In our previous article, we examined how financial services can ensure they are making the most of their customer data by effectively capturing and integrating it to form a cohesive customer view. But just having your data integrated does not mean it is actionable, or telling you anything useful. For this to be the case, businesses need to turn to attribution.
What is attribution?
Attribution is the process of “accurately assigning value to each digital marketing touchpoint across the complete user journey, providing a great understanding of what combination of events drove conversions" (Fospha, 2017). Marketers are becoming ever more reliant on attribution due to the increasing complexity of the customer journey – which now spans multiple, channels, devices and sessions. Without an attribution model in place, businesses find it hard to understand which of their multiple channels are responsible for a customer conversion, and therefore where they should be investing their marketing budget. And, as a result of this misunderstanding, the cost of customer acquisition in the financial services industry is around $303 (Hubspot’s State of Inbound Marketing Research Report).
Typical attribution models
Many businesses believe they are accurately attributing value to their marketing channels, but this is not always the case – rather, they are using out-of-the-box attribution models that rely on arbitrary metrics to assign value, such as:
Last click: 100% of the conversion is attributed to the last step in the customer journey.
First click: 100% of the conversion is attributed to the first step in the customer journey.
Linear: Each step in the customer journey is attributed equal value.
Position Based: The first and last step in the customer journey are credited a higher proportion of the conversion i.e. 40%, whilst the middle steps share the remaining value i.e. 20% each.
Time Decay: The last step in the customer journey is attributed the largest amount of value, with the steps further back in the journey are assigned diminishing value.
A further limitation of these models lies in the fact that they often underestimate the value of generic keywords, or social channels, in the path to conversion. As a result, marketers struggle to get an adequate budget for these activities – despite the fact they may be leading to customer conversions.
Indeed, in a study by Bizible (2017), it was found that 36% of finance and banking marketers say they aren’t using the right attribution model to measure their marketing performance. In a further study by Jawing (2017), only 20% of respondents used an advanced attribution technique.
So how can businesses ensure they have an accurate understanding of their marketing spend?
The answer lies in data-driven, multi-channel attribution. Unlike aforementioned out-of-the-box models, these intelligent models take into account all historical visitor journeys in order to assign an accurate value and cost of each step in any customer journey that leads to a conversion. In doing so they provide a true reflection of your costs and revenue across multiple channels, rather than relying on a proxy measure on how these channels deliver conversions. For more information on data-driven attribution, read this blog.
Accurate attribution is also contingent upon cohesive, granular customer data – and this is where our previous piece comes into play. If businesses haven’t taken the necessary steps to integrate their customer data in a Customer Data Platform, they will be feeding inaccurate data into their attribution model. No matter how good their modelling is, if the data source is not rich and granular, insights will continue to be inaccurate.
Below are three reasons that make data-driven attribution key for financial services marketers wishing to optimise their marketing channel spend:
Analyse every step
Data-driven attribution models ingest every step of a customers’ journey – across multiple devices and channels. It then employs intelligent machine learning to analyse these step, taking into account the specifics of a businesses’ marketing campaigns, customers, brands and customer journeys, as illustrated in Figure 1. This provides a granular view of how each marketing channel contributes to a customers’ path to purchase, enabling marketers to gain a deeper understanding of the specifics steps their customers took to reach their website and convert– such as whether it was straight from an email newsletter, or from an affiliate site and then back through a paid keyword.
Figure 1: Data-driven attribution modelling analysing each step in a complex customer journey
Assign values between channels
Once visibility of the entire customer journey has been harnessed, a data-driven attribution model then assigns a value to each of these conversions. These models differ drastically from out-of-the-box models because they do not automatically assign value to any one step – instead, they calculate a unique value for every stage in every customer journey, based on previous and future steps, as illustrated in Figure 2. As a result the value is inherently more accurate than alternate models. With these insights, marketers can understand the real value and cost associated with each of their marketing touchpoints. From here, they can identify where marketing activity plays little to no role in driving conversions, see any inefficient and underused channels, and redistribute their spend accordingly. For instance, if affiliate channels are being over invested in marketers can reallocate spend to a channel that results in conversions, such as PPC.
Figure 2: Modelling calculating the value of each stage in the customer journey
Identify value within channels
Finally, the key to effective attribution, and to lowering the cost of customer acquisition in the financial services industry, is to not only optimise between channels, but to also optimise within channels. For instance, knowing that paid search channels results in more customer conversions than affiliate websites helps marketers to redistribute spend to this more successful channel. However, with the average cost per click on Google being $3.72 in the financial services industry – and only $0.88 in the e-commerce industry – businesses need to be able to lower costs within a channel too.
A data-driven attribution model enables marketers to understand their marketing spend in the most granular detail – down to specific keyword and ad level. This provides visibility on which keywords lead to conversions, and which are being clicked on but bring in no ROI. With these insights, you can save on PPC spend by redistributing your budget to drive conversions, and increase ROI. This level of granularity cannot be found in standard models, but is evidently key for complete marketing channel optimisation.
Attribution plays a key role in financial services being able to understand their marketing channel spend. And with the increasing complexity of the customer journey, this has never been more important. The key lies in using an effective attribution model – without this, businesses will never see the true value of their activities, and risk increasing their cost of customer acquisition and decreasing their ROI.
In our final piece, we will examine how businesses can operationalise these insights; through advanced machine learning and integrations with marketing systems, financial services marketers can take their attribution insights and scale them in real-time.