Happy Sunday, Everyone!
As a marketer, today is one of my favorite days of the year: it’s the Big Game™. In a few short hours, a total of ~$650M will be spent on ~100 0:30 ad spots: ~60 in-game commercials (at a reported price of $7M to $8M per spot), and ~40 commercials before and after the event (at a reported price of $3M to $5M per spot, depending on placement). And that’s just the actual cost of the spot – it doesn’t include the production cost, other channel deployments (like social, out-of-home, other TV networks, digital, etc.), supporting material costs (mailers, in-store signage, banners, events) and talent (do you think Matthew McConaughey is supporting the great football conspiracy for free?). Conservatively, brands will invest several billion dollars – all for a chance to “win” customers during one of the biggest moments of the year.
Put another way: today is a moment (and a big one at that). It’s a celebration of advertising and consumerism and capitalism as much as a championship sporting event (also: Go Birds). It’s the only event where people are as excited to watch the commercials as the game. But - for as exciting as moments like these are, they are not all sunshine and rainbows. There’s a darker, more insidious side lurking just below the surface: a willingness to attribute everything that happens around a moment to the moment.
A concrete example: we’re in the midst of an audit for a large fitness brand. As part of our scope, we’ve reviewed reports from their current partners, ad account notes, dashboards, decks, actual creative, landers, etc. Over the course of doing this, one thing has stuck out to me: a willingness to attribute “down” performance to outside factors – the economy, the algorithm, macro-disruptions, the list goes on and on. Now, to anyone reading these reports, each of these explanations, on its own, seems eminently reasonable: Prime Day definitely can impact conversion rates. The most expensive election in history is likely to impact ad markets. Wide-ranging policy shifts and significant uncertainty from the world’s largest economy are probably going to roil consumer sentiment. The Holidays (writ large) are a capitalistic free-for-all like no other. People go on vacation (or just don’t want to buy stuff) during New Years.
Taken together, a darker pattern emerges – one that I’ve seen countless times across brands large and small: a refusal to accept control and accountability, and instead, blame the algorithm (or the market, or the macro-economic conditions or the moment). If you’ve been around enough businesses or meetings, you’ve probably seen it, too. And, on some level, I get it. The allure of attributing poor performance to some esoteric, ephemeral factor is compelling. About the only thing everyone - brands, agencies, politicians and the general public - can agree on is that the algorithm (or the market, whatever) is a phenomenal punching bag.
Here’s the dirty secret: it’s probably not the algorithm (or the market, or the macro, or the moment). It’s probably something else. That something is probably in your control, at least to a certain degree. And you should probably have spent more time finding that thing (vs. blaming everything else).
It’s harsh, but it’s true. Accountability is a bitter pill.
All of this leads to a single question: if it isn’t the algorithm, then what is it and (more importantly), what do we do about it?
Step #1: Have a Theory of the Case
I firmly believe that most of :gestures broadly: this is a direct result of marketers lacking a clear perspective on how various types of events (or economic shifts, or catastrophic events, or moments) will impact their campaigns/initiatives/ad accounts. The most successful marketers, on the other hand, are crystal clear. They have a POV on how impacts will manifest, the magnitude of those impacts and their expected duration.
A great example of this was the 2024 election. As far as two months before the election, I observed people blaming the election for poor performance, slowdowns in sales velocity, increased ad costs, and a dozen other things. That continued through most of December (six weeks after election-related ad spending ceased). Did the election have some impact on ad markets? Absolutely. Was it responsible for your ad account sucking for a third of the year? Absolutely not.
This is where having a theory of the case can help marketers make better life choices.
If your working theory was that the 2024 election would be the most watched and the most expensive in history (something I wrote about here), then your theory of the case should have included the following: (1) most ad spending will ramp up in the ~4 weeks leading to Election Day; (2) that spending will be overwhelmingly concentrated in swing states and swing districts, not distributed evenly; and (3) this was going to be a divisive event, so there was likely to be an escalation online (more comments, trolling, etc.) than you’d expect from a (so-called) normal election. This, in turn, was likely to result in increased time online (more inventory, so the impacts to CPMs might be less pronounced), more donations (less $$ available to buy stuff) and more distractions (lower-than-expected conversion rates).
