A recent ad-effectiveness study by Meta and Nielsen confirmed what media planners have known for 70 years: the four most important factors in media impact are:
As digital media giants like Google and Meta have grown over the last twenty years, many marketers have built their careers around deterministic attribution, algorithmic optimizations, and more data than they know what to do with.
There has been so much trust developed in the marketing community around "the algo" (as if it was some kind of omniscient marketing deity that decided who made money and who didn't) that we've forgotten the basics.
If you don't have a compelling message -- repeated multiple times, for a sustained period of time -- people are less likely to buy your product.
That's all that used to matter before the rise of Big Data, but even today it's still the most effective.
People get bombarded with thousands of ads per day across hundreds of products. What are you going to do to stand out among the noise and actually convince people to make the purchase?
Here's how to interpret the study, and what to do to maximize your Facebook investment.
Research shows time and time again that creative is the #1 most impactful media lever. It should be eye-catching, compelling, and appropriate for your brand, product, audience, and environment.
If you are not 100% confident in your creative, do it again. Even if you are 100% confident in your creative, it's probably not good enough. No amount of creative optimization is enough (and that does not mean inundation of quantity, but true quality).
You've got a few fractions of a second before someone scrolls on, so it has to have that initial thumb-stopping component, but then quickly and convincingly establish credible value. If you feel like your product is too complex to explain in 15 seconds, then split up your message across multiple ads by funnel stage (brand > problem > solution > action).
The study found that showing an ad 2.25 times per week was 80% more effective than only once per week. This reinforces the common refrain that people need to see an ad 8-10 times a month in order to generate affinity.
If your Facebook account-level frequency is lower than 8 over the last month, you're spreading yourself too thin. More refined audience targeting, retargeting past video ad engagers, or aggressive audience exclusions are some ways to increase impression delivery against the same target.
I see hundreds of brands make a big mistake and give up on a campaign before it has even exited the learning phase.
Consumers need continuity in order to develop brand affinity and purchase intent. It's the constant reminder of a product that eventually gets someone to buy (some viral products can get away without doing this, but that's luck, not strategy).
It's fine to have test initiatives, but running an “evergreen” campaign for nearly a year is significantly better at driving incremental revenue than a short flight.
Spend more money, get more money
If you hammer the same 1,000 people hundreds of times, and all of them buy, it’s still not going to be a major needle mover. Even if you have a high frequency with awesome creative, honing in on a niche that is too tiny will never scale.
For most brands, that means spending more money. Reaching about 15% of the population has a much better shot of generating incremental revenue than serving ads to only 1%.
But crucially, in order to do so at a frequency that is heavy enough, you will need to invest. Building a brand and moving significant numbers of people to take action isn’t a cheap process.
Stop having blind faith in the algorithm
Somewhere during the last ten years, performance marketers have forgotten what it’s all about: connecting with potential customers. There are certain fundamentals that will likely never change, and simply giving Facebook your credit card, testing thousands of creative variations at $100 a piece, and hoping for the best won’t work for every brand (or arguably, most brands).
Follow the basics -- creative, frequency, continuity, reach -- and grow your revenue.