics to avoid building new campaigns.
- Avoid significant budget changes (larger than 20 percent).
- Avoid too many changes to your creative, messaging, ads or landing pages.
- Schedule out significant changes either on a monthly or quarterly basis to batch changes together and limit relearning frequency.
We’ll work with you and make suggestions and monitor settings throughout the life of the campaign. For example, we’ll check your settings to see if your conversion tracking is fully set up for the bid strategy you’re using, and review and optimize campaigns daily.
Don’t go it alone. Digital advertising is a science and you can trust our experts to help you build out your strategy and optimize your campaigns. Contact us today to get started.