COVID-19 Paid Social Trends
This is part of a series of updates that provide data-driven insights and experience-based anecdotes that will hopefully help marketers and brands weather the COVID-19 storm. If you have feedback or questions, please send us a note at firstname.lastname@example.org.
The COVID-19 Landscape
The week of March 9 signaled the start of major impacts on paid social platforms due to COVID-19. Initial factors at play were two-fold:
1) with more time on users’ hands and a growing need for virtual connection, consumption of content on Facebook and Instagram began to soar,1 and
2) with an economic recession on the horizon and loss of retail revenue eminent, many paid social advertisers were forced to pull back on working dollars1.
This resulted in a surplus of available inventory and caused a month-over-month decline in CPMs for clients and campaigns across the board. Other metrics were also affected: engagement and click-through rates grew with app consumption which lead to a decrease in cost per engagement and cost per click.
As the COVID crisis continues, these trends are evolving week over week. As mentioned above, CPMs are tied to two external factors: increased consumption and decreased advertiser spending. Neither of these trends have reversed direction yet, as stay-at-home orders have been extended and many advertising budgets are tied to a calendar or fiscal year process. Notably, remarketing CPMs were still experiencing losses as of last week and may take longer to rebound (discussed in more detail below).
In the face of the COVID-19 challenges, many brands and destinations have turned to new customer acquisition, give-backs and promotional strategies on paid social as a way to effectively grow their purchase funnel by staying supportive and top of mind for customers. Backbone’s prospecting campaigns saw a notable increase in impressions starting mid-March. This spike also corresponded with normal seasonality and the launch of many already scheduled campaigns, in addition to COVID-19 reactions. While CPMs have seen an overall decline during the pandemic (-13%), rates have not consistently declined YTD.
As mentioned above, with the early drop in overall platform CPMs, many marketers shifted dollars away from low-funnel tactics to focus on new customer acquisition and positive brand association. Adjustments were quickly made to avoid sale and product-focused messaging in order to stay supportive and non-exploitative for customers. These adjustments actually exacerbated CPM trends further down in the funnel, where the decrease in advertiser demand for inventory caused rates to drop even more drastically. Remarketing CPM’s only started to pick up again this week, even as other metrics have rebounded closer to normal.
There is no denying that platform conversions have taken a hard hit due to decreases in consumer spending on non-essential items and, in some cases, closed distribution and production centers. While the decline was significant, results did start to rebound in April from their low point in March. We expect this is largely due to updated messaging and national adjustments to a new “normal.” With brick-and-mortar retail shut down but outdoor recreation still largely available at the local level, much of the demand in the outdoor industry moved online. While each brand’s situation is unique, on balance, we saw e-commerce performance of Backbone’s clients to be surprisingly adaptable to these external shocks of COVID-19 – with an average of 51% year-over-year increase in revenue for April so far.
The paid social environment is in a very different place than it was a month ago. While CPM declines are still being seen for remarketing, the shift in dollars to customer acquisition kept prices about flat for prospecting campaigns. While e-comm revenue took a hard hit in mid-March, we’ve seen positive progress in April with new tactics and adjusted consumer sentiments. One trend certainly hasn’t changed: the use of social platforms for digital connection, communication and entertainment remains high.