Happy Sunday, Everyone!
As both 2025 budgets and conference season have kicked into high gear, there’s been one question I’ve heard more than anything else: what should we be doing and what should we be avoiding in order to hit our growth/profitability goals for 2025?
It’s a question I spend a significant amount of time thinking about, both for our agency and our portfolio companies, as well as for our clients. And, over time, I’ve developed a running list of things to avoid (the “Nots”) and things worth thinking about (the “Hots”) that I want to share with you this week.
I’ll warn you that (like most weeks), some of these will be unpopular. A few will likely get me un-invited from a few SaaS-sponsored conference dinners. But I genuinely believe that - for the vast majority of brands - avoiding the “Nots” and doubling down on the “Hots” will be a net-positive for your business in the coming 12 months.
In order to keep this manageable, I’m going to split this into two issues: this week, I’ll focus on the “Nots” (the things that are overvalued + should be discontinued); next week’s issue will be all about the “Hots” (i.e. the things that are undervalued and you should do more of or consider starting).
With that, let’s get to it:
The Nots:
Programmatic Display Advertising
The dirty not-so-secret about 95%+ of display advertising (especially Google Display Advertising) is that it is absolute garbage. The easiest way to tell: go pull up your placements report, then visit the sites yourself. Read the content. Check out the videos. Download one (or two) of the not-so-shady-looking apps. Go through the experience. Then tell me if you think it’s worth deploying your brand’s dollars on those sites.
The reality is that virtually every programmatic vendor is overstating ad impact via long view-based attribution windows (some 60 or even 90 days – meaning they’ll take credit for the overwhelming majority of your conversions), fraudulent clicks/impressions (yes, it happens) and the perpetually-nefarious “automated” targeting (this is the digital equivalent of mixing Vladimir vodka with Flint, Michigan tap water: it takes something that’s already pretty bad (most programmatic impressions) and mixes it with something that’s somehow worse (irrelevant placements)).
Does that mean that all programmatic advertising is bad? No. In the right context, with strict placement controls informed by sound audience research, the right measurement setup and robust creative, display ads can work. But: the vast majority of brands are still better off spending their next dollar on Google Search, Meta Ads, TikTok or YouTube – not programmatic display.
Scaling Pricing Model SaaS
Variable-based-on-your-success models (i.e., percentage of revenue, percentage of GMV and percentage of ad spend) are the worst. I tolerate them for a few exceptional platforms (like Optmyzr), but my advice to our portfolio companies is always the same: avoid them like the plague.
The feedback I always hear when I say this is the same: but what they’re offering is appealing: “We’ll keep your fixed costs low and grow with you!” or “You only pay for what you use!” or “The only way you pay more is if you’re growing.”
I’ll grant that this sounds nice. It’s tempting (and often, imperative) for early-stage businesses to keep fixed costs low. But these platforms are running businesses, not charities. They aren’t keeping costs low at the start out of the goodness of their hearts; they’re doing it because they know that once they’re embedded in your tech stack, the incremental cost of extracting and replacing their tech with a fixed-cost vendor will likely not make financial sense.
The best advice I can offer to any company considering a variable-fee SaaS platform is this: defend every basis point of margin you can. You are better off biting the bullet and paying a flat fee (or even a tiered fee based on number of users/seats) than you are giving up 1%, 2% or more of each and every dollar you bring in the door. It’s difficult enough to scale your business without leeches in your tech stack.
“Awareness” Meta Ads Campaigns
Awareness-oriented campaigns (i.e. reach & frequency, traffic) have come back into vogue recently – I’ve seen more Meta accounts running them in the last three months than I think I saw in the preceding twelve.
The rationale behind these tends to be pretty simple: the CPMs are lower, and they (allegedly) allow brands to reach “latent intent” audiences (roughly translated: people who aren’t in market for whatever it is you’re offering, but will miraculously be compelled to buy whatever it is you’re selling upon seeing your ad in their feed).
As with anything, there’s a time and a place for awareness-focused campaigns – but that usually starts at about $12M-$20M/year in Meta Ads. If you’re not spending that, you’re better off optimizing for the closest-to-people-and/or-profit conversion objectives (qualified/converted leads, net-new customer sales, etc.), and using your targeting/offers/angles to unlock new audiences.
Event Marketing & Sponsorships
Let me start by saying that this is NOT directed at brands that have an intentional, well-researched, consistently-executed event or sponsorship strategy. I am a strong proponent and believer in the power of experiential marketing.
But what I see from the vast majority of brands is not that - it’s ad hoc sponsorship of some random event/team/production/influencer with no support, accountability or strategic intent.
