blog | 4 min read

Power Up Your Attribution with Predictive CLV

March 1, 2020

Image Displaying mathematical graphic.

Most marketing analytics teams know that attribution is important, and they also know that customer lifetime value (CLV) is important. But it’s far less common to find teams that consider these two together and blend them into an integrated approach.

Without considering CLV as part of your attribution measurement, you could be focusing your efforts on the wrong channels, resulting in short term return on ad spend at the expense of choosing the best channels for building value over time.

In last week’s blog on attribution, we wrote about the importance of a solid data foundation for greater accuracy in tracking which marketing dollars connect to which customer. This week we take the next step: once you have a clear view of each individual customer, you can start to think about attribution more strategically and with a longer view.

Conversions Don’t Tell the Whole Story

No matter what attribution method you use, market mix modeling (MMM) or multi-touch attribution (MTA) or a combination of both, the default measurement is conversions. This is a sensible place to start when looking at which marketing actions led to a customer making a purchase.

The problem arises when companies treat conversions as a standard unit, all equally valuable. This doesn’t account for the fact that different customers spend different amounts with a brand over time. In other words, a conversion may be an endpoint to a sale, but it’s the beginning of a relationship with your brand. And naturally, some relationships are longer and more fruitful than others.

The top 10% of your customers often represent as much as 50% of your revenue, so focusing your efforts on these customers represents a huge opportunity. On the other hand, not paying attention to CLV in your attribution may mean you’re missing out.

For example, you may have a high volume of conversions from search and a lower volume through social. If you were only measuring by conversions, you would conclude that search is your top-performing channel.

But if you assign predictive CLV to each conversion, you might find that the customers who came in through search tend to be one-time shoppers or customers who only purchase at discount. The customers you got through social, on the other hand, are likely to stay with your brand for longer and provide better returns for the length of their relationship with you.

Sounds Great, But How Do I Make it Work?

In order to take a long view of attribution, you need an individual level CLV score and the ability to tie each user to a given stream of engagement.

This isn’t always easy, which is why it isn’t always done. CLV takes calculation, and effective predictive CLV requires advanced mathematics scaled across all your customer interactions. Fortunately, there are now tech tools that lighten the lift and spare you the heavy number crunching. And like the clarity and accuracy in attribution measurement we discussed last week, the first step in making this possible is a clean, solid data foundation.

Once you have the CLV score infrastructure in place you need to get access to the channel data you’re going to use. For MMM, this means share of voice by market and by channel. For MTA, this means clickstream or path-to-purchase data, available either by collecting it yourself or working with an MTA partner.

Then you link CLV data with market and channel level data. For MMM this involves CLV of customers by market or trade area; for MTA it’s the CLV of customers by path-to-purchase.

Finally, decide how you want to allocate CLV across the purchase journey. Different brands have different approaches. If the intent is to use all paid media as an acquisition vehicle, then all predicted value will be attributed to the first purchase. Other brands might favor an approach where only a customer’s predicted future “owned channel” spend is attributed, and “earned channel” purchases are attributed against the channel when they occur.

And then… watch the magic start!

Getting your data straight and integrating CLV into your measurement lets your attribution strategy contribute to your long term business goals rather than just the next campaign. That way you can truly understand which of your channels are most effective at bringing in high-value customers and track the effectiveness of your marketing on your customers throughout their whole relationship with your brand.