A Guide To The Adgo Algorithms And How They Work

We build programs and algorithms that automate optimization tasks and make your Facebook campaigns more effective by at least 20-30% based on what we’ve seen to date.

Why? Because Facebook is focused on optimizing its overall platform. This means they won’t always work in your best interests when determining what ads to show users or effectively optimize bids to get you the lowest CPC, CPM or CPA—this needs to be done by you, the advertiser. As soon as a campaign grows beyond a small scale however, doing this manually becomes an impossibility; taking a high level of skill and hundreds of working hours.

So we built the Adgo platform to do this for you—across hundreds of ads and ad sets at once—automatically.

This is how it works.

The Creative Selection Algorithm

Deliver the Right Message to the Right User

The first Adgo algorithm selects your best performing ad creatives to ensure your most effective messages are getting seen by target customers.

Does Facebook already do this?

Yes, but only at a superficial level. Facebook makes decisions on best performing ad creatives too quickly and then favours the ones that meet one specific goal. It then sticks with these creatives regardless of subsequent performance.

This is why you’ll see certain ad creatives accumulating thousands of impressions and others with only a couple of hundred, and when you see a drop off in performance of those ad creatives, nothing much changes. This isn’t smart marketing.

So how does Adgo fix this?

  1. We rotate through each of your ad creatives (however many you have) in groups, giving each one an equal chance of being shown, until statistically significant top performers are identified.

  2. We then turn off your lower performing ad creatives and only show the top performers.

  3. When we begin to see a drop off in performance of those ad creatives we then turn your lower performing ones back on.

  4. The process restarts automatically to continuously find a new top performer.

Icons used in this image were made by Freepik from www.flaticon.com 

Icons used in this image were made by Freepik from www.flaticon.com 

How do we measure top performers?

It’s up to you. The Adgo system allows you to select multiple goals to optimize towards; Purchase, Add To Cart, Clicks, CTR, Engagement (and many more, including your own Custom Conversions) and weight these in order of importance to you.

This means you’re not just basing the selection of top performing ad creatives on one goal, giving you a more nuanced approach to ad optimization with multiple data points.

How does this benefit you?

Ad creative is what gives life and meaning to the message behind your product or service. It’s what drives action. The slightest change in this creative can make the world of difference, influencing the decisions users make and impacting your ROAS.

Automating the testing and management of multiple variations of ads allows you to perform optimization tasks based on granular data that no individual or team would be able to achieve manually—quickly identifying and connecting your best performing messages to each target audience.

The Bid Optimization Algorithm

Optimize Spend for the Best Targets and Value Propositions

The second Adgo algorithm adjusts ad set bids to ensure your ads are being seen by the best performing target audiences.

Does Facebook already do this?

Nope. When you launch your campaign you can choose between automatic bidding and manual bidding at the ad set level. Manual bidding allows you to set your own maximum or average cost per click, impression, like, view or action. Auto bidding does this for you but not in a way that adjusts your bids favourably to get the lowest cost per result, or to target your highest performing audiences. It optimizes bids with the goal of spending your daily budget in its entirety—which is rarely the best course of action in most cases.

So how does Adgo fix this?

  1. When you launch your campaign we initially set bids to ensure your ad sets begin getting impressions, clicks and actions.

  2. We compare these costs with historical performance data to ensure that we never bid to high.

  3. We then increase the bids of the ad sets performing well and decrease the bids of the ad sets performing poorly.

  4. We continue to monitor the performance of each ad set ensuring every bid will be adjusted to your benefit if performance increases or decreases.

Icons used in this image were made by Freepik from www.flaticon.com

Icons used in this image were made by Freepik from www.flaticon.com

How do we measure top performers?

Just like with our Creative Selection Algorithm—it’s up to you. Our system allows you to weight multiple optimization goals in order of importance and the algorithm will optimize towards these selected goals.

How does this benefit you?

Focussing ad spend on your best performing target audiences is vital for any business; ensuring your message gets seen ahead of competitors and safeguarding against wasting money on audiences that aren’t performing.

Automating bidding across ad sets allows you to perform optimization tasks based on granular data that no individual or team would be able to achieve manually—ensuring you prioritize spend on areas where it’s most likely to generate a return.

The Brand Guard Algorithm

Protect Your Brand Integrity

The third Adgo algorithm measures the Reactions to your ad to ensure that only effective messages and content are delivered to the appropriate audience—protecting your brand integrity.

But doesn’t Facebook already do this?

Afraid not. Facebook optimizes ad delivery towards your best performing ad, which is dependent on what your one optimization goal is; impressions, reach, views, likes, link clicks, conversions. It has a negative feedback metric which is based on the number of times your ad was hidden by a user but it doesn’t measure negative reactions to ads, shift budget away from ads with negative feedback or provide you with a breakdown of reactions in the dashboard.

This leaves you vulnerable to running ads that may be generating negative Reactions and sentiment towards your brand, until you manually catch this and turn off the ad—by which time the damage is already done.

So how does Adgo fix this?

  1. We gather all the Reactions to your post—including Like, Love, Haha, Wow, Sad or Angry.
  2. We also measure the number of times your ad was hidden by a user and you as an advertiser were blocked by a user.
  3. We then turn off and flag ads that are generating a negative sentiment.
  4. Finally, we feed this into our algorithm and give you a detailed breakdown of all this data via your dashboard.

How does this benefit you?

Unknowingly running ads that produce a negative audience reaction can inflict damage on a brand that has a long and costly impact. But there’s no way to be 100% sure before you launch an ad that it won’t be received negatively.

Automating the monitoring of audience Reactions to your ad allows you to be sure that any ads eliciting a negative sentiment from your audience are quickly identified and turned off—ensuring you protect your brand integrity from any serious damage.

Where do we get the data?

We pull all this performance data directly from the Facebook API—but we also go one huge step further. We can help you configure Adgo to trigger actions based on external data that may be of importance to your business; meaning you can start, stop and adjust bids for ad sets based on what’s happening in the real-world, outside of Facebook.

We already have some external triggers set up, such as our Weather Optimizer. This enables you to optimize your bids based on regional weather conditions. Do certain ads or audiences perform better on sunny days? Our algorithm will work this out for you and optimize your bids to make the most of those conditions favourable to your campaign.

Stronger Together

Adgo’s Creative Selection and Bid Management algorithms work in tandem to ensure your most effective ad creatives are being shown while bids are being optimized for each audience you’re targeting across your ad sets.

The algorithms work at any scale—even if you have thousands of ads spread across hundreds of ad sets this optimization will be taking place in real-time.

Conclusion

Automating and optimizing these core tasks gives you insight, power and security that no team, however knowledgable, could achieve manually. If you want to take your social advertising program to the next level and be way ahead of your competition, intelligent automation is the only way to go.