ArticleThe Road To Ruin Is Paved With ROI
As the leaves start to change and the air gets a bit colder (at least where I am in the mid-Atlantic), the number of meetings appearing on my calendar with titles like “2025 Planning” and “FY25 Budget Preliminary Discussion” and “2024 Year In Review” increases at a truly astounding rate.
Now that I’ve attended hundreds of these meetings, I can say with near-absolute certainty that the vast majority of these meetings will include the following three core themes – the exact terminology might differ, whether it’s contribution margin or cost per lead or cost per SQL will vary based on the business, but the underlying macro themes will be the same:
- Leadership wants to increase top-line revenue by [X]% – 10%, 25%, 50% – depending on the stage/size of the company & the industry.
- They [Leadership] wants to achieve this top-line growth while maintaining or expanding the company’s current contribution margin (as a percentage of revenue).
- With rates high & a choppy economic outlook, money is harder to come by – so we must squeeze every ounce of value out of each dollar invested in growth. As a result, we’re going to be: (1) phasing out discounts and shifting our promotion structure, (2) increasing ROAS targets / decreasing CAC targets (instead of $200/customer, $175/customer is the new target) and (3) holding every single dollar accountable via [insert stupid third party attribution tool here].
At first blush, these seem eminently reasonable – most are straightforward, logical, (likely) reasonable and would be deemed fiscally prudent by any CFO, CEO or Investor. After all, who is going to object to trying to grow 10% or 25% while keeping costs in check? Why would anyone oppose more stringent accountability for growth dollars – after all, money is tight! In fact, many of the people in these meetings go one step further: anyone who pushes back on these objectives is probably afraid of accountability or not up to the job or simply is unwilling to put in the work necessary to help the company/brand succeed. The rationale that drives this contention is (ironically) hyper-logic: the plan makes too much sense. And since no one can see the flaw with the strategy, if the strategy fails, that failure must be due to the people who implemented it.
As someone who has been in (far too many) of these meetings, what I can tell you with absolute certainty is (1) that the above is almost exactly what is being discussed in these meetings and (2) that the above is – substantively – the entirety of what’s said in that meeting.
The strategy and the mechanisms required to achieve those goals are not considered, nor debated, nor discussed. The details are an afterthought.
But, as the old saying goes, the Devil lives in the details. Put another way: the road to ruin is paved with ROI.
For most marketers, that last sentence probably sounds categorically insane. It runs afoul of everything we’ve (collectively) been taught and told for the last 20+ years. And somehow, it’s still true.
It’s true because these conversations – and the directives that follow from them – are divorced from the underlying realities and their second-order implications. It’s simply a distant cousin of the always-popular “privileging the hypothesis” fallacy. Most economists or financial people would say that the plan outlined at the outset of this issue is reasonable and logical – and thus, if it fails to have the desired effect (growing the enterprise value while improving cash position), the flaw is with the execution, not the strategy.
There are two flaws in the above reasoning – and these flaws are the things that kill growth:
- The Assumption of Continuity + Extrapolation: all of us have been trained for decades – going back to pre-K or kindergarten – to view the world through a lens of continuity: what happened yesterday will happen tomorrow. In some disciplines – like astrophysics (the sun will rise tomorrow) and molecular biology (breathing oxygen won’t suddenly kill your cells tomorrow), these assumptions are good. But in most others – for example, marketing and human behavior – they’re worse than wrong. We’ve likewise been overexposed to continuity’s cousin, extrapolation: if I did 1.0*X and 1.0*Y happened, then if I do 1.25*X, 1.25*Y will happen. Extrapolation makes sense in deterministic systems – which is great for hard sciences (at least until you get into quantum physics, then things get wonky again), but less great in chaotic systems, like….all of human civilization and society.
- The Inability To Think Non-Linearly: most people are quite terrible at understanding non-linear growth and decay – for whatever reason, we struggle to grasp how seemingly small things can have outsized impacts if given enough time. This mental shortcoming typically manifests itself in things like loans (95% of people don’t understand interest rates), personal finance (the single-easiest way to become a millionaire is to invest $100/week into a low cost S&P 500 index fund from the time you’re 25 to the time you retire), compounding knowledge (reading 50 pages a day for a decade is the single-most-accessible way to 10x your earning power), even hockey-stick growth (every VC wants it, few understand it). But there are more subtle cases where small changes in underlying distributions have outsized impacts on businesses.
So, how do these two flaws apply to our upcoming board meeting (and the decrees that are bound to follow)?
The simple answer: the details.
The longer answer: play out the scenario. There are three goals: (1) increase revenue; (2) reduce discounts and (3) maximize the attributed return on each dollar invested in growth/marketing. How would you go about executing this?