We recently spoke with a brand that sent well over $100,000 (cost basis - which (for this brand) translates into $300,000+ in MSRP value) in free products to various influencers, shows, groups, etc. over a multi-year period. What did they get back? Absolutely nothing. Not just no sales - but we struggled to even find tagged posts or photos from the events where the products were (allegedly) featured.
While this may seem like an extreme example, it is anything but – even smaller brands are inundated with requests for partnerships, sponsorships, event support, etc. A dinner here, some free product there, a sponsorship for a little league team, an “advertisement” in a diner menu there – add it all up, and $50,000, $100,000 or more has walked out the door. That’s real money that could be used to fuel real growth.
Most Loyalty Programs
Loyalty programs are on my $#!% list twice: first because they (typically) fall into the “scaling pricing model SaaS” category, and second because - on top of getting progressively more expensive - they get progressively less profitable as you scale.
Here’s the simple reality: customers aren’t buying your [facial cleanser, creatine gummies, single origin coffee, built-it-yourself furniture or whatever else] because of your loyalty program. The points (or whatever else) you offer to the people who religiously use your products are not the reason they use the product – and moreover, the overwhelming majority are not price sensitive when it comes to your product.
I recently re-ordered my favorite body wash (as I do pretty regularly), and was prepared to pay the usual cost (I think it’s $50 or so for a couple bottles, shipping, taxes, etc.) - only to be alerted at checkout that I had enough points to cover the cost of this order. I (of course) applied those points and paid nothing. Was it a pleasant surprise? Absolutely. Was the potential for a future free order the reason I placed the preceding 15 (or however many) orders? Absolutely not. Would I have still made this purchase without the points? Yes! All this loyalty program succeeded in doing was costing the company money.
That’s the point: loyalty programs provide the wrong incentives to the wrong people.
If I’m a loyal user of your product, I’m going to buy it and I’m going to use it regardless of whether or not you give me a discount. The same can’t be said for first-time users (maybe they do need some kind of incentive to de-risk an initial purchase).
A better use of these dollars is to focus on either (a) converting your existing loyal customers in one product category (say, body wash) to another (like cologne or face cleanser or deodorant) or (b) allowing your loyal customers to “gift” a friend a trial product.
Performance Max Advertising
As a Google advertiser, my last three years of communication with Google have been filled with a non-stop refrain of, “Try performance max (“PMAX”) - it’s the best!” Well, I’ve tried it. I’ve used it in eCommerce, I’ve used it for B2B + B2C lead gen, I’ve used it for SaaS, I’ve used it for donations - you name it, I’ve probably done it.
The same thing that’s touted as the benefit of PMAX is its downfall: it tries to do too much. If you’re unfamiliar, PMAX is a campaign type that combines all of Google’s available inventory: search, shopping, display, local, YouTube, discover) into a single, fully-automated campaign type. The end result? A campaign that is the equivalent of unleashing a swarm of locusts into your ad account.
PMAX cannibalizes everything else - from your search campaigns (you’ll see PMAX ads serving for search terms where your existing search/shopping campaigns already have excellent quality scores, staggeringly good historical performance and ample budgets) to serving in place of your hyper-relevant YouTube and Demand Gen campaigns, despite inferior performance. And the cherry on top? The presence of display ad units in PMAX means that all of your excess budget can be dumped into those ad units with (virtually) no accountability.
The best analogy for PMAX comes from The Big Short: it’s fish stew. It’s all of the inventory Google can’t sell anywhere else, thrown together into overly generalized affinity/in-market groups to jack up the CPCs and passed off as something that’s beneficial.
As with anything, there are good uses for PMAX. Feed-only PMAX (as one example) can be a semi-decent approximation of smart shopping (which was good back in the day). Page-feed powered PMAX can be great for hyper-niche brands.
Uber-Polished (read: Expensive) Creative
I’m perpetually amazed and depressed by the number of brands that invest hundreds of thousands of dollars per year in uber-polished, hyper-produced creative. While there’s nothing wrong with making a nice-looking ad, the reality is that the more expensive the ad, the more difficult it is for that ad to break even. This is just math.
If you spend $10,000 on a single ad for a product that produces a contribution margin of $10 per sale, well, you need to sell 1,000 units just to cover the cost of creative (before you get to the making profit for your business part). That’s a high break-even point when talking about a number of sales - one that many ads won’t get to on Meta or Google or YouTube.
Contrast that with an ad that costs $500 to produce – sure, the quality is lower (well, maybe - a decent creative with today’s AI-enhanced tools can do some cool stuff) – but the break-even point is just 50 sales.
The response I typically get when I point this out is a variant of, “Well, we’re not just paying for the fancy creative; we’re also paying to have someone who knows what sells and understands how to create ads that are going to work at a much higher rate.” And, as with many other sales pitches, this one sounds both appealing and compelling.
But let’s break it down.
