The Importance of Conversion Architecture

The final entry in our Advertising 101 series is about the importance of conversion tracking to having success in media buying. If you missed it, you can start by reading the first entry, Which Audience Should You Target First? and the second one, Which Ad Channels Should You Start With?

The advertising landscape has evolved significantly over the years, and the ability to measure and optimize your ads has become a critical component of a well-functioning ad account. Previously, ad measurement was primarily important for media buyers to make optimizations after ads had already delivered. But now, having effective conversion tracking is absolutely necessary to feed data to the machine learning algorithms in order to optimize ad delivery in real time. In this post, we'll delve deeper into the importance of conversion architecture and provide insights to help marketers and business owners maximize their digital advertising efforts.

Machine Learning & Advertising:

Advertising platforms like Google and Facebook have adopted machine learning algorithms to optimize ad delivery. With these algorithms, humans have less direct control, as most of the ad delivery optimizations are handled by the machine learning system. This may sound complicated, but at the core it’s a fairly simple process. The advertiser needs to define a goal for the machine, then the machine will optimize ad delivery to hit that goal. For the machine learning algorithm to be successful, it requires a signal to optimize towards – this is where your conversion architecture strategy comes into play.

Conversion Architecture Strategy:

It is vital to carefully consider which conversions to track, how to track them, and how to assign value to them. You need to be strategic in this process, because–in a very tangible way–defining your conversions is in fact defining your campaign strategy. 

Machine learning algorithms can be quite literal in their optimizations, meaning they will optimize towards whatever you instruct them to, even if it leads to poor business outcomes. As such, it's crucial to choose the right actions to optimize towards.

For e-commerce businesses, optimization can be relatively straightforward, as the primary goal is to optimize towards purchases. It’s always good for people to give you money, so this is always a good optimization target.

Lead generation businesses, however, must be more cautious, as optimizing towards simple form submissions can lead to low-quality traffic. If all the machine has to do is find traffic on the internet that fills out forms, it can easily fall prey to spam and low-quality leads. This is why it's crucial to provide the machine with a signal of lead quality. This can be achieved by scoring leads based on the information provided in the form or even better by importing offline conversion data once your sales team has scored the lead or  the lead has reach a determined qualification level in the sales process. 

Going through the process of defining your conversion architecture requires strategic thinking about your marketing objectives, as well as technical solution development in order to make sure you can deliver an accurate signal to the machine learning platform. It can take some time to get this right but it is absolutely essential, or else you will send you machine learning off to spend your money going after audiences that are irrelevant to your business. 

The Learning Phase:

Once your ads are running, most platforms begin a "learning phase." During this phase, the machine learning algorithm works to optimize your campaign by identifying target audiences, assessing creative performance, and determining bid amounts for keywords, among other factors. Each platform has its own best practices, and it's crucial to follow them to ensure your campaigns learn and optimize effectively.

Allowing the learning phase to progress without excessive interference is essential. Be cautious not to make too many adjustments, as this can reset the machine learning and introduce new variables it must account for. It's also important to have patience and give the algorithm sufficient time to learn and optimize before expecting optimal performance. Monitoring the learning phase progress and understanding when to make adjustments is a key skill in digital advertising.

Reporting:

Understanding and implementing a solid conversion architecture strategy is essential to achieving success in digital advertising. By carefully selecting the conversions to track and optimizing towards the right actions, you can harness the power of machine learning algorithms to drive significant growth for your business.

As your campaigns run and the machine learning algorithm optimizes performance, it's essential to have comprehensive reporting that provides visibility into your ad performance. This enables you to make informed optimization decisions and evaluate your business and advertising strategies collectively.

Wrap up:

That completes the Advertising 101 series. I hope this has helped answer some of your initial questions about how to begin thinking about an advertising strategy and given you a framework for thinking about these things moving forward. There’s no one size fits all playbook for how to run your ads at this point. You must go through a consistent process of testing, reporting, strategizing, and optimizing to make sure your ads are performing well, your budget is optimized correctly, and your advertising campaigns are set up to achieve your business goals. 

At Data Raven, we specialize in helping businesses navigate these difficult questions and create advertising strategies that enable marketers to achieve their business goals. Reach out today for a free consultation if you’d like to chat more. We always love a good conversation.