My strategy would include the following:
- Hammer the Base via Email: Dramatically increase the number of emails + SMS messages sent to both current and prospective customers – email is near-free to send, and the potential costs (unsubscribes) are comparatively low to the likely benefits (a lot more sales that I don’t have to pay for).
- Re-Distribute Traffic: from a pure quantitative point of view, the most valuable impressions – defined as the expected value minus cost – most brands buy are retargeting. This makes good sense – if someone has already been to your website, they know you exist. They presumably have a problem, or may have a problem, or are otherwise interested in the problem that your company solves. That, over a sufficiently large sample, makes the people in your remarketing audience more valuable to your brand than the mass of people who have no idea that you exist, let alone what you do or what problem you might be able to solve. Thus, if my goal was to maximize return, I’d allocate much more of my budget to these campaigns.
- Deprecate Mid Maze, Non-Branded Search: On the other side of the cost/value equation is your mid-maze, non-branded search – think terms like “things to consider when [X]” or “most important features of accounting system for agencies” or “high-performing meta ad types” or “how to run youtube ads” – these have some commercial intent, but it isn’t immediate (unlike the end-of-the-maze search, where there’s an immediate link between click and revenue). Most of these campaigns don’t convert at anything resembling a high level, and they come with a moderate cost – so cutting these efforts has a negligible impact on your top-line revenue, but a moderate-to-high impact on your costs. Result? Higher marketing ROI.
- More Branded Search: Speaking of those end-of-the-maze searches, I’d definitely fully fund any branded campaign – these clicks have exceedingly high expected value and, due to your high QS for any branded search, relatively low cost. Easy money. Whether these are incremental is another battle for another day, but if our goal is just to make marketing’s ROI look as good as possible while juicing revenue, a stupendously well funded branded search campaign is going to be part of the plan.
- Lean Into Creative Winners: For what’s left of our prospecting, we’re going to dramatically reduce our creative output (saves time and agency/creative service fees), and instead run the prior top-performing ads.
- Affiliate Becomes The New Prospecting: Finally, since I do need some kind of new customer acquisition, I’d spin up some “affiliate” deals with a unique offer code + some kind of revenue share deal. Executive leadership would agree because, in theory, this de-risks prospecting as the brand is now only paying (both to the affiliate and the user) in the event of a successful outcome. Of course, that theory is seriously flawed – there’s an entire cottage industry dedicated to harvesting promo codes and distributing them to every person with Capital One Shopping, Honey, etc. installed in their browser. But, for the purposes of this effort, the desired result will be viewed favorably by both the Big Wigs and Triple Whale / Northbeam / whatever.
I guarantee that any reasonably-sized business that does the above will see a marked improvement in their top-line revenue, marketing ROI and EBITDA during the first 3-6 months, maybe even the first year.
Why? Because this entire strategy is geared toward extracting value from your existing pipeline and customers over a short period of time, not compounding value by expanding your customer base over a longer period of time. The end result of that is always the same: the active pipeline and customer pool shrinks as customers naturally churn and are not replaced with new ones (which were, historically, produced by the campaigns that were sacrificed in order to make the ROAS numbers go up).
The business (quite literally) consumes itself in an effort to grow.
The most common retort to this is: but that’s what the third-party attribution tool is for! We’re going to use Triple Whale or Northbeam or Terminus or Ruler or CaliberMind or whatever to ensure that we’re avoiding this disaster. These platforms will provide us super-granular data that tells us what’s working, so we can hold every dollar accountable!
Here’s the thing: this is exactly what every TPA provider wants you to think, because they know it conflates two core concepts: attribution and incrementality.
Attribution is the process of apportioning credit for outcomes to the marketing touchpoints that were involved in the outcome.
Incrementality is the process of identifying outcomes that would not have happened without the activity/channel.
Don’t believe me? Here’s an actual link to an “Attribution” guide from Triple Whale. Notice that the word “incrementality” does not appear on the page. In fact, the page includes this passage:
Which states (in very cloaked language) that TW has no earthly idea what drove the sale – or if it would’ve happened either way. All TPAs do is allocate credit for “marketing” conversions without regard to whether or not they would have happened in the first place.
Put another way: attribution tells you how many vegetables and chicken tenders and fries moved around your plate; incrementality tells you how many you ate. If the goal is to consume a well-balanced meal (and yes, chicken nuggets can absolutely be part of a well-balanced meal), which one do you want to know?
In this scenario, your third-party attribution tool is – at best – going to provide no incremental value to your marketing efforts, and serve only as a drag on your non-working dollars (TPAs aren’t cheap!). In working with many of these platforms, it over-indexes to branded search (duh), remarketing (double duh), and low-funnel non-branded search. That all sounds good in theory, but again, these are not base-builders; these are value-extractors. These channels aren’t creating new demand, they’re capturing existing demand.
