In the world of PPC marketing, everyone fails at some point. It’s inevitable.
In fact, it will probably happen to you. Many times.
But you can learn a lot from your PPC failures.
Failure is the opposite of success, so when we talk about failure we need to understand what success was going to be.
In the following case study, success was defined as discovering profitable opportunities available in Google Ads.
The Situation
This client is a niche player in the health and beauty industry, which features numerous national brands with large PPC budgets. We knew this going into the engagement.
We were also aware that there would be keywords, products, and categories that just wouldn’t be profitable.
However, we figured that if we could find a handful of profitable areas we would be able to add a profitable channel to their marketing toolbox.
We were focusing on new client acquisition and planned to exclude existing customers. We uploaded a Customer Match list into Google Ads to accomplish this.
Shopping ads were being tested as well as text ads with keyword targeting. The test budget was $1,500 for click charges. The test was scheduled to run during Q4 of 2018 to take advantage of the holiday gift buying season.
What Happened
After completing keyword research, building out the account structure, writing ad copy (and copy for ad extensions) and getting our product feed active in Google Merchant Center we started advertising during the first half of October.
The search volume for our prime keywords and products was large enough that we had more than enough traffic to exhaust our budgets. The issue was more about cherry-picking the best possible keywords so that conversion rates would be high.
TL;DR version: we stopped the experiment after spending $1,000 on clicks and generating about the same in sales.
The 1:1 ROI wasn’t nearly sustainable for the company, even taking into account lifetime value (LTV) of customers.
Postmortem: What Went Wrong
There were several bumps in the road. None by itself was fatal to our efforts, but added together they definitely put a damper on things.
Some issues we ran into:
Google Changed the Policy on Customer Match Lists
The new policy stated that an account had to have $50,000+ in lifetime spend to use Customer Match (basically) so we weren’t able to effectively exclude existing customers like we had planned.
Our Product Feed Expired
The vendor in charge of the product feed didn’t set the feed to automatically pull so we lost a few days of shopping ads traffic during the experiment.
(I fully admit fault here in that I should have double checked the feed settings and also could have caught the impressions and clicks going to zero more quickly.)
Our Credit Card Got Denied
This happens far too frequently with new advertisers. I’m not sure it’s Google or the card issuing companies, but the card was denied and the account stopped running.
(Google’s message that “Automatic payment declined for $XXX. No reason provided by your financial institution” doesn’t help much.)
By the time we jumped through all the hoops, we lost a couple more days on all our efforts.
Conversion Rates Were Lower Than Expected
The site had a sitewide conversion rate over 4 percent before we ran the test.
We predicted that PPC traffic wouldn’t achieve this level (email and direct traffic frequently convert at higher rates because they include existing customers repurchasing, which is common in this industry) but the 2 percent conversion rate we did achieve didn’t make the math work out.
The biggest factor here was obviously the conversion rate. The math just didn’t work.
$1.25 CPC / 2% conversion rate = $62.50 cost per conversion
With an average order value of about $61 it just wasn’t financially sustainable. The budget didn’t allow for enough traffic and testing to improve conversion rates.
Our positions were top of page, but barely, so reducing bids was going to kill visibility. Again, when you only have so much budget you have to get things right from the start and hope for a little luck, too.
Turning PPC Failure Into Success
As I mentioned at the beginning of the post, success was defined as finding profitable opportunities in Google Ads. This experiment was a failure according to that definition.
However, I strongly believe that experiments and tests are only failures if you don’t learn from them. There were several takeaways for the company that justify the expense and effort of this test:
1. Establish a Baseline
We learned that CPCs for shopping ads and all of our top categories would be at least $1.25-$1.40 to get exposure. Not great exposure, but top of page.
We saw that our best campaign had a 3.25 percent conversion rate and account-wide was 2 percent. So even in a best-case scenario with bottom-of-the-range CPC and top-of-the-range conversion rate we still had a cost/conversion of about $38.50.
That still wasn’t good enough to continue in the client’s eyes.
2. Know Which Products Sell & Which Keywords Lead to Sales
Admittedly, there weren’t a ton of conversions at this budget level, but we saw which products sold through Shopping ads and which keywords got us sales in text ads.
We also learned which keywords did not get us sales, which is equally helpful.
3. Actual Search Query Data
In the future, this company now has real data from real queries typed by real people and customers.
While keyword research tools are helpful, they are estimations and approximations. This data is a roadmap for future tests.
4. Tested Ad Copy
We had two primary messages in our ad copy.
- One centered on third-party validation of the product’s quality and used “Award-Winning” prominently.
- One centered on discounts and featured a percent-off discount that new customers could earn by signing up for the newsletter.
The messaging that used “Award-Winning” was the winner. That information can now be used across other channels.
The fact that the percent-off messaging didn’t work could indicate the percentage isn’t high enough or that people don’t respond as well to that type of discounting (and a dollar-off promotion could be tested).
Conclusion
Nobody likes to fail. Pulling the plug on an experiment early makes it even more painful.
However, even in failure you can learn valuable insights that guide your future efforts and inform efforts in other channels and areas of the business.
As Malcolm Ford said, “Failure is success if we learn from it.”
More Resources:
