Advertising can be a complex business
Knowing where to place your ad spend is not a new problem. Since the dawn of advertising, marketing departments have wanted to know how to get the best return on their budget.
Take this early 20th century quote that neatly sums up the uncertainties surrounding channel optimisation:
“Half the money I spend on advertising is wasted; the trouble is, I don’t know which half.”
It seems to make common sense. How can we tell which specific advert generated a certain lead? How do we know if the final touchpoint before conversion led to the sale? Before the digital era, these questions seemed unsolvable.
Fortunately, things have moved on since newspaper adverts and billboards were the best ways for companies to reach customers.
Now, we can use data to tell us which paid search ad a user clicked on, which Facebook ad first showed them your product and which Twitter post directed them to your business. Using a combination of these platforms and an holistic reporting dashboard, we can start to build a picture of which touchpoints are driving a return on our marketing investment.
Marketing knowledge and hard-earned experience are still the most important factors in any good ad campaign. But data and analytics can help answer questions that marketing departments have long been asking.
So which attribution model should I use?
Marketing attribution models are a data-driven method for assigning value to each touchpoint along the customer journey.
In simpler terms, it tells us which sales interaction worked hardest in selling your product or service. When we know this, we can start to allocate our resources more efficiently.
There are three basic models:
- Last-touch attribution
- First-touch attribution
- Multi-touch attribution
Underneath multi-touch attribution, there are four standard sub-models:
- Time-decay
- W-Shape
- U-Shape
- Linear
Last-touch Attribution
Imagine a football team. And imagine they’ve just scored a goal. Last touch attribution is the equivalent of awarding all credit to the striker – the person who scored the goal. On the face of it, this model seems to make the most sense. The ad that finally converts a user into a customer persuades a customer to buy your product is surely the one that matters, right?
This is the default model for many digital marketing platforms such as Google Analytics. It gives a clear, albeit simplistic view of which channels have driven a conversion:
As with all models, last-touch attribution does have its plus points:
- It’s easy. It’s easy to set up, easy to measure and easy to explain.
- It works well for very short customer journeys.
But there are also cons:
- For many brands, it’s too simplistic. It doesn’t reflect the reality of the customer journey.
- It can lead to a skewed view of your entire marketing strategy. Giving too much credit to the last touch will eventually lead to ignoring demand generation. Eventually, nothing will be coming through from the top of the funnel.
So while the striker might receive the applause, it was the entire team that got the ball in the right place for them to score and credit for these assists needs to be considered.
First-touch Attribution
As you may have guessed, on the other end of the scale, first-touch attribution places all of the value on the customer’s first point of interaction with a brand. If we go back to our football analogy,
a first-touch attribution model would give all the credit to the goalkeeper.
The goalie caught the ball, looked around for the best pass and started the attack. Without the goalkeeper making the initial move, the ball would have no chance of reaching its final destination.
In the first-touch attribution model, all value is assigned to a customer’s initial interaction with your brand. So if it was via a Facebook ad where they were first introduced to your brand, we’d place 100% of the credit there.
This model has some benefits:
- It shows how the early efforts of a campaign can lead to bottom of funnel conversion.
- It’s simpler to use and easier to set up.
- It’s great for marketers whose only focus is demand generation.
But this model has its drawbacks:
- It lacks nuance and doesn’t take into account the various other factors & touchpoints that have influenced your customers’ decisions.
- It can place much importance on the first touchpoint which leads to money being spent in the wrong place.
How do you think the rest of the team would feel if the goalkeeper got all the credit? Especially the striker, the player who ultimately puts the ball in the net?
Multi-touch Attribution
There are four standard multi-touch attribution models and they all emphasise the importance of different touchpoints along the customer journey.
But while they all recognise the importance of the whole funnel, each model assigns varying levels of credit to each interaction.
Importantly, multi-touch attribution models can give us a more accurate picture than the first-touch and last-touch ones.
Because as all proper fans know, football is a team sport and everyone plays their part in success.
