Article

3 Things To Do More & 3 Things To Do Less

Analytics Data Ecommerce Marketing Testing

November 13, 2023

It’s difficult to believe that it’s actually November – BFCM is around the corner, 2024 budget planning is in full swing, and the Charles de Gaulle airport Starbucks (where I’m writing this) is changing their posters from pumpkin spice lattes to Gingerbread-based concoctions.

I’ve spent much of the last week in Bucharest, where I was invited to be the closing keynote at GPeC, the largest eCommerce conference in Central & Eastern Europe. Across many conversations with marketers, store owners and vendors/partners, one topic kept coming up: what should we be doing better? Where are the opportunities to thrive in the next 12-14 months?

While I don’t have a crystal ball, I do have a few ideas. I’ve broken them up into two, 3-item lists: what we should do more of, and what we should do less.

3 Things I’d love to see more:

Real, Financial Accountability:

I’m leading with this one because I believe it’s just that important.

The simple reality is that most marketers have benefited from smooth seas for the vast majority of their careers (Q2-Q3 2020 exempted) – growth was largely up-and-to-the-right, interest rates low and regulation scarce. Most of us have only ever known good times. But things change. Interest rates are at 20/30/40 year highs (depending where you live), and there are strong indicators they may be going even higher in the short-to-mid term.

Economic growth masked a lot of sins and stymied a significant amount of the philosophical evolution that we needed as an industry: evolution from (always-bogus) attribution to true, big-picture measurement; evolution from vanity metrics to real financial measures. Make no mistake: we will get there as an industry. That is inevitable. The smart money is always on the optimists, because they’re the only people who ever change the world (and, if I can be sure of anything, it’s that the world will change).

From a strategic and tactical perspective, this leads me to an inevitable conclusion: the organizations and people who actively invest in the tools and skills to connect marketing activities to financial outcomes will create a significant advantage.

This goes beyond the basics of accurate revenue numbers and relative lead values – those are things most advertisers have (at varying levels of accuracy).

I’m talking about metrics such as contribution margin, net-present LTV, risk-adjusted marketing return rate (MRRa), Marketing Efficiency Ratio (aMER + rMER), lead velocity and net-present pipeline value – all of which connect to both your forecast and your P&L.

I’ve written about the idea of net-present metrics since 2016. They’re still not used to the level they should be – but then again, when the risk-free rate is <2%, maybe they aren’t necessary. But when the risk free rate is pushing 6%, that’s another thing entirely.

If you’re curious, check out this article from SEL about 3 years ago, that spells the whole thing out.

Growth is going to look a lot different in 2024 than it did in 2022. Having real financial metrics and real financial accountability will enable you to capitalize on the opportunities that will be there for the taking – as well as stop doing the things that are currently hurting your business.

Robust Data Infrastructure Investments

Let’s just cut to the chase: Data is the only optimization lever that matters. Every ad platform is becoming more black-box. It will not stop; if anything, it will accelerate. Marketers and agencies who focus on lever-pulling all the live-long day will soon find themselves with fewer levers to pull and fewer results to speak of.

The simple reality is that the best data wins.

You don’t get the “best” by relying on the same in-market, affinity and interest audiences that everyone else uses – there’s no advantage there. If you do the same things as everyone else, you’ll get the same things as everyone else. Pretty simple.

Fortunately, platforms have provided progressively more powerful mechanisms to create data advantages:

Enhanced/Offline Conversions (CAPI for Meta) – these are must-use features at this point; there’s simply no better and more accurate way to connect outcomes on your site with both the individuals who completed them and the business results your (or your client) achieved as a result. If you’re curious about these, I highly recommend you check out this issue, which provides an in-depth analysis of each of these methods.
Custom Segments + Combined Segments – Google Ads – to their credit – has invested substantially in building more robust audience targeting; Custom + Combined Segments are examples of that work. The ability to build custom segments of people who conducted specific Google + YouTube searches is powerful for prospecting, as is the ability to layer together your data (in the form of customer lists, website/app audiences, etc.) and Google’s data.
Lookalike Audiences (LALs) – I’m still a massive fan of LALs, if they are done right. This means creating segmented seed lists by product/service/offer type (and perhaps, even more segmented by various attributes, such as income or title – if you have a sufficiently large customer/prospect base). LALs are incredibly sensitive to initial conditions, so the more homogenous the seed audience, the more accurate the LAL. The more accurate the LAL, the higher the probability that it will provide a data advantage moving forward (i.e., the people included in the audience will more closely resemble the seed).
Business Data Uploads & Value Rules – the three methods above are all focused on customers + prospects; these two are focused on your business. The ability to provide Google/Meta/Microsoft with more information on your business – whether that comes in the form of relative customer values (a value rule) or additional data about your business (business data).