Yes, there was a ton more going on than this, but the above framework provides the basics for a working theory on the impacts of the event on the ad markets, along with some clear ways to test that theory:
- Virtually all digital inventory is bought/sold on CPM (if you buy on CPC, you’re just trusting someone else’s math) – and you have both historical CPMs and trending CPMs. If you expect political spending to drive up ad costs, then you should be able to see the impact of that by looking at rates of CPM increase in 2024 vs. historicals during this period. That gives you a benchmark for how much your cost is going up – so if your CPA increase is greater than the CPM increase, something else is going on.
- If you drill into your costs by location, you should expect to see that costs are rising faster in swing states vs. “sure” states (i.e. there’s not a compelling reason for either campaign to spend too much money in Kansas or Connecticut).
- The nature of a significant election is that it will suck some of the proverbial oxygen out of the room – some of the money that would ordinarily be spent on goods will be donated to candidates/campaigns (expected result: lower conversion rates); some people will hold off on major purchases until there’s more clarity around who’s going to be in charge (expected result: lower conversion rates and/or slower deal velocity); overstimulation does tend to correlate with lower action rates (expected result: lower CTRs + CVRs in the most advertised-to markets).
- Political content can be divisive - and algorithms love controversy. You’d expect that blander/safer/more generic content will not be served as often down the stretch (read: October-Election Day). Easiest way to observe that? Impressions + post actions for your organic content.
The thing to notice for each of these four points: there’s a clear pathway to validate each point (or, at the very least, begin to validate it). That, in turn, gives you (the marketer) a quick read on whether (or to what extent) what you’re seeing is a result of a larger moment/macro situation, or something else.
If - hypothetically - your working hypothesis was that the increase in CPA was a result of the election, but you opened your ad account/analytics platform to find:
- CPAs in swing states were flat
- CPMs were increasing faster in “safe” states vs. swing states
- CVRs were unchanged
- CTRs were declining across-the-board
- Start-To-Finish rates had plummeted (i.e. Form Start - Form Finish or ATC - Checkout) across the board
Well, the election is likely not to blame for your rising CPA. In fact, not a single one of these points aligns with your theory of the case – which means - more likely than not - something else is going on, and you need to figure out what that is (and do so expeditiously).
What I love about this method is that it forces you to think through the expected impacts of major moments before they happen - and make predictions for how those moments/events will manifest in your data. You’re not always going to have the luxury of time before a major event (i.e. a sudden market crash, a natural disaster, an unexpected tragedy) strikes, but you can still have a “playbook” for how you handle them developed in advance (similar to how PR teams have crisis planning documents for hypothetical events).Whether it’s a known moment or an expected event, share what you think will happen with your client/boss/team in advance (or as quickly as possible). That lays the foundation for accountability and honest conversations, vs. rote boogeyman blaming.
I should also highlight – even in situations where your theory of the case IS right and where a major outside influence DOES appear to be the culprit behind suboptimal performance, that does not give you license to throw your hands up and blame the [whatever]. In that case, your job shifts to finding ways to win despite the adverse conditions. The reality is that IF it is a market-wide impact, then everyone else (including your competition) is being hammered – which gives you an advantage: while their team is stuck in neutral, moping about and blaming the algorithm, you can be proactively trying to overcome these conditions – finding new audiences, testing new offers, improving your landing page experience, refining your exclusions, etc.
Step #2: Triangulate The Problem
Every marketing activity has points of failure – and it’s the job of the marketer to identify which of those points is responsible for the change in overall outcome (whether that’s lower sales, fewer leads, lower lead quality, lower conversion rates, higher cost per acquisition, etc.)
