Cost per Purchase up 20%, as low conversion combines with higher CPM’s – Our Strategic Facebook Recommendations! (Week 7)

May 1, 2020 | Industry News

The new reality of marketing under COVID-19 has entered week seven. It’s time for our seventh edition of “Latest Pricing Trends and Strategic Recommendations on Facebook During  COVID-19”

Executive Summary (Full Details Below)

This week we are seeing that Facebook and users are getting accustomed to the current advertising trends, with many of the metrics continuing similar trends to the prior week. Our data shows a rise in CPC, driven by a steady 5% per week rise in CPMs (and not offset by a rise in CTR). We recommend that our clients stay the course this week with Facebook advertising strategies (and you will notice that  our “suggested strategies” are largely unchanged from last week). 

You can sign up for future updates here.

Summary of Changes

  • CPM rising 5% per week, continuing our weekly trend.
  • CTR staying flat, as customer interest is steady.
  • CPC rising 5% per week, as cpm is rising on flat CTR.
  • Conversion rates are drifting downwards.
  • Cost per Purchase up 20%, as low conversion combines with higher CPM’s

What Does it Mean?

  • CPM has bottomed and is now rising steadily. People are slightly decreasing the amount of time they spend on Facebook, and ad budgets for brands are stabilizing-to-increasing.
  • The “Prospecting Honeymoon” is over. In late March, prospecting got inexplicably cheap but is now returning to a more normal level making it less of a bargain.

What Will Happen Next? 

  • Performance continues to be less jumpy each week as things stabilize.
  • CPMs will continue to rise since people are spending less time on social media.
  • Overall, we expect costs to continue to increase gently over time.
  • We expect the recent trend in prospecting to continue as many brands plan to retarget and retain this new audience.

Strategies to Adapt: 

  • Look at your geographic cost & spend breakdown, and start a campaign focusing on low-cost, low-spend cities.
  • Test creative that focuses on delivery speed & delivery visuals.
    • Include images with boxes on doorsteps to gauge interest from potential consumers. Try using copy that says, “Shipping to anywhere in the US in 48 hours.” Phrases like this are testing very well.
  • In previous weeks, we advised clients to expand prospecting and add new audiences as prospecting was having a honeymoon. 
    • The relative performance of prospecting, retargeting, and retention is back to normal. Now is the time to consider bringing your budgets back to equilibrium.
  • Consider testing large lookalike (3% through 10%) audiences since they are strong performers half of the time and customers may not be the same this month as last month.

Detail: What is ad performance like right now? 

Cost Per 1k Impressions (CPM):

Facebook CPMs have hit bottom and are increasing 5% each week.

  • What are we seeing?
    • CPMs dropped 40% in the early days of the crisis.
    • CPMs are continuing to rise about 5% per week starting the week of 3/22. 

Click-Through Rate (CTR):

Facebook CTR has been slowing rising in recent weeks, but CPMs are rising faster.

  • What are we seeing?
    • CTR had been rising by about 5% per week, but is now flat.
    • Retention was inexplicably strong in early April and is now reverting back to where it was in late March (we don’t have a good explanation for this, but we saw it on multiple accounts)

Cost Per Click (CPC):

CPM increases are larger than CTR growth, so CPC is slightly drifting upwards.

  • What are we seeing?
    • Early on, CPC dropped aggressively on falling CPM.
    • In early April, CPC has been flat, as rises in CPM have been offset almost exactly by rising CTR.
    • In late april, we are seeing a rise in CPC as rising CPMs have been matched with flat CTR.

On-Site Conversion: 

Conversion rates have been drifting downwards over the past 3 weeks.

  • What are we seeing?
    • Conversion rates rose in late March and early April.
    • Conversion rates are down about 20% in the past 3 weeks (and partial data for the week of 4/26 continues this trend).

Cost per Purchase: 

Due to CPC rising slightly and dropping conversion rates, we are seeing a 20% rise in recent cost per conversion.

  • What are we seeing?
    • We see the prospecting honeymoon phase (which was very noticeable in late March) is clearly over.
    • Retargeting continues to get cheaper for brands.
    • Retention costs are rising from a very low point.

How can we improve performance?

Tip #1: Look at your geographic cost and spend breakdown, and start a campaign focusing on low-cost and low-spend cities.  

(Note: This is the same tip I highlighted last week, but the update is that we have tested this more widely to more clients in the past week, and are even more confident in the importance of this tip)

In Normal Times:

  • If you created a Facebook campaign for “Whole U.S.”, Facebook would normally allocate the spend by city efficiently.
  • Facebook normally increases & decreases spend by geography until all locations have roughly similar costs.
  • Typically, running a hand-budgeted geo-specific campaign makes performance worse, not better.

Currently the Geo-Allocation System is not currently working:

  • For unknown reasons, the geo-distribution system at Facebook is not allocating everything properly (we think it’s using too long a trailing average so it’s using pre-Covid data).
  • If you look at your spend and cost by geography, you will find that there are certain cities that have very low costs and low spend. This is a flaw and shouldn’t be happening.

You can fix it by launching a “Underserved Geography” Campaign:

  1. If you go to the “breakdowns” in your ad manager, you can look at your costs & spend. You will find the DMAs that have well below average costs and have low spend per capita.
  2. Make a campaign that targets all your normal audiences, but only targets those underserved DMAs.
  3. Run this “Underserved DMA” campaign in parallel with your normal “Whole Country” campaign.
  4. Adjust budgets between these two campaigns until they have the same costs.

Tip #2: Test large Lookalikes (3% through 10%) in prospecting just in case your understanding of the customer is out of date. 

(Note: This is also the same tip as last week, but we have tested this more thoroughly, and are even more confident in the importance of this.)

We Normally Suggest:

  • We use lookalikes aggressively in prospecting for most of our accounts.
  • We don’t normally see useful performance with lookalikes larger than 2%.

In Light of COVID-19, We Recommend:

  • We have tested large lookalikes (3% through 10%) for a number of clients.
  • About half of the time they have been very strong performers.
  • We think they are doing well now because your customers today may be somewhat different than your pre-crisis customers, so larger lookalikes pull in the newer group better.

Tip #3: Start to re-allocate campaign budgets back towards retargeting & retention, as the strange “Prospecting Honeymoon” is over.

  • In late March, Prospecting became cheap relative to retargeting & retention.
  • At the time, we suggested an aggressive budget move towards prospecting.
  • The honeymoon is now over, and the relative performance of prospecting, retargeting, and retention is back to normal.
  • So now is the time to consider bringing your budgets back to equilibrium.

Tip #4: Test creative that focuses on delivery speed & delivery visuals.

  • Include images with boxes on doorsteps to gauge interest in potential consumers.
  • Try using copy that says, “Shipping to anywhere in the US in 48 hours.” Phrases like this are testing very well.

To stay up to date on our state of the industry metrics, test results, and opinions on everything Facebook, sign up for our R&D Labs newsletter here!

0 Comments