At the end of the day, the importance of good data for digital advertising is only going to increase; investing in the systems required to capture, clean and connect that data to your ad platforms is one of the single-most-impactful thing brand and advertisers can do today to future-proof themselves again the relentless march of ad platforms toward automation.

Proper Account Structure & Diligent Management

As stated above, growth masks a wide array of sins. This was true for brands, and it’s also true for accounts. For the last 3-5 years, the combination of relatively cheap attention + increasingly online populations meant that most ad accounts performed well relative to legacy tactics, regardless of their structure. The simple version: advertisers could do whatever and produce results; those results were likely suboptimal, but still pretty darn good. And pretty darn good was good enough.

But now?

CPMs are way up. Competition on digital ad platforms is fierce. And platforms have compelling incentives to crush surplus value – after all, every dollar of surplus value that an advertiser captures is an inefficiency. There is a reason ad platforms resort to tactics like obfuscating incremental costs; it isn’t because they’re evil, it’s because they are self-interested.

The right response to this is to protect your downside while keeping your upside as close to unlimited as possible. The methods by which you do this are threefold:

Data – the better your data, the higher the probability that you’ll gain a competitive advantage from it.
Account Structure – while there is no “perfect” account structure, there are objectively better (and objectively worse) ways to structure ad accounts. Some indicators:
- Machine-Learning Friendly – ML thrives with larger, homogenous datasets. The days of hyper-segmentation are over. A well-structured account segments only when necessary; that may be due to vast differences in expected value or performance (such as branded vs. non-branded, or customer list vs. prospecting), or due to vast differences in target audience (i.e., different services). Bottom line: segmentation comes at the cost of data fracturing; sometimes, that cost is worth paying, other times, not so much.
- Cost Controls – I can’t overstate the importance of cost caps / bid caps / tCPA in ad accounts today. There is simply no better defense to ad platform surplus value capture than cost controls. Properly implemented, these provide something unheard-of in markets: genuinely free insurance. An account with proper cost controls + unconstrained budgets can dynamically adjust to changes in supply without overspending; that’s true value-capture.
- Proper Exclusions – what you exclude is more important than what you include; ensuring that you’re providing ad platforms with the right exclusions (and updating them over time) is paramount. I can’t emphasize enough how powerful exclusions are – especially in an ecosystem where positive targeting has lost its specificity.
Proper Management – this one has been overlooked for far too long; the era of set-it-and-forget-it ad platform management is largely over.

Digital advertising is very much a battle against gravity – with the gravity being platform’s tendency to crush surplus value over time. The role of the advertiser, then, is to defy gravity – to continually find new sources of surplus value that can be leveraged to deliver exceptional results for their business (or their clients).

So, what does this look like? It looks like iterative audience testing and keyword discovery; it looks like identifying top-performing placements (contextual isn’t dead!) and excluding garbage placements; it looks like continually finding new ways to create + share better data with ad platforms; it looks like an “eyes-on, hands-off” approach to updates.

There’s no one-size-fits-all for digital ad account management. But there are objectively better ways to do it.

3 Things I hope We All Do Less:

Constant Ad Account Tinkering

Let me cut to the chase here: you’re not helping anyone by making daily “optimizations” in your ad accounts. All you’re doing is confusing the machines, slowly learning and ultimately shooting yourself in the foot while trying to run a marathon.

Instead, focus on the following:

Batched Changes – review your accounts every day; note changes that should be made, batch them together, and push them to the ad account before your lowest-performing day each week (every business has a day or two that is objectively worse than others).
Immediate Interventions – in the event something is truly wrong (for instance, the ad platform has gone off on a random tangent that is objectively terrible), intervene immediately and exclude. Limit your intervention to the narrowest-possible options; you’ll do more by doing less.
Rely on Structure – fundamentally, your structure, data and financials (from the first section) should be doing 80% of the work in driving value from your ad account; the remaining 20% is your management. If you’ve set the proper targets and are passing back the right data, machines will auto-correct pretty quickly; after all, they have an incentive to do so.
Embrace Short-Term Variance – I’m not saying that every day is going to be filled with sunshine and rainbows – there will be days (even if you’re doing everything I’ve written here) where an ad platform fails to hit your targets. There will be days where it blows targets out of the water. The urge to do something after a bad day can be overwhelming – but your job is to resist it. Your job is to take a step (or two) back from the keyboard, look at the bigger picture, ensure your targets are set correctly…and trust the machine to course-correct. Intervene only when necessary.

