See Beyond the Click
Why Your Attribution Model Should Give Credit Where It’s Due
Aug 4, 2016
Think about the last time you bought a car. Maybe you saw some TV commercials, researched online, clicked some online ads, asked friends their opinions – and after weighing your options you finally went down to the dealership and drove your brand new car off the lot. So why did you decide to make your purchase? Was there only one factor that influenced your decision? Or was it a combination of all those elements that eventually convinced you to pull the trigger?
This is the core of the debate when it comes to attribution in digital advertising. Attribution – the act of giving credit for a conversion to a digital ad or channel – is a hotly debated subject, and for good reason. There’s no set gold standard for attributing credit for conversions; the customer journey is far too complicated and involves too many variables to lock in a reliable, foolproof standard.
That’s not to say there haven’t been attempts at creating accurate attribution models, but at the end of the day they are still educated guesses (fueled by data and research, but guesses none the less). No customer is the same and each is in a separate moment in their purchasing journey, meaning different touchpoints have different impacts depending on who is interacting with them. With different touchpoints having different weight for every customer, it’s difficult to lock in a foolproof attribution model.
But that doesn’t mean all attribution models are created equal – they may all not be perfect, but some are clearly better than others when it comes to giving credit where credit is due.
Plenty of attribution models have been created over the years, ranging from multi touch attribution (where credit is spread between multiple touchpoints) to single click attribution (all credit is given to a single click in the customer journey). Here is a quick primer on some of the most commonly used.
These all-or-nothing models favor a single action in the customer’s journey.
– First Touch | In this model, the first view or click on an ad gets 100% of the credit for the conversion.
– Last Touch | This model gives the last view or click on an ad 100% of the credit for the conversion.
These models favor giving credit to touchpoints throughout the journey, and recognize the contributions of each.
– Algorithmic | A percentage of credit is given to each touchpoint based on where it is located in the sales funnel.
– Linear | This model spreads credit out to all touchpoints evenly, giving each one equal credit.
– Time Decay | This one spreads credit out, but not evenly. The closest touchpoint to the actual conversion gets the largest percentage of attribution.
CLICKS ONLY GIVE YOU PART OF THE STORY
When it comes to attribution, there is a lot of faith put in clicks – but they don’t tell the full story. Go back to the car analogy – you wouldn’t give 100% of the credit to the dealership, just as you shouldn’t give total credit to the click in a customer’s purchasing journey. There was a lot of exposure to ads for that car, even if they did not drive immediate action. Only by conducting an honest evaluation of the customers’ exposure to your ads can you accurately attribute proper credit.
The truth is advertisers are demanding their ads drive customers to an eventual purchase, and that doesn’t necessarily involve a click. In research exploring the effectiveness of display ads, Mediative found that display ads are 39% more likely to influence web users when they are researching a purchase versus not looking to buy anything and are just browsing, which shows that merely seeing an ad can have a strong influence on a consumer and their decision to buy.
It’s important to remember that clicks aren’t the only action a customer can take after seeing your ads. They can punch in your company’s URL and go directly to your site, conduct a Google search for your company product, or head straight to an ecommerce site to do more research. Because users all act differently, it’s important to have an attribution model that takes as much of that behavior into account as possible – for example, tracking impressions instead of clicks can give you a sense of whether a customer saw your ad before eventually converting.
Which is why a single touch attribution model like last touch doesn’t make much sense for tracking performance. It gives a false impression on how effective an ad is – it’s essentially saying “don’t bother with other ads or channels – this one ad is what did all the work and deserves all the credit.” That’s an absurd statement to make, and it should show you how using a model like last touch when running display campaigns is not a good idea.
VIEW-THROUGHS GIVE YOU THE BIGGER PICTURE
Clicks only give you a sliver of insight, which can be frustrating for any marketer trying to determine which ads are drawing customers’ attention. View-throughs, on the other hand, provide a bit more leeway when it comes to tracking customer actions, thus giving more insight into the customer journey. Using view-throughs in a multi-touch model makes sense because it allows you to measure the effectiveness of ads that may not be generating clicks, but are still driving performance.
View-throughs are different from clicks in that they don’t track immediate user action. Instead, they measure a specific amount of time where a customer can take another action (outside of clicking an ad) that will result in a conversion. For example, if a customer sees an ad and then Googles the product or brand name which then leads to a sale, then that would be considered a view-through.
By tracking view-throughs, advertisers can take these types of scenarios into account. By going beyond the click and expanding how you track conversions, you give yourself more information on how customers are interacting with your marketing efforts. Armed with that, you can make more informed decisions on how to steer your campaigns. View-throughs go much further than tracking clicks when it comes to accounting for the complete customer experience that leads to a conversion. And while they don’t give you the much sought-after perfect attribution model (they do have their own limitations such as not giving 100% certainty that your ad was actually responsible for the sale), they are a step in the right direction.
Advertisers also need to know which touchpoints are having an impact so they spend their budget wisely on channels that are actually influencing customer behavior and driving sales.
To better understand how each touchpoint plays its own important role in leading to a conversion, let’s take a look at the sports world. Google does an excellent job of explaining the importance of multiple touchpoints, and to borrow their specific analogy, we’ll use a basketball team as an example. To win games, you need a full team of players to move the ball down the court and score points. You not only need someone who can put the ball through the net, you also need players who can feed that person the ball, play defense, and set up big plays. Each player is valuable and contributes to the team, even if they aren’t personally scoring points.
The same can be said of online advertising – multiple touchpoints help move the ball down the court, and get your customer to convert. And while they aren’t all getting the glory of getting the last click, you cannot underestimate their value. If you aren’t recognizing the value an ad is providing, you’re less likely to put budget behind it – and that can hurt the bottom line.
CREDIT WHERE IT’S DUE
Attribution is incredibly important to ad tech companies – it determines whether they get paid or not. But a sound attribution model matters to advertisers too because it shows which ads and channels are performing and are worthy of more budget.
Clicks alone just aren’t enough – yes they tell advertisers when the customer finally decided to click the ad, but they fail to take into account the value a view-through provides in tracking other actions. Advertisers must take into account there is more than just the click, and by attributing credit where credit is due, advertisers can be smart with their budgets and spend money where it matters most.