How many of the below questions/statements have you heard this past year?
CEO: “When is the new campaign launching? It’s been a month since the last one went live.”
CMO: “We’ve been running that case study ad for a couple months now, can we refresh it?”
VP of Demand: “Remember that ad set we ran back in Q1 with the before/after visuals? Looking back on demo request volume from the past year, it seems like they peaked around the same time + started to dip after we replaced them with the next ad set.”
Or even more pointed at myself, I know there have been times this past year where I’ve been thinking we were overdue for an ad refresh + started the process to kick off the next round.
Butttt have we ever stopped to wonder what logic was driving the need to launch the new ones? Most of the time, it seems like it’s just a function of time.
Campaigns are given an arbitrary set of days to run
A content calendar has pre-determined dates for ads to be running
Or, as David Ogilvy put it in Confessions of an Advertising Man (and what really made me stop + think about this):
Scores of good advertisements have been discarded before they lost their potency, largely because their sponsors got sick of seeing them.
You aren't advertising to a standing army; you are advertising to a moving parade. Three million consumers get married every year. The advertisement which sold a refrigerator to those who got married last year will probably be just as successful with those who get married next year.
Sponsor: Primer
Outbound is broken + LinkedIn somehow keeps getting more expensive. Now what?
Every B2B marketer is paying premium prices for the same shrinking pool of attention on the same two channels (LinkedIn + Google).
Meanwhile, channels like Meta, YouTube, + display cost a fraction of LinkedIn…but most B2B teams can’t use them. Standard account lists match at 15%, AKA audiences are too small to learn + too noisy to convert.
Primer fixes the match rate problem. They turn your ICP, target accounts, and net-new prospects into 45-85% matched audiences across every major ad platform. Now the efficient channels we couldn’t use before due to low match rates are suddenly usable.
Get started with a free account + 30-day trial of paid plans today!
Fun fact: Rippling credits them for their $145M Series B success.
And the timeline is only speeding up
With all of the recent AI mandates or AI shaming (“What do you mean you aren’t using AI as part of your demand gen workflow?"), the overarching message is that the solution lies inside of the “magic” formula: volume x speed.
We used to be limited by creative production, but AI can create more for us now, so why ship 10 ads when we can ship 100?
We used to run 1 campaign every quarter, but AI can move faster for us now, so why not run 1 campaign every month?
Do more. Go faster. Get more at bats. Hit more audiences.
But there are three things that have me going hmm when I see/hear this advice:
Where’s the feedback loop? Except for a handful of people I see using AI from a systems perspective vs solely as execution, for the one thing that all of those additional at bats are giving you (data), almost no one is taking those insights out + incorporating that feedback loop back into what’s created next.
Our attention spans are shorter than ever + we’re taking that out on our audience. Just because we see the ads daily when we’re inside the ad platforms doesn’t mean our audience has. We’ll have seen the ads hundreds of times from brainstorm to edits to campaign build in the platform before our audience even sees it once. Then compound the fact that our now 100+ ads in rotation means budget is spread really thin (because let’s be real, few are being given more budget right now), so we’d be lucky if the ad frequency for our market is ≥ 1 for most of those ads.
We’re iterating on the winner(s) at best, not continuing them. This was a big one for me this year. We’d have a handful of ads that just crushedddd it. Then the next campaign came around, we’d turn off all the ads in the old campaign, take the learnings from what did/didn’t work, and apply that to future ads. But after that Ogilvy quote had me facepalming myself, I thought to myself “why am I pulling something that’s still working and hoping that maybe a few ads in the next campaign can match this?”
What I’m doing about it
Shrugging it off + hoping for the best.
Kidding.
The data side of me is checking the numbers inside platforms for datapoints other than how many days it’s been since launch:
Frequency at the ad level
CTR/view rates within the same audience
Performance decay rate relative to other ads launched at the same time
But then since we do have to keep up with leadership’s expectations of shipping more/faster, I changed the account structure we used for which ads were running at any point in time + how much budget to give them.
Old way we ran:
New campaign gets the heavy majority of the budget (would often be a mish-mash of product/feature ads, case study ads, etc.)
Old campaign is retired
Rinse + repeat the above until exhaustion or mental breakdown occurs first.
New way I recommend:
New campaign (20% of budget)
This is usually oriented around time-based relevance: a new product/feature release, a seasonal offer, what’s popular with the market right now (aka our adaptive POV), etc.
Case studies/customer proof (20% of budget)
Thought leader ads (20% of budget)
Use case content (20% of budget)
Then last + most importantly - top performers (20% of budget)
Instead of ending all the ads in the “old” campaign cold turkey, I’d take the handful that I deemed as top performing and/or still performing, and roll them into an evergreen “Top Performers” campaign. This extends the life span of them for as long as I want, I can toggle various ones on/off if I know some are more seasonal in nature, and/or when a proper system is in place to iterate off what’s worked, we have a beautiful library ready to be used as inputs.
I really like this new setup for two main reasons.
First is the balance it brings. No more big swings in top of funnel performance if a campaign hits (or flops).
Instead of taking 80%+ of the budget and hoping for the best, it’s now roughly a fifth of the budget. If it takes off, great, push more budget to it. If it flops, great, pull back some budget from it. There’s built in containment here.
Second is the compounding performance it brings. We’re stashing top performers as time goes on, so our library of “what works” keeps growing and we can push harder on that campaign budget at any time if needed.
We also get to learn from that pool + can either manually or via AI systems build a feedback loop from that campaign so that everything new we creates has more data available to inform what’ll increase the likelihood that those ads are also successes.
Charlie Munger may not have been a marketer, but I’m pretty sure he’d agree with this principle.
See you next week,
Sam