After AT&T ran the first online advertisement in 1994, a new marketing era was born. Digital advertising contrasted traditional advertising, like TV and print, which were extremely difficult to measure. Now, companies could hire science and marketing teams to monitor and drive digital marketing campaigns.
A wave of change has hit the advertising industry, yet again. This one is spurred by iterations of AI bidding practices of top advertising platforms, like Google and Meta, combined with new privacy laws like GDPR. Now, detailed tracking of marketing performance is no longer as relevant or as possible.
Instead, marketers face a black box that is difficult to analyze and influence. To overcome this challenge, digital champions rely on quality first party data that goes into the black box. Read on for best practices to conquer data-driven modern digital marketing.
Think long-term profitability vs. cheapest revenue
Traditionally, marketers granularly manage marketing accounts and use ROAS (return on ad spend) and other revenue-based metrics to measure the success and efficiency of campaigns. However, because efficiency does not always reflect profitability, more advanced metrics are critical, particularly with fewer campaign and bid management levers available due to Google automation.
Data activation, integrating first party information like margins and new customers, is the secret to measuring true profitability and informing bidding AIs to optimize toward it.
Use data activation to drive bottom-line growth
First, not all businesses have the same potential to improve profitability through data activation. A business with high margin variance will see a more significant impact when moving from revenue-based optimization to a margin-based model than a company with similar margins across categories.
We’ll go over how to activate five elements that can establish the potential a brand has to improve profitability and influence measurement and bid management, to maximize:
- First order profitability: Profitability when an order is placed (short-term profitability), influenced by margins, returns, attribution, and incrementally.
- Long-term profitability: Influenced by new customer acquisition and lifetime values.
First, evaluate net margins — the higher the variance, the more impact moving from a revenue-based to margin-based optimization strategy will have on your business. Consider the example:
You spend $50 advertising two products with 60% margins and generating revenue of $150 each, resulting in 3:1 ROAS. In other words, for $100 invested in advertising, you gain $300 in revenue or, a return on investment (ROI) of 80%. Both products are profitable.
Yet, if one product has a margin of just 30%, ROI would be negative, and you’d spend more than you earn.
By relying on measurements based on revenue only and training Google algorithms to only evaluate revenue, products appear to have similar value. Therefore, ad spend will be distributed equally among both products, and you’ll lose ad budget.
Instead, you can feed margins into Google’s algorithms so that it invests more in high-margin items, improving the profitability of your business. The more your margins vary, the more potential for uplift.
We tested this approach with a leading U.S. office supplier and saw huge improvements in profitability compared to their previous set-up.
2. Return rates
Similarly, high variance in return rates indicates that feeding return data back into your marketing channels will have a greater impact for your business.
Unlike margins, however, returns have a time lag after purchase. As a workaround, marketers can apply product or category-level averages so that bidding solutions retrospectively insert real return values as they become available.
3. New vs. existing customers (+ predictive CLV)
By differentiating between new and existing customers, you can estimate future margins a new customer can generate within a 12, 18, or 24 month period after calculating the lifetime value for each customer.
Depending on how aggressively a business wants to grow, this value can be assigned to new customer purchases across marketing channels, thereby increasing order value, and forcing bidding algorithms to invest in similar customers.
We recommend using predictive customer lifetime value (CLV), which uses indicators like the product, full price vs. sale price, and time of year to establish each new customer’s lifetime value.
Measuring new customer rate and applying lifetime values will help businesses single out the marketing channels that acquire more new customers vs. others. For instance:
PPC brand and paid social retargeting
While these often look great on the order level, both channels tend to have very low new customer rates, typically targeting users who are already familiar with the brand and intend to purchase it anyway.
PPC non-brand and paid social prospecting
Unlike PPC brand and paid social retargeting, these tend to contribute strongly to new customer acquisition.
Measuring ROAS with a last-click attribution model usually assigns too much value to converter channels with low new customer percentages, such as PPC brand.
Let’s consider the user journey for someone shopping for shoes online:
Typically they start googling “shoes online” on their way to work, which brings up ads from different retailers. The user clicks on several similar ads to browse online stores before buying. They may click on your result first, but move on to others before clicking on your ad again and choosing not to make an immediate purchase.
Later, at lunch on Instagram, they click on your business’s ad before returning to browsing. At home that evening, the shopper decides to purchase the shoes from your website, typing in your business’s name and clicking on a brand ad before making purchases.
With last-click attribution, the paid search trademark is assigned the search, even though it was the last step in the journey. In reality, the user likely would have purchased the shoes elsewhere if they hadn’t seen the non-branded campaign when first web surfing and hadn’t been reminded of your business during lunch.
Relying on the last click means that introducer channels receive less budget than converter channels, despite their significance in generating sales. This ad spend distribution decreases the percentage of new customers and even the overall number of orders while users purchase from competitors over time.
To discover specific business potential, compare changes between first and last-click attribution models in margin vs. new customer rates per channel. Again, the higher variance, the more potential to improve long-term profitability by changing your model.
This advice particularly applies to channels like PPC Brand, which tend to have very low incremental value — in 90% of cases, no competitors actively bid on brand terms meaning that after a brand ad appears in the search first position, it’s followed by a free, organic result.
So, if brand ads disappear, over 90% of users will click on the organic result. Conversions still occur but save ad budget. If brand ads aren’t incremental to revenue, why invest any money in this channel?
Conversely, channels exist that create a spillover effect, such as in the attribution example above. Without the initial non-brand ad, the brand conversion wouldn’t occur, so the incremental value to the business is higher than the revenue generated in the channel initially reflects.
Ready to use first party data like digital champions?
Conquer modern digital marketing with Crealytics performance marketing solutions, including our Performance Marketing Due Diligence offering.
*Blog written in collaboration with Emily Hunt