Facebook advertising has positioned itself as a huge player for the 2014 holidays. Digital marketing agency 3Q Digital took a close look at the Black Friday-to-Cyber Monday performance of Facebook campaigns for two of its clients, a leading children’s educational toy subscription service and a leading online jewelry supplier.
Both clients ran substantial discounts in ads specifically mentioning Black Friday and Cyber Monday specials: for both days, the subscription service offered 40% off of the first month, and the jewelry supplier offered 50% off of everything in its catalog, and both featured direct calls to action (“Shop now”; “Join the movement”).
In focus: performance by targeting types, CTR (click-through rate) and CPC (cost per click) patterns, distribution of mobile vs. desktop traffic and costs, and trends for time-of-day performance.
For the subscription service campaigns, 3Q Digital found the following:
– The click-through rate on the Black Friday specific creative ads produced a 326% higher than average CTR in comparison to the current evergreen campaigns.
– CPCs followed this same pattern for the Black Friday-specific creative; they dropped almost 60% to $.38, compared to the account average of $.94 across devices.
– On both Black Friday and Cyber Monday, mobile drove more than 75% of the traffic, validating analysts’ predictions that mobile would continue to support growth and performance across all verticals.
– The Desktop News Feed ad placement, in the past referred to as the most valuable real estate in all of social advertising, produced CPCs 89% higher than the average on Black Friday and 71% higher than average on Cyber Monday versus mobile. This was in line with the expected increases in cost resulting from the holidays and increased competition on the platform.
3Q Digital also analyzed the performance for lookalike audiences across devices, utilizing the full range of percentages (1-10%, with 1% being the most relevant or similar to the seed custom audience and the 10% audience less relevant but at a far larger scale). The findings:
– CPCs were roughly consistent, averaging $.78 with little fluctuation across the 1-10% audiences.
– The CPA (cost per acquisition) of the 10% audience was $43, almost double the CPA of the much smaller and more relevant 1% audience. The CPA of the 5% audience, surprisingly, nearly mirrored the 1% audience’s CPA.
– The 4% audience posted the lowest CTR, approximately 41% lower than the 10% audience.
For the leading jewelry supplier, the holiday Facebook campaigns revealed patterns in time-of-day performance and revenue split between targeting types, including the Website Custom Audiences feature released in January 2014. The results:
– Facebook appears to be following similar “prime time” patterns as its offline cousins. On Black Friday, 10am PT was the highest revenue-generating hour, driving nearly 50% more revenue than the early-afternoon hours immediately following lunch.
– The next “money-making” hour occurred in the early evening; the 6pm PT hour posted 12% of the overall revenue generated that day.
– On Cyber Monday, the clear “winning” hour was 5pm PT, accounting for slightly over 20% of the revenue that day as consumers shopped around daily work hours.
– On Black Friday, the lookalike audiences combined to account for 8% of revenue. The other 92% of revenue came from retargeting. Of the revenue produced by retargeting, 44% came from fans, 46% came from Website Custom Audiences, and 10% came from existing customers.
– The allocations are slightly different for Cyber Monday, with 14% of revenue generated from lookalikes and 86% from retargeting. For Cyber Monday, Website Custom Audiences accounted for 51% of overall revenue.
Desktop News Feed ads proved costlier with increased competition, even as mobile accounted for the majority of traffic. Tailored and holiday focused ads with a clear call to action, combined with Website Custom Audience targeting, produced tremendous performance for the jewelry supplier’s “burst”-style sale. And campaigns for both clients showed the challenge of scaling performance across audience segmentations of different sizes.