How you calculate the value of a conversion can have big implications for your marketing budget and your company’s bottom line. Many marketers optimize their search campaigns, for example, by looking at the cost per conversion of a given keyword. For display or video campaigns, analyzing performance by the conversion cost per ad or ad type is common.
But what happens if the person that clicked on a Facebook ad then clicked on a search ad a few days later? How much value would you attribute to each source?
You could assign 100 percent of a sale to the source that the user touched last or first, or use an equal split approach. Or you may have a marketing automation tool that keeps track of user behavior, which can also be used for custom attribution.
But attribution does not take a customer’s conversion value into account. For example, a website may sell vases that range in price from $15 to $1,500. A consumer could see an ad and purchase a $30 vase. Standard conversion tracking would assign the value of this conversion at $30.
But that could be wrong.
Consider this scenario. A few months later, the same person that purchased the $30 vase could move to a new house and need décor items. So she might purchase another vase for $250 at the same store. Therefore, that conversion of $30 has now turned into $280 ($30 + $250).
Thus the question is how to calculate the true conversion value. There are two approaches, depending on your business.
- The traditional method is to calculate the conversion value — also called “lifetime value” — by summing up all purchases that a customer has made or will make. To be more predictable, you could cut it off after a year or two. After summing up all sales for all customers, your next step is to take the average and use it as the lifetime value of your entire store. This traditional way of calculating the lifetime value of a customer is helpful when all your products are similar in value and price. For example, if you sell socks that are all within a dollar or two in price, this is the simplest way to calculate conversion value.
- The segmentation method groups customers based on different attributes, such as price points. This works well for stores that sell high-end and low-end products. For example, someone who purchased a shirt for $59.99 likely has a different conversion value than the buyer of one for $199.99.
Another segmentation strategy is to use the first item purchased and group the customers who bought that item and assign their lifetime value on their subsequent purchases.
Regardless of how you want to segment your list to calculate the value of conversion, the key is to make sure the calculation is consistent.
To compare the methods, let’s examine the following scenario.
A gourmet grocery store sells cheeses and other food products. Customer A purchased a $5.99 jar of jam. Customer B purchased a $49.99 assortment of cheeses. Taking all customers into account, the grocery store computes an average overall conversion value of $59.99.
- The traditional method would use the average lifetime value of all customers and assign $59.99 — for both customer A and customer B.
- Using the segmentation method, the store would assign customer A the value of $79.99, as it knows the typical, subsequent purchases by other customers who started their journey with the same product. And the store would assign customer B a value of $49.99 as it knows customers that purchase $49.99 cheese assortments are typically one-time buyers.
Assume the gourmet grocery store spent $29.99 to acquire customer A and $39.99 on customer B. What is the return on investment for each approach?
- Traditional method. Using the traditional approach, the grocery store assigns a customer value of $59.99 for both customers A and B. The acquisition cost for customer A was $29.99, for a profit of $30. The cost to acquire customer B was $39.99, which represents a profit of $20.
- Segmentation method. Using the segmentation approach, customer A has a value of $79.99, which represents a profit of $50. Customer B has a value of $49.99, for a profit of $10.
In other words, the two methods produce different returns on investment!
Having a more accurate approach to conversion tracking can help with your digital marketing and attribution models. What looked like an unprofitable keyword or channel could be very profitable, and vice versa. Test your attribution model periodically — every quarter or every year — to ensure that your original model is still correct. Do not be surprised if you need to keep tweaking it, especially with new product offerings and different marketing initiatives.