During the holiday season, there’s an extra focus on the naughty and cute menu.
Successful campaigns gain access to additional resources, while failed initiatives pivot or are retired.
This month’s question gets to the heart of optimizing and scaling digital marketing. asks Garland of Orlando:
“At what point do you consider campaign testing to be a failure? For example, $5,000 spent on data and little return on spend.”
In this post, we’ll look at understanding success/fail signals, as well as unpack how they create your brand.
This question requires a lot of variables, so we’ll do our best to handle the most common ones.
Prepare reasonable tests
Before embarking on any digital marketing test, it is really important to define the success and failure criteria.
The most basic step is confirming what is known absolutely (ie, do you trust your conversion tracking, does form filling work, is your sales force strong, etc.).
If these constituents are not set correctly, it won’t matter how well the variables you test perform.
This is why it is so important to bake at least 1-2 months in preparation expense.
Along with finishing learning periods, you will ensure that your performance reflects genuine success.
It is also essential that the tests check only one variable at a time.
If you decide to test everything at once, you will have a hard time getting definitive conclusions about whether variables have positive or negative effects on campaigns.
Finally, it is important to note that all digital ad networks have different learning periods and interaction rules to communicate effectively with the algorithm.
For example, Google requires at least five days, while Facebook (Meta Ads) requires you to meet a conversion limit.
Determine successes and failures
Once you’ve set up your ground conditions, you can begin to define what success and failure look like.
If you are testing Conversion Rate Optimized (CRO), your tests will likely focus on the following levers:
- landing pagesDo they inspire more, less, or the same amount of engaged users?
- advertisementsDo they have a healthy click-through rate (CTR) for conversion rate?
- Targeting optionsIs the target group of people better, worse, or the conversion rate itself? And value?
The ROAS tests will focus on the following options:
- auction priceDo the auctions the campaign enter lead to better, worse, or the same return on investment?
- user journey: Is the user directed in a way that gives way to the higher, lower, or same conversion value?
- creativeDoes creativity help pre-qualify customers better, worse, or the same as before?
Testing a new channel requires slightly different considerations:
- Easy maintenanceCan you reasonably build and maintain a campaign on the new channel, or will it require completely different resources?
- Market valueIs this channel too focused on your best customers or is it new ground?
- budgetHave you allocated enough budget for the channel?
- GoalIs your target audience on this channel?
You’ll want to give any initiative at least 60 days to prove itself; However, if there are obvious signs of failure, you will need to adjust.
Clear signs of failure
The following should be taken as clear signs of failure in the calculations.
- Campaigns cannot be spent after more than five days.
- Conversions in the account do not translate into quality leads/sales.
- High spending is much higher than normal spending rate.
- The variables being tested yield worse results than the control group.
The final takeaway
It’s easy to feel that any spending that doesn’t yield a return is a waste – but it’s never wasted if you’re learning something.
Make sure to create your primary data points and also honor the original pass/fail signals.
Do you have a question about PPC? Submit via this form or tweet me at @navahf using the hashtag #AskPPC. See you next month!
More resources:
- How to A/B Test Responsive Search Ads: Step by Step with Tips
- How can I build on my SEO knowledge to be better at PPC?
- The biggest PPC trends for 2023, according to 22 experts
Featured image: Paolo Bobita/Shutterstock