The four most common multi-touch attribution models are…
Time-Decay
In Time-Decay, we assign a decreasing amount of credit to each touchpoint as we move further back in time from the point of conversion.
Simply put, the further away from conversion, the less important an interaction is.
It recognises that the initial advert your customer saw was influential in creating demand. And it takes into account that mid-funnel touchpoints helped push them further towards conversion.
But the idea behind time-decay attribution is that the final ad was the one that turned your prospect into a conversion. It was this interaction that did heavy lifting. As author and analyst Avinash Kaushik wrote: ‘If the touchpoints were magnificent, why did they not convert?’
W-Shape
In this model, we attribute credit to the beginning, middle and end of the marketing funnel. In theory, we assign value to the moments when someone first becomes aware of your product, when they become a viable lead and when they become a customer.
Think of it like this:
- Your initial ad on mobile piques someone’s interest.
- A few days later, they see the same product again on their social feed.
- A few days after that, after they’ve weighed up the pros and cons of buying the product that’s been following them around the internet, they search for your brand on Google and you are there at the top.
This model allows you to spread your resources over most of the customer journey. And if you think that multiple touchpoints along the way are essential to conversions for your brand, then it makes sense to give them more credit.
U-Shape
The U-Shape model gives equal weighting to the value of the first and final touchpoints. The interactions along the way are given some recognition, but less than those at the top and bottom of the funnel.
This model suggests that initial demand creation and the final touchpoint before conversion are where we should be assigning most credit for a conversion.
This model has its benefits. For shorter sales funnels, it accurately reflects reality. A person sees your product on their social feed. They look for it online and it comes out at the top of their search via a brand PPC ad. The U-Shape model reflects this scenario.
But it has its drawbacks. The longer a sales funnel grows, the more important the interactions in the middle become. This model fails to recognise these touchpoints and may lead to the misplacement of time, effort and money.
Linear
This is the last of the four standard multi-attribution models. And it’s very simple to understand and implement.
The Linear model gives equal credit to every touchpoint along the customer journey. From demand creation, to mid-funnel nudge, to final interaction before conversion.
It’s a great model to use on luxury products that have a slightly longer sales funnel, such as shoes, jewellery or a car. Customers buying products such as these often encounter many touchpoints before making a purchase.
Unfortunately, this model doesn’t attribute value to the most important interactions. This means we can’t tell which stage of the funnel was the most effective in driving conversion. So we aren’t necessarily placing our resources in the most efficient places.
Mixing experience with data
As you will have noticed, all of the models above have their own benefits and drawbacks. They can certainly all inform decisions about where to place your ad budget. Increasingly, advertisers are turning to data-driven attribution models that act as a hybrid of many of the above models. The advertising platforms attempt to fill in the blanks that exist when tracking across multiple channels to build a more accurate picture of which touchpoints are driving revenue and ROI.
Whilst these are more of a black-box in terms of how they work, our recent testing in platforms such as Google Ads have seen performance increases of 5 – 10% when switching to a data-driven model.
Ultimately, in order to decide the channels in which your advertising budget can be most effectively spent, you need to consider all of these models. It is crucial to take reporting from all available data sources into account, using a combination of all attribution models and looking at campaign performance at both a macro & micro level.
But most importantly, good marketers rely on their experience and knowledge, and will take into account the specific nuances of the business they’re working with. One size really doesn’t fit all when it comes to digital ad campaigns & attribution – because every brand and every product is different.
At Adbetter, we create completely bespoke reporting dashboards that pull data from all available sources and collate it into an easy to digest format, tracking the KPIs that matter to your business. By doing this for all our clients, we ensure that we take all attribution models into account and can zoom in & zoom out on your data as necessary whilst we are A/B testing, allowing us to establish the unique marketing channel mix that works for your particular brand or product.
Ultimately, attribution models and reporting dashboards are very powerful tools, not magic bullets. It’s a marketer’s job to interpret what the data is telling them, and implement a strategy that uses this effectively.
Need help establishing an effective attribution model for your business? Get in touch today to find out how we can help pinpoint the combination of channels & ads driving your revenue.