Say that your fancy creative has a “hit probability” 10x higher than my quick-and-dirty “cheap” creative. Based on a broad set of data, ads hit (defined as an ad that drives 25 conversion events in a 30-day period, spends 25x your target CPA AND produces those 25+ conversions at or below your target cost during that period) at a rate of 5% - 10%. So, for every 100 ads in an account, 5-10 will turn into hits. To be as conservative as possible, I’m going to use the low end of that range: 5%.
That equates to a cost per hit of $500/.05 = $10,000 for the “cheap” creatives. Not great, but certainly not terrible.
Now, do the same math for the “expensive” creatives: $10,000/.5 = $20,000.00.
Despite the hit rate of the fancy creative being 10x higher than my cheap creative, the cost per hit is still 2x the cheap creatives.
When it comes to creative, you’re better off focusing on creative velocity + volume: quantity drives quality. If you’re curious about this particular topic, check out this issue.
Attribution SaaS / Software
Over the years, I’ve bashed third-party attribution heavily and repeatedly. This will be no different, because the underlying reality hasn’t changed: multi-touch attribution (or MTA) is nothing but a colossal waste of money for the overwhelming majority of brands, for a host of reasons:
- High Costs - attribution platforms aren’t cheap, and (as a bonus) they also tend to fall onto my “variable pricing SaaS” $#!& list. The real kicker here is that most free tools (like GA4) provide data and insights that are as close to accurate as the TPA tool, at NO cost.
- Limited Accuracy - third-party attribution models rely on fragmented data sources – something that has only gotten more difficult as privacy regulations have forced TPAs into using more synthetic and/or modeled data.
- Platform Bias & Walled Gardens - every major digital ad platform (Meta, Google, Amazon, TikTok, Microsoft) are walled gardens: they don’t share cross-platform data with third-party tools. This makes it virtually impossible for TPA tools to assemble an accurate picture of what’s happening in each channel and across channels.
- Overemphasis on Multi Touch Attribution - every attribution tool pushes their own “proprietary” model that attempts (poorly) to assign a value to each interaction/touchpoint in the customer journey. As you can imagine, this can get quite complex, quite quick – and that’s exactly what happens: the TPA spits out a massively complex picture of the customer journey with little-to-no actionable insight, resulting in the marketing equivalent of a gold-plated book report.
- Conflates Attribution with Incrementality - the holy grail of marketing is incrementality (defined as positive outcomes that would not have occurred but for the marketing) - but (as even Triple Whale will sheepishly admit), that’s not what TPA tools provide! All they do is divvy up “credit” for what happened, with no regard to whether or not that thing (channel, ad, tactic) had any causal relationship to the outcome. Those are two fundamentally different things.
- Can’t Add Relevant Data to the View: in many industries where TPA is being pushed (i.e. eCommerce, senior living, legal services, etc.), there are significant amounts of offline + inaccessible data (i.e. referrals, personal interactions, in-person engagements, health condition data, etc.) that has a disproportionate impact on the overall outcome, but can’t be captured by the TPA tool. The end result is the digital equivalent of trying to guess the fine details of a painting with only the frame present.
- Disintermediates Media Buying Decisions From In-View Algorithmic Data - Finally, and most dangerously: TPAs disintermediate media buying decisions from the data available to the platform. That may not sound like a big deal, but it is. To understand why, consider this example: imagine you’re a student who submits assignments to your instructor every single day for weeks or months on end. During this time, you don’t hear a single thing - no comments, no grades, no feedback – until suddenly you’re taken to a remedial class or promoted to an advanced track. You would likely be (understandably) flabbergasted – what resulted in the change? What assignment was good (or bad)? Why weren’t you told about what was happening and given a chance to improve? That’s a (rough) approximation of what’s happening to the ad platform when TPAs are used to make media buying decisions: the platform doesn’t have the data in-view (like the student doesn’t have access to the grades/feedback from the teacher) that is being used to make major decisions (i.e. adding or subtracting budget, pausing campaigns, etc.). Of course, the platform (like the student) can try to guess which assignment(s) led to the result, but without the in-view data, making that determination is (at best) a crapshoot and (at worst) a fool’s errand.
As with everything on this list, there are some edge cases where attribution tools can make sense – but those are few and far between. The overwhelming majority of brands are better served taking the money they’d spend on a TPA tool and investing it in something that will actually help them grow.
And speaking of that - if you’re curious about what platforms/tactics/strategies I’m bullish on, make sure you check out next week’s issue – it has 12 “hot” tactics that are (in my opinion) undervalued and overlooked by too many marketers.
Until then, have a great week!
Cheers,
Sam
Loving The Digital Download?
Share this Newsletter with a friend by visiting my public feed.
Follow Me on my Socials