This is the pitfall of these strategies – they all inevitably lead to a vicious cycle of forced competition at the most competitive times (remarketing, high-intent non-branded search), often in the most competitive ways (commoditization leads to a value + price-based decision-making, which in turn leads far too many brands to introduce larger discounts), for a progressively-smaller bucket of potential customers. While I’m (usually) loath to make comparisons between marketing and gladiators, that’s essentially what this becomes: a gladiator battle. And while you may win a few rounds, the reality is that 99% of gladiators never made it out of the arena. Maybe your brand is in that 1%.
The alternative, as Sun Tzu wrote, is to defeat the enemy before the battle ever occurs. In war, that’s usually via hsing (translated as a strategic advantage gained through superior positioning or deployment) and shih (loosely defined as a force multiplier that results in a material advantage – think of a crossbow or a catapult or a wall). Now, marketing isn’t warfare – but the same principles apply.
If your goal is to defeat your enemy (read: competitors), then the optimal way to do that is via superior deployment (marketing in the places where value is created but missed by TPA tools + competitors alike) and positioning (a focus on building a credible, trustworthy brand that can command a premium price without discounting), alongside force-multipliers (integrated strategy + data that focuses on balancing demand-generation and demand-capture across a range of touchpoints and channels).
So, how do we do that? How do we avoid falling into this trap and being forced to fight a losing battle in an arena we’re never going to leave?
The first – and best – way is to understand that the trap exists, learn to recognize it and be unafraid to call it out to your colleagues. The first time you do it, you’ll be wildly unpopular. People (especially the CFO) will not like you. That’s OK. You can take solace in the fact that you’re right and they’ll love you when time proves you wise (or they’ll rue the day when they ignored you and signed their own resignation letter).
Once you have that down, there are six things you can do and principles I’d recommend:
1. Set different targets for acquisition and retention:
Blended targets suck because they mask the underlying distribution – if I were to say that I reduced a brand’s Google Ads CAC from $200 to $100, most people would say that’s incredible. But the reduction isn’t as important as how I did it: if I accomplished this by reducing non-branded search by 80%, removing current customer exclusions and doubling down on RLSAs, I have (most likely) done more long-term harm than short-term good to the brand.
Every business needs distinct targets for new and returning customers, as well as for prospecting and remarketing. Your thresholds for prospecting should be higher than those for remarketing – after all, you should be willing to pay more to acquire net-new customers/clients/leads than you are to be reintroduced to people who have already met you.
2. Get Your Data Right:
This should not need to be said, but it needs to be said: before you start doing stuff, make sure you’re doing it based on the correct (or mostly correct) data. To the maximum extent possible, always optimize based on the data closest to people & profit – push for SQLs vs leads, for contribution dollars vs. revenue. The better and more accurate the data you can share with platforms, the higher the probability that those platforms will be able to deliver what you ask.
3. Insist on Metric Trios:
Most people have never heard of Goodhart’s Law, but it’s something every marketer should know. Roughly stated, it states: “When any metric becomes a target, it ceases to be a good metric.” Put another way, when we only have one metric, we optimize for it – no matter what. If you make ROAS or CPA the target, then the optimization for it likely comes with massive trade-offs (lower lead quality, over-distribution to existing customers, pushing more high-cost, low-margin products, etc.). The antidote to this is to insist on metric trios – composed of an efficiency metric (like CPA or ROAS), an absolute value metric (like revenue, contribution margin, or SQLs), and a predictive metric (a forecast or baseline/benchmark).
The tension between each of these metrics is what results in growth – if efficiency is pushed too far, the absolute value and performance relative to benchmark/forecast will crater (bad); if absolute value is pushed too high, efficiency will tank. This natural resistance is what allows brands to grow responsibly while mitigating waste.
4. Think Scalpel, Not Hammer:
If you read the above few pages, you’re likely coming to a conclusion that, as someone who works in marketing, I don’t want brands to cut marketing. Nothing could be further from the truth, but I could see how one might arrive at that conclusion. Waste happens in marketing all the time – but it’s rarely as obvious as a single channel or single campaign (though, to be very fair, I have seen a fair few examples of truly heinous waste of late). Instead of smashing with a hammer, your goal should be to methodically, but quickly, cut out as many chunks of waste from your efforts as possible – perhaps there’s an ad group with a long-standing, sky-high CPA or a set of KWs that have not produced a single conversion in a year. Start there. Audits are often one of the things most marketers scoff at the most, but the return on invested capital of one is often staggeringly high – after all, who wouldn’t pay $20,000 once to be able to reduce a $500,000/mo ad spend by 10% with no impacts to revenue? Pay $20,000 to save $600,000 a year? Yes, please. Just as a surgeon (generally) doesn’t cut into a patient hither and yon, so too should a marketing director or CMO not cut into a marketing machine willy-nilly; a well-done audit is like the x-ray or MRI before the surgery.