So, the next time you’re tempted to go log-in to your Meta Ad Account to make some quick “optimizations” – just don’t.

LLM-Generated Content

Over the past 3 months, I’ve seen an explosion in LLM-generated content –from spammy blog posts to engagement-bait social content to weird DALLE-2 images.

I’m genuinely impressed at how far LLMs have come in just under a year. I’m somewhat surprised at how rapid adoption has been. But I’m a contrarian by nature. When I see the vast majority of people running in a specific direction, I look for the opportunity to go the other way.

The simple reality is that LLM-generated content consolidates in the middle of the bell curve: it’s around average (maybe it’s a C+ or a C- or even manages a B with a great prompt and some editing). It’s representative of the internet as a whole.

There’s no durable advantage in being average. Mediocre isn’t a differentiator, it’s a cry for help.

Aaron Orendorff put this better than I ever could nearly a year ago, when the ChatGPT hype was just starting to build:

He was right then. He’s right now. The value of truly remarkable, brilliant, so-good-it-hurts content has never been higher. So do that. Invest in exceptional writers, creators and storytellers.

Note: I’m not saying LLMs / “AI-generated” content doesn’t have a place in marketing’s future – it 100% does. Recart’s AI-driven SMS is a great example: it takes your content and tailors it, using your data + parameters, to create personalized texts. That’s a good use case. Code Interpreter is a great use case. And I’m sure there are a million more.

But true, genuine creativity is – and will be for the foreseeable future – the domain of people. And the reason for that is simple: disruptive, attention-demanding, thumb-stopping creative isn’t created by following a pattern or reducing a question to a statistics problem; it’s created by breaking a pattern in a particular way. Machines that rely on patterns don’t break them. People do.

Overwhelming Emphasis on Acquisition

Last, but certainly not least: I think we’ve all (collectively) focused far too much on acquisition at the expense of customer loyalty + retention. As we enter a lower-growth economic period, acquisition will become more expensive, more competitive and more difficult. The brands that survive will be those who can keep, upsell and earn the loyalty of their existing customers.

This doesn’t mean neglecting acquisition, but it means re-thinking the balance between the two. This looks like:

Eliminating Loser Customers – This may seem like a strange one to start, but I assure you, it makes perfect sense. The simple reality is that every brand has “loser” customers – the ones who drive negative contribution dollars, the ones who place inordinate strain on your resources, the ones who take, take and take without ever providing value. Cutting these customers is what frees up the resources you need to delight the rest of your customers.
Canceling Discount Culture – I genuinely don’t understand most brand’s insistence on discounting – especially among existing customers. All this does is train your customers to expect those discounts, which cuts into your margins and hurts overall financial health, all while subverting your ability to create genuine differentiation long-term. Discounts are fine once in a while, but if your entire retention strategy is emailing people coupons, maybe rethink that.
Creating Community – I genuinely believe that people don’t join brands; they join people. The brands they choose are just outward signals of what kinds of people they want to join/attract. And that is where the opportunity lies: creating communities where like-minded people can participate and join other people. I wish more brands invested in genuine community building (which is really, really difficult!). This is more than having a FB group or a Slack channel or a discord; it’s cultivating genuine connections; it’s providing real, tangible value to members; it’s listening 10x more than you sell. That’s often a hard pill to swallow for executives (especially up-front), but the payoff for doing it well is massive.
Email + SMS Are Journey Steps – Finally, I wish more brands would embrace the reality that email and SMS are just part of the journey. The more it feels that way, the better they will perform. Stop the email garbage. You don’t have to spam your entire list daily with another 30% off coupon. You don’t have to test emojis in subject lines because Neil Patel or eMarketer said so. Just treat your actual customer communications like part of the journey and you’ll be fine. Segment your list, create remarkable flows, and only send blasts when you have something so valuable, so awesome, so time-sensitive that you’d be doing your list wrong by *not* sending it.

Customer retention is where a LOