When it comes to media buying, there is a general optimization cycle that the activity goes through.
The first step is to configure the campaign. Once the campaign is active, the budget is spent, and there is activity, reports are generated and analyzed, and the campaigns are once again configured and optimized.
This cycle can be done in three ways:
Manual Optimization Cycle
This method is actually the most common, albeit the least efficient. It involves going through the entire cycle manually, which means manual reports, analysis, and configuration for every single campaign. This is extremely time consuming, open to human error, and not very effective. While it is the most common method today, we believe it will slowly fade out as technology evolves and IoTs and omnichannel targeting increase, as it is becoming impossible to properly manage it all manually.
Build Your Own Algorithm – Outside
Not everyone takes the manual approach. Some companies realized that a large part of the optimization cycle can be automated, and created their own algorithms to do the work for them where possible, for example reports, analyses, and even bid adjustments. This is done by creating an algorithm that is not part of the bidding system, but connects through an API with its own rule and commands on how and when to adjust the budget and bids, create reports, etc.
This method isn’t perfect. Because the algorithm is outside of the system, it works on a periodic basis, not real time. When you build your algorithm, you define how often you want it to analyze results and adjust the campaign, for example every hour or every three hours. It works at a campaign level, not a bid level, so it provides more control and is more effective than manual optimization, but it is still not ideal.
Build Your Own Algorithm – Inside
By building your own algorithm and uploading it to the system, it can work within the system in real-time at much higher resolutions – I’m talking impressions and bids, not just campaigns. It also means you can build your logic on historical and aggregated data, which means making bid decisions that will increase your chances of showing your ads to the right people at the right time. For each user, you can analyze their usage data to see when they are most active, the probability that they will act, and more to choose the best bid based on your goals. This method allows you to benefit from a more efficient optimization cycle, more control, and much higher optimization.
Spotad allows you to build your own algorithm and plug it into the bidding system for real-time bid-level optimization.