Recently I’ve been part of podcasts and even had a talk at the Forbes Europe CMO summit with the great Seth Matlins asking questions framed around performance marketing vs brand marketing. I particularly liked Seth’s take of “+/vs” and I just wish we could retire the term “performance marketing” as a distinction from brand marketing. We have a marketing funnel, at the very top of the funnel is brand building and the very bottom is direct response. I believe two things are true about the funnel:
- The WHOLE funnel Performs
- Pure Brand and Pure Direct Response only exist in theory
I talk about this at length in my upcoming book Click Here.
Let me break those down:
The whole funnel performs
In my work I have spent billions of dollars throughout the funnel. Just this year we ran hugely successful ad campaigns to grow WhatsApp in the Northeast States of the USA and Texas. In these campaigns we used the entire funnel from pure Brand Awareness building to direct response conversion optimized ads in Facebook and Instagram.
From pure ads that just said Download WhatsApp:
To a fun brand spot with modern family:
We measured the whole funnel and how each part of the funnel interacted with each other part using this test plan:
And guess what we found? Yep, it worked best where we did EVERYTHING. We drove up awareness through consideration to actual retained active user growth. At the same time we continued to build WhatsApp’s brand leadership as the most private messaging service in the USA.
We have run the similar tests on Instagram, Ray Ban Meta Glasses, Meta Quest and more across the US, CA and globally in the last year and the results keep coming back the same it isn’t brand vs performance, it’s brand + direct response PERFORM better together.
Pure Brand and Pure Direct Response only exist in Theory
In physics we were often thought to think about theoretical limits in any experiment as they were great to perform thought experiments for what the extremes of a function would look like. But the thing is the limit often does not exist in these functions. I worry we’ve done the same thing in marketing but forgot that the limit is theoretical.
Something we are focused on is that our headsets and glasses are premium products at amazing prices. If we run every direct response ad focused on price point, promotional price cuts and so on we will cheapen the brand which will limit our pricing power in the future. So here (for example) we look at long term impact and only have 15% of our impressions leading with price. So ultimately realizing these direct response ads also impact the brand and nothing can be too extreme. This part of the work is very hard to measure but also very logical and so we do 2 things: 1) we try and look for any impact on pricing expectation/premium nature of our products in hold outs but 2) above all we simply argue on logic we shouldn’t go too far and make sure with our product partners and leadership they agree with the logical balance we are striking to give long term performance.
More broadly though if you have billions of impressions of direct response ads (like my teams) the idea they won’t impact the brand is just silly. They will and you need to not treat them like an extreme that is only direct response, you have to use judgment and measurement to make sure they don’t damage the brand and ideally help it. We manage this by having sentiment guardrails on campaigns where we can and measuring if they are impacted for our core brand metrics. Where we can’t we observe longitudinally to see if there’s any correlation between Direct Response ads and brand metrics we care about. Finally we apply logic.
Now the reverse is true as well. Almost no pure top of the funnel brand campaign exists without impact all the way through to action. If you are driving awareness of a RayBan Meta Wayfarers without ending up driving people to buy, are you really driving good awareness? If you are moving the sentiment that (for example) IG is about every day creativity and not growing creation on the platform are you moving that sentiment metric in any way that matters?
We need to pretend the limits exist and understand all marketing performs and no part of the funnel is truly isolated from any other part of the funnel.
Conclusion
In the end this debate, most likely, won’t die for years but I wanted to set out my stall on this issue. I believe all marketing performs. Yes you measure it all in different ways but as a dyed in the wool direct response marketer who grew up with affiliates, search and social and sells primarily direct response marketing for a living I buy ads right the way through the funnel, across all channels for what they are most effective for and I don’t see this simple dichotomy as a real or helpful way of looking at things.