5. Do The Unexpected:
Often, in leaner economic times, brands resort to discounting, bundling, value-based messaging and service/quality reductions in an effort to grow the top-line (discounting, bundling) and the bottom-line (reductions). Customers – whether they’re businesses or individuals – expect that.
This expectation creates opportunity for brands willing to do the opposite – raise prices. Improve service quality. Highlight the insane details. Restrict service/product lines. When everyone goes in one direction (cut), the opportunity is found by going in the other (increase). Don’t get me wrong, you’ll sound patently insane for suggesting that you should be more exclusive, raise prices and redouble your customer service efforts – after all, that’s quite irrational in the context of our current economic realities.
But, then again, there are plenty of studies that suggest that higher prices actually increase demand (in industries where quality is difficult to ascertain, price is often used as a proxy). The same is true for restricting supply – the harder something is to get, the more even the savviest of people want it (just ask yourself why some of the most sophisticated investors in the history of capitalism tripped over themselves to put money in Madoff’s fund). And when customers are beyond frustrated with your competition’s lack of service, what becomes the most important criteria in their selection of a new brand/partner? Service quality.
6. Obsess About the Delta:
Reduction is often the central character in every “efficiency” story – what can we cut, how can we reduce this expense, where are we overpaying. Make no mistake: these are valid questions. But they are, at best, half of the equation.
The ignored other half of the equation is the value side: what is produced by those dollars out? I’ve heard countless marketers say things like, “LinkedIn CPCs are so high – why would I pay $25/click when Meta is only $2.50?” The answer, of course, is the expected value of those clicks. I’d argue that the cost per click / cost per 1,000 impressions is largely irrelevant – what matters is the value those clicks or impressions produce and the delta between the cost and the value.
Consider this (anonymized) conversation with a CMO from a few months ago:
CMO: We want to turn off Google Search – the CPCs are nearly 5x higher than our Meta CPCs, and [attribution tool] says that these are converting at the same rate to leads.
Marketer: OK, but before we do that, let’s pull a lead analysis to confirm these leads are all equal in terms of the rate at which they became MQLs, SQLs, Proposals & Closed/Won.
CMO: Our leads are pretty uniform
Marketer: The analysis will be quite quick then, and if it confirms it, we’ll pull the plug on PPC.
CMO: Fine.
A few days later
Marketer: We’ve pulled the lead analysis, and it does appear that Meta has produced substantially more leads than Google, at a cost per lead of $92 vs. Google’s $327. You were also correct that the CPCs on Google were about 5x that of Meta – $11.78 vs. $2.17.
CMO: I knew it. This was a waste of time. Pull the plug.
Marketer: That would be a terrible idea.
CMO: You have about 30 seconds before I fire you.
Marketer: During the last the ~500 leads generated by paid media in the last quarter, ~390 have come from Meta and ~110 from Google. ~18.4% (92) have attained SQL status.
CMO: OK, so?
Marketer: Of those 92, 80 came from Google. Google’s cost per SQL is ~$449.63. Meta’s cost per SQL is $2,990. And if you go by pipeline value, Google’s contract value is roughly 20x Meta’s: $82,500 vs. $1,624,000.
CMO: I deserve a raise for not firing you.
The moral of this story is that – especially when efficiency is top-of-mind – the best marketers keep their eye on value deltas. The cost of something is only relevant in the context of the value it provides. Marketing terminology lends itself to fungibility quite easily, despite the fact that what’s being compared (in this case, clicks on Meta vs. clicks on Google) are not the same. Explaining this to a CEO or CFO is often quite difficult, so I typically resort to currencies: there are many different “dollars” – there’s Canadian Dollars and East Caribbean Dollars and Australian Dollars and New Taiwan Dollars and New Zealand Dollars and Brunei Dollars and US Dollars – all of which have different values. Everyone would think you’re categorically insane if you were willing to exchange one New Taiwan Dollar ($0.031) for a USD ($1.00) because they were both “dollars” – yet marketers are all-too-willing to exchange a Google Ads click for a Meta Ads click. Obsess about the delta.
“Cost is only half the equation. At the end of the day, growth is the objective, and growth comes from the delta, not the revenue or the expense.”
From all of my conversations with CEOs, Investors, CMOs, agency owners and our clients, I know there are 1,000s of conversations like the one I mentioned at the outset happening over the next few months. My sincere hope is that this gives you some insight into the second- and third-order consequences of following through on these proposals, as well as ways you can constructively push back and save some CEOs & CFOs from themselves.
Cheers,
Sam
