The Biggest Mistakes we see on Paid Media Platforms.

Having audited thousands of accounts, and efficiently ran millions of dollars across multiple campaigns for various clients, we catch up with some of our platform experts to get their opinions. What are the biggest mistakes you see on various paid marketing channels?


1.  Facebook / Instagram

Not seeing hard enough exclusions between audience type, or the campaigns aren’t broken out by audience type (prospecting, Retargeting, Remarketing). It’s either that the buyer knows what they are doing and wants to take more credit for sales that were already going to happen, so their performance looks better. Alternatively, it could be that they don’t understand that if you don’t put in the hardest exclusions, you’ll get the same people being targeted across your different campaigns. Skewing results for CPA, ROAS, and ultimately where performance is really being driven from.

See our Case Studies that drive positive incremental ROAS or our FB/IG services page for more information.


2. Search 

One of the biggest mistakes we’ve seen in search is not reviewing your search term report to have a consistently updated negative keyword list. With so much query variance, you need to go broad initially to test which keywords you might show up for but there’s so much potential overlap of keywords with different meanings that you might be paying to show up for a question you don’t have the answer to. Which leads to a lot of wasted money.


3. Programmatic

Provides no lift in impact. We have seen the contrary but we wanted to unpack this a bit further. Working with our data scientist team, we found that the stereotypical issue lies in a few key areas but mainly in the design flaws of the test. The flaws have and continuously occur when the test and control groups are not properly pre-defined, pre-split, and controlled correctly. If for example an industry standard of 90%/10% test/control is not accurately set up based on issues with audience identification then that’s just starting the test out on a loss. It is key to get over that speed bump of identifying a user between platform to platform, publisher to publisher, or device to device. That way an advertiser can make sure a user is placed properly in a control or a testing group. Many overlook the planning and understanding behind creating the proper parameters for a test, leading them to undefined or inconclusive results. It is essential to understand the tech stack you are testing in to ensure that you have all variables defined and accounted for to yield actionable insights. 


4. Google Shopping

Going too broad or too narrow, depending on the inventory that you have and the number of products. We see that many accounts will have either all their products in 1 campaign, which will end up just showing your top products and run out of budget before the rest of your inventory has a chance. Or on the opposite side having the products so segmented that your conversion-based optimization doesn’t have enough data to properly target who would actually buy your product. I think too many people try to apply cookie-cutter methods to very different client needs.


5. YouTube

2 words, bad creative. YouTube is an incredibly creative forward ad platform that requires you to convey a message in 5-seconds before someone – who is actively waiting to skip your ad – skips it. If you aren’t able to tell that story in a short amount of time or at least mention your brand, then you’re wasting your money. Many people try to rehash social ads into YouTube creative when the objectives are generally different. Meaning the creative ads needed should be different. We see this often with businesses, applying a one-size-fits-all to channels, which rarely, if ever, works.


6. Snapchat

Not selecting pre-load on standard video ads for a better user experience and decrease in website abandonment.

People often comment that pre-loading ads cause issues with their Google Analytics reporting. E.G. every impression on Snap shows up in Google Analytics as Traffic. This is true, but there are scripts you can implementing Google Analytics to mitigate this, so traffic only shows up for ads that have been clicked on.

It’s fine if you don’t select pre-load because one of the following reasons, but in general, this probably doesn’t apply to 95% of advertisers.

    • Your site is extremely fast on mobile and you want to reduce server requests or analytics discrepancies.
    • You have 3rd party advertising pop-ups
    • Prompting for the Snapchatter’s location

7. TikTok

Missing out on low rate high-value opportunities, which won’t come around again for cheap.

As a platform, TikTok is developing at an incredible speed to keep up with its more mature competitors, but also adds its own unique features and value. E.G. From Hashtag Challenges to spark ads, and now an ‘affiliates’ marketplace being tested in beta. 

Whatever way you look at it, TikTok has allowed businesses to be guinea pigs in return for cheap rates to prove product success. Obviously, a business as part of this test takes on some risk, but it is hugely outweighed by the potential reward. 

This is where I think part of the problem lies. For a platform like TikTok, brand-led businesses find it difficult to give up an element of control to creators in what they produce. And when they do get too involved and start telling a creator what it should be and how it should look, the results they drive are usually poor. The flip-side to this is many businesses don’t understand how to make TikTok content (not ads), and how to become part of the thousands of trending hashtag conversations. Herein lies some of the problems.


8. Creative

While this isn’t a channel, it does apply massively in driving performance. 

A lot of businesses don’t put enough focus on channel-specific performance creative. Research by Nielsen suggests that creative impact explains up to 47% of ad result variance, yet businesses rarely give it more than 5% of their attention. This is due to the platform algorithm needing to drive the best and most relevant experience for audiences, based on data collected from interactions, engagement, and conversion events.

I believe that many businesses are brand-led in their approach to marketing campaigns, and try to work a one-size-fits-all across their marketing channels. It keeps brand messaging and image consistent, cuts down creative production timelines, it’s overall easier, and makes resourcing more manageable. However, because of this, it’s generally the reason you won’t see desirable ad performance.