Why Display Ads Are a Waste
The data that made us stop.
Hey folks 👋, I'm Kenneth Burke. #BurkeBits is where I share stories, data, and frameworks to help you become a better marketer. Subscribe for free to level-up.
We’ve done display ads for years as a marketing “best practice.”
To be clear, display ads are those ads that follow you around the web after you’ve been to a particular site or viewed a specific product. E.g. you go to nike.com and then start seeing ads for running shoes on your local news site.
I’m sure these ads are very successful for some brands and products, but as a business texting service, we saw mixed results at best and eventually pulled enough data to tell us to stop.
Here’s the story.
Our First Shot at Display Ads
We started with AdRoll as a platform for retargeting viewers who’d already been to our website, and for finding net new prospects to market to.
We didn’t have a designer on staff at the time, so we outsourced the image ads. They were okay.
Our performance on AdRoll seemed promising, with better results coming from retargeting than from prospecting ads. Except, it was tough to tell the full impact (or lack thereof).
Eventually we paused display ads because other channels were more clearly and consistently bringing us returns. But we’d be back.
The Power of Good Creative
Sometime later we started using Google’s network for display ads, primarily for retargeting people who’d already visited our website. We also hired our first full-time product designer, who helped with marketing assets.
They created new ads for us that were much more appealing, and ad engagement went up 30-40% pretty much instantly.
We still had trouble clearly attributing leads and sales to display ads, but there were enough positive signals that we kept going.
In general, display ads are low-risk with high-reward potential. They’re so cheap, you might as well use them until you find a reason not to.
The Big Experiment
We’ve run experiments forever, and now we wanted to see if display ads were worth it. So here’s what we did:
Overview: We compared a period with display ads running to a period without them running, and looked to see if there were any differences.
Test Hypothesis: Display ads don’t bring us qualified leads, so we’ll see no change when we turn these ads off.
Experiment Period: 2 months of active display ads, followed by 2 months with no display ads.
The Metric: Number of qualified leads that come in through our SMS Chat (our live chat widget for texting).
Why SMS Chats, specifically? It’s a high buying-intent action, there’s plenty of volume to measure quickly, and we’re able to track the source, medium, and campaign of each chat. Then we can skim each chat and categorize it as a lead, existing customer, or junk.
There were 3 qualified SMS Chat leads from display ads during the active period, and there were 230 junk chats. We could not attribute a sale to display ads.
Once we turned these ads off, we saw a decrease in junk chats by the same amount we were getting from display. Total number of leads and sales were unaffected.
Grains of Salt:
The active period was a low point for us, and for most in SaaS. It’s possible that skewed the numbers.
If someone visited our website from a display ad, and clicked around a few pages before chatting in, the particular tracking we were doing broke down.
Most of the junk traffic seemed to think we represented the website our display ads were shown on. Not sure what that means, but worth noting.
Advertising usually has a notable lag effect. It may be several months before the effects show themselves, and we only experimented for a few months.
Still, if you can cut an ad channel off for months and see no difference in leads or sales, it’s a good call to reallocate that budget. So that’s what we did.
Why Display Ads Are a Waste
I don’t actually think display ads are a waste for everyone, but there seems to me to be clear indicators of when they will and wont’ work.
Won’t work when:
A purchase has multiple decision makers
It’s sales-led (a rep is guiding the process)
The ads are on sites that have nothing to do with your product (that’s just an interruption)
Will work when:
It’s a consumer product, or single decision maker
You don’t have to talk to anyone to buy (contactless)
The ads fit naturally into the websites they’re on
I’m not saying we won’t ever use display ads again, but we’ll need a pretty good case that they’ll bring us more qualified leads and sales before we start.
Was this helpful or interesting? Who else should see it? Be sure to subscribe and share it with them.
Have questions about this topic or something you’re working on? Ask away! I’m an open book and happy to help.