The Stats: Conversion Rates, Goals & Assists
My personal favorite Wayne Gretzky statistic? If you take away all his goals and just count his assists, he would still have more total points in the NHL than any other player’s goals and assists. Combined.
It is worth noting, however, that a certain ageless wonder named Jaromir Jagr might just make this fact obsolete, depending on how much longer he can evade the chapped, grim clutches of Father Time. Especially remarkable, however, is that while highlighting the baffling number of career assists tallied by Jagr, the Great One remains the all-time leader in goals scored. It’s therefore safe to say that not all of his records are threatened by the mullet-icon.
The Conversion Controversy
In online advertising, we place a large focus on goals, or in other words, conversions. We establish what our client’s objective is, break it down into the identifiable actions which we want users to take, and set up markers along the way. These markers, we define as conversions, and we have spoken about tracking conversions in past blogs (and if you would like more information on this subject, feel free to check out our other blog posts). Beyond the subject of what constitutes a conversion, however, it is important to be aware of how they are attributed, or in other words, how their source can be traced. Traditionally, online advertising platforms employed what is known as “last click” attribution, which means a conversion is awarded to the last ad clicked or prior to the user completing a conversion, even in cases where the user clicked on multiple ads prior in a small time-window. This sufficed for a long time, but unfortunately, it is no longer adequate.
Google AdWords provides the ability to choose how conversions are attributed. Available models include:
- Last click – credit for the conversion goes entirely to the last-clicked ad
- First click – gives all credit for the conversion to the first-clicked ad
- Linear – distributes the credit for the conversion equally across all clicks on the path
- Time decay – gives more credit to clicks that happened closer in time to the conversion
- Position based – gives 40% of the credit to both the first-clicked and last-clicked ad, with the remaining 20% spread out evenly across the other clicked ads on the path
The Case for Linear?
Each of these models has its use, and the argument can be made for any, depending on the client’s business, specific objectives, and how their customers behave. For the majority of our client portfolio, we have migrated our conversion tracking to the “linear” attribution model.
This makes sense for many reasons, not the least of which being that it matches how most of our clients track and assign commissions within their own sales staff. For example, if a salesperson at a car dealership meets a brand-new prospective customer, conducts the entire sales process, and closes the sale directly, this one salesperson receives the entire commission. However, if the same salesperson teams up with another salesperson to complete the same transaction, the commission is typically split evenly between them. And in cases where three or more salespeople all claim some degree of responsibility for the sale, their manager most likely will attempt to determine whose involvement was legitimate and whose involvement was limited to merely directing the customer towards the washroom. Often, management will begrudgingly end up simply slicing it into a fraction for each. Fair’s fair, right?
The Path to Conversion
We have long observed that many of our AdWords campaigns achieve similar results for different clients. A branded Search campaign, for example, nets a large proportion of conversions because, obviously, if a customer is searching for the business by their official brand name, there is a good chance that the user is searching with intent to act. They have probably made up their mind where to conduct business rather than search for the product or service, being open to suggestion where to get it. But how did they make up their consumer’s mind? Did they search for a product or service, learn about it, search again for nearby businesses, research some more, then search for the one they like best and convert there? If so, shouldn’t all the prior ads with which they interacted receive partial credit for the conversion? With a “last-click” attribution model, all the hard work done to get the user to that point is essentially unacknowledged. It oversimplifies the complexity of the user’s path to conversion. Emphasis is placed on the branded campaign (which receives a misleadingly large proportion of conversions) while other campaigns lose the opportunity to add up the various partial conversions and show their own importance in the sales funnel.
Linear Spreads the Love
With a “linear” attribution model this issue is addressed, and we have been pleased with the results. Of course, many of the conversions tallied by a branded campaign exist without influence from other campaigns, but over time, the legitimacy of these other campaigns is more evident. If you’re a current client of ours, and have noticed that your monthly performance reports feature numbers with a decimal place and digits following rather than nice round numbers of conversions, that is the reason. Total conversions are not affected, but credit for the conversion is being shared where credit is due by being divided up – across different campaigns, ad groups, ads, and even keywords. This allows us to make better-informed decisions as to how significant an impact is represented by certain elements, regarding a given account.
Goals and assists are counted equally. Even people who argue that goals are more important wouldn’t argue that Wayne Gretzky is the greatest hockey player of all time. The total is ultimately what matters, regardless of how you slice it.