The word optimize is used way too often. Like all buzzwords in the AdTech world, optimization has come to mean something that, counter-intuitively, doesn’t hold much weight anymore.
Every Tom, Dick and Harry claims to be able to optimize marketing solutions. Nowadays, optimization often refers to an option, as simple as an opt-in box to check on a screen that will somehow magically improve ad campaign performance.
In the spirit of correcting misunderstood marketing terms (see constraints), we’re here to set the record straight in regards to optimization.
The Scale of Optimization
The above graph shows how much more effective the results of search marketing “optimization” are once you move away from humans and towards machine learning, like Bid and Budget Management.
Despite the many qualities humans possess, optimizing digital ad campaigns is not one of them. A person can be smart, intuitive and game-changing, but a human is limited in the number of actions they can make, and humans famously make mistakes and forget. Humans can develop special tricks and analyze results well, but people struggle with recording, remembering and repeating successes reliably.
When a person sets automatic rules, they will run and ideally flourish with just a bit of maintenance that requires the campaign manager to check the rules to be convinced they are doing what they were told to do. But that maintenance does limit the scalability of rules.
When algorithms are introduced into the equation, that is when accounts can truly be optimized, especially with machine learning. A person is not meant to analyze large amounts of data, process them and programmatically run campaigns at scale, and “if then” rules fall short.
People vs. Algorithms
Marketers and marketing agencies still feel uncomfortable trusting their campaigns to an automated process rather than their instincts because they are unwilling to abandon their gut feeling and relinquish control.
Comparing humans to algorithms is like comparing a horse to a jet and calling them both modes of transportation. Yes, both a horse and a jet are forms of transport, but if the goal is to get from A to B and cover long distances as fast as possible, there is a clear winner.
True, a human can manage a handful of accounts reasonably well, but algorithms manage tens of thousands of campaigns, which is much more impressive.
For example, after reviewing more than 20,000 accounts over a one month period, with some accounts using our proprietary machine learning algorithms and some not, those using our optimization solution were 3 times more likely to spend the daily budget, without going over, and the cost-per-click was 40% lower than those accounts not using machine learning.
In order to get good results from your search marketing efforts, at a minimum you need to:
- review ads
- review landing pages
- update KPIs
- add new keywords and negative keywords
- pause underperforming keywords
- check budgets
- change bids
When you check budgets and change bids every day, there’s less time to focus on keyword performance and review.
With algorithms automatically respecting budgets and changing bids, and enabling easy campaigns review and analysis, you can devote your time to necessary keyword, ad and landing page improvements and adjustments. Without automation taking away most of the daily burden, it’s difficult for humans to “optimize” campaigns at all when there are so many updates to be made.
Just look at the results of an account, originally under human management, after implementing advanced optimization algorithms:
In just one week after launching bid and budget management, clicks more than doubled and the cost per click was reduced by more than 33 percent! Those results, that fast, are inconceivable without machine learning in place.
Algomize Instead of Optimize
Instead of optimize, let’s ditch the buzzword and call algorithmic optimization “Algomize”.
Algorithmic stock trading dominates the stock markets, so it’s about time algorithm-based bidding and budgeting dominates the digital marketing field.
In the past, automation has been sold as a magic black box, where advertisers are asked to trust that things will work out. But now with machine learning, it is possible to provide clear feedback that proves that the algorithms are reliably doing what they were programmed to accomplish.