Covid-19 has impacted life around the world with various shutdowns, stay-at-home orders and just generally rethinking your lifestyle. Businesses shut down, advertisers pulled back on spend and media consumption abruptly shifted.
With this being said, there were a number of advertisers that sought to push through these shutdowns, maintain top-of-mind awareness and push sales in new ways. With that, let’s take a look at how navigating the first couple of months through COVID-19 was.
Industry average and historic metrics don’t ever fully depict what your advertising campaigns expectations are. For instance, WordStream does a really good job of identifying the industry averages for Search, Social and Display campaigns and provides the following for expectations on costs:
We generally see our internal metrics for campaigns because our delivery metrics are skewed by the number of Pet Care advertising campaigns we run monthly, peripheral data usage for stronger optimization and an emphasis on unique creative that matches a local customer. For instance, our 2019 averages ebbed and flowed across the different quarters, and ranged far and wide based not only on industry but optimization tactic.
Historic averages constitute precedent which does not support the needs of “unprecedented” times. As such, it is important to look past “what [insert advertising metrics] are” and start thinking about “Why average metrics are”? Understanding the reasons behind what things are allows you to plan more effectively in new situations.
For instance, if you look at the average CPC on Google search and starting considering the why behind that, you get the following:
The average CPC for a Chrysler Dodge Jeep RAM dealer, focused on purchase intent keywords, is roughly $2.46 or $2.08, industry average and internal CPC respectively. Consider why. There is overlap of CDJR dealerships competing over the same customer. There are competing brands fighting for the same customer (F150 dealers pushing for RAM 1500 keywords), there are pre-owned dealers with FCA brands on their lot and then there is anyone else out there that may find those keywords relative to their business (and that’s not even considering “wreckless” Google Ads accounts that are bidding on things unrelated to their business and causing us grief). Google Ads, and most other digital placements are auction based ad deliveries. To get the better placement, you have to bid higher (we’re keeping this simple and not taking into consideration the effects of Quality Score). Everyone that is showing up for those keywords is inflating the auction price. This continues to build, and build and build, each preceding day building up the days in the future, which is why you see Google Ads CPC’s constantly increase year over year.
The following trends across search, display, video and social media platforms take into consideration trending performance figures over $2.5 million in managed advertising spend over a 16-month period. Some things to note:
Throughout 2019, we saw some very similar cyclical performance changes as we did in 2018 with cost metrics giving way to the changing of seasons:
Average 2019 CPC for Search and Social Media
A slight dip in January and February, followed by a slow summer and then constant increases through Q3 and abrupt increases throughout Q4. Throughout the year, paid search saw a 40% swing in terms of CPC where social media saw more than a 75% swing in CPC.
These follow global patterns of advertising trends.
Switching gears to averages in Programmatic Display CTR and percent change in cost (using CPM in this graph), a similar story was seen. Monthly CPM increases were noted for every month in 2019, except for August and Q4, where we actually had regression in CPM.
The difference in these metrics compared to Search and Social are due to the landscape of display advertising. There are a lot more available units in terms of available inventory and placements for display advertising.
Consider: There are generally only 3-4 paid search units that any company cares about for a select number of searches on just Google.com. For Programmatic Display there are 3-6 ad units per page across 95% of websites and apps out there.
In Q4, the increase in competition from other advertisers was far outweighed by the increase in visitors to websites, which inflated the amount of display auctions that were available. This reasoning is also justified by the change in CTR over the same period of time, with nearly a 25% decrease in Q4 vs. Q3.
Nearly every year, first quarter retains some of its December momentum with high search volumes, resembling retail sales. As January progresses, metrics do begin to decrease into February and March, making way for the Spring/Summer lows.
During 2020, there was a significant decrease in these numbers, which was significantly more than that of previous years. And, as these metrics moved into Q2, they didn’t begin to rebound. Throughout our advertising efforts, which stayed pretty consistent in terms of total spend, saw a stark decrease in average costs, with more than a 50% slide in Paid Search CPC and a 90% decrease in CPC in social media (comparing May through earlier highs).
The above graphs are a clear indication of two things: Reduced competition due to advertisers pulling out and increased inventory due to more people using Internet connected devices and properties.
The increase in inventory is certainly shown by the overall numbers of people using the web during this time, which has been illustrated as web based curiosity shopping, similar to “Window Shopping”. The basis of this is the bottoming out of conversion rate and the constant battle in search campaigns to negate “window shopping” type search queries, and focus on buyer intent.
Even with the reduced conversion rate, the faster paced reduction in CPC actually led to a significant decrease in Cost Per Conversion during the same period, meaning that advertisers that stayed on during this economic downturn, actually saw a significant overall increase in conversions. Year over year decrease in CPA for our advertisers was roughly 60%.
As a final thought, it’s important to not put too much stock in industry benchmarks. Too much focus on benchmarks can bread complacency, as you start to focus on the fact that “I’m already beating industry averages”, which even when it begins to trend down, continues as “Well, we’re down but still over industry averages”, and by the time you are below industry benchmarks, it’s too late to turn that negative trend around. Additionally, benchmarks are void of insights in unprecedented times, as they mostly rely on, well, precedent and at the end of the day, most metrics are relative so they need great context when analyzed.
Rather than keeping an eye on static benchmarks, keep your aims on positive trend lines over time. This allows for a continued eye on growth and allows for correcting momentary downturns.