For digital ads, there are a handful of points of failure:
- Targeting - are your ads going to the right audience? Has that audience changed (i.e. a targeting criteria removal/redefinition, or a change in search behavior)? Do you have the correct exclusions in place (i.e. negative KWs, audience exclusions) to ensure that we’re not serving ads to people who are not relevant or able to convert? Have you checked your search terms report + placement report? Have you updated your audience research?
- Ecosystem - this is primarily for search, but has the ecosystem where the ad unit appears shifted? The most common manifestation of this is changes to SERPs to include AI overviews, change the number of above-the-fold ads, or shift the type(s) of ad units served on the SERP (i.e. going from search ads to shopping ads). Ads aren’t served in a vacuum, so it is imperative that you know the context in which they’re being served. If your SERP has changed, but your strategy/approach/creative has not followed suit, then expecting the same results is either insanity or delusion - neither of which most brands can afford.
- Settings - has Meta/Google/YouTube/TikTok made a change - an auto-applied recommendation, audience expansion, broad match KWs, etc. - that has resulted in a change to who is seeing your ad, or how they’re seeing your ad? Ad platforms are getting progressively more insidious in how they hide/conceal/position absolutely critical settings – to the point where, if you aren’t vigilant, you could inadvertently waste thousands (or more) in just a few hours. In the digital advertising + media buying classes I teach, the very first slide I share with students (after the university-required ones) has just six words: “The devil hides in the defaults.” Platforms don’t set defaults because they want to help you; they set defaults because those settings are what makes them the most money with the least accountability.
- Ad Creative - If your creative is dull, uninspired, irrelevant to the audience or otherwise generic AND you're shipping it at a time when competition for attention is already heightened (i.e. during a major moment), what do you think is going to happen? In fact, in this very audit I mentioned above, we found dozens of examples of the creative not aligning with the audience, not presenting the correct offer and/or being woefully generic. There’s a reason any good marketer will tell you to constantly self-scout: because if you don’t see the competition in the wild, if you don’t know what else is appearing in the feed, if you don’t understand who else is competing for your audience’s attention, how can you hope to win?
- Lander - Landers are (probably) the most ignored part of the entire ad value chain, for reasons that (to this day) boggle my mind. The lander is (arguably) the single-most-important component of the experience, because it’s where curiosity is transmuted into bona-fide intention. The ad gets attention; the lander (and the offer) close the deal. Yet, somehow, I routinely see brands spend $10k, $20k, $50k on dozens of influencers, creative angles, etc. only to send all of the traffic from them to a generic, boring, milquetoast lander. In “normal” conditions, with a great product and a strong brand, that can work – it isn’t ideal, but you’ll win in spite of your suboptimal marketing choices. But, during a moment? When competition is higher, attention is harder to keep, money is tighter, ad spaces are more expensive? Those details matter more. That landing page that previously didn’t help or hurt turns into an anchor, dragging down your campaign performance.
- Offer - Finally, there’s the almighty offer – something most marketers rarely consider (and almost never actually investigate). The example I use most often is of a home services company (part eCommerce + part lead gen - a little something for everyone) that saw performance crater repeatedly. Their old agency blamed the algo (and the market, and interest rates, and (probably) Zeus and avocado toast, too) – but when we looked into this further, the real culprit was simple: their offer was being dominated by their competition. They were running (for example) a “Buy 3 - Get 1 Free” sale on windows; their primary competitor was running a “Buy 1- Get 1 Free.” The reason why their cost per conversion was rising while their number of new customers was falling was simple: their offer was being blown out of the water by their competition – and instead of pivoting to a “beater” offer or shifting their approach to counter this (i.e. focusing on the quality difference, talking about the exclusions/fine print their competitor relied on to avoid actually honoring the discount), their agency just threw more and more money behind this loser. It was the marketing equivalent of putting a jet engine on the back of the Titanic: the only thing you’re going to succeed at accomplishing is sinking it faster.
- Technicals/Experience - is your site/app/platform performing as expected across all devices? Is the checkout working? Are there broken links on the page? Are costs (like shipping/taxes) calculating properly? Does the form actually submit? Any time there’s an issue, this is the first question I ask our team: when is the last time you physically went through the conversion process step-by-step? What device(s) did you use? Have any negative events fired? Did you review heatmaps or session recordings?
I have seen/found dozens of examples of situations where a payment gateway failed, or a tiny detail on a lander was broken, or a call-only phone number had a digit wrong. Check everything - making assumptions is often the worst thing you can do.
- Go To The Sales / Customer Support Team - If everything else fails, go to the sales/customer support team and ask to review their notes/calls from the last X days (however long the issue has persisted, or a few days before). Listen to the actual calls. Review the customer files. If you don’t want to spend the time (or don’t have that kind of time), upload the calls to an AI (like ChatGPT) and have it summarize them and find commonalities for you. The answers (and insights) I find from those conversations never ceases to amaze me – whether it’s learning that our benefits aren’t actually mapping onto their needs, or that their challenges are completely different from what we believed (don’t assume your audience research is right!), or that the features we think are right for development aren’t actually what they are prioritizing, or that our pricing isn’t as competitive as we thought.
Step #3: Improvise, Adapt & Overcome
While I’ve never been in the Marines, this is easily one of my favorite phrases to use when something is going wrong. It’s easy to blame a problem on some amorphous boogeyman (the market, the algo, whatever), throw your hands up, and wait for everything to work itself out. The issue with that approach (aside from the obvious - you’re being paid to fix it, not sit around) is that things might never work themselves out, or by the time they do, you’re so far behind your competition that catching up is near-impossible.
Instead, I demand our team adopt this approach in situations where - for whatever reason - we’re not getting the result we want.
Improvise: Systematically + creatively deploy your available resources in ways that maximize impact / have the highest probability of solving the problem. If (for example) Meta CPMs are spiking in certain markets, shift budget to another platform (like YouTube or X or Reddit) in those markets, or allocate more budget to traditional channels (like TV or Radio). If your offer is being dominated, shift your offer (if possible), or re-frame the comparison into terms you can win (i.e. turn a competitor’s low prices or steep discounts into a referendum on their quality or viability – the flip side to someone having to discount so steeply is that they’re not selling; the reason they’re not selling is probably because people don’t think their product is worth what they’re charging). Find the opportunity.
Adapt: One of my favorite movies of all time is Margin Call. In it, the legendary Jeremy Irons has a wonderful line: “There’s three ways to win in this business: be first, be smarter, or cheat. And I don’t cheat. And while I think we have some pretty smart people here, it’s a hell of a lot easier to just be first.”
Adapting is all about being first. When macro conditions change, it’s evolve-or-die time. The faster you can (1) diagnose the macro change (hence why Step #1 is critical) and (2) start the process of improvising + evolving in response to that macro change, the higher the probability that you’ll not just survive, but thrive. This is a core tenant of an anti-fragile organization: when system-wide adversity hits, you get stronger because you adapt faster.
There’s a fine line between trusting the gameplan/process and myopic fidelity to the process. I’m not saying change course every time you encounter resistance; but I am saying that you need to have a way to diagnose when your strategy is no longer effective.
Overcome: Henry Ford’s famous quip, “Whether you think you can or think you can’t, you’re right.” is one that I appreciate more the longer I work in this industry. No matter what brand or campaign or platform we’re working on, there’s always challenges that arise – and what separates the successful from the unsuccessful is the resilience in the face of adversity and the willingness to improvise, improve and adapt.
I’m not going to pretend that doing this is easy (it’s not) or that it’s fun (it really is not) – but it’s essential all the same. It is my sincere hope that the next time you feel that urge to blame the algo (or whatever), you open up this issue, bite the bullet and work to find and address the actual problem – not just blame an amorphous bogeyman.
Until next week,
Sam
PS. Go Birds.
Loving The Digital Download?
Share this Newsletter with a friend by visiting my public feed.
Follow Me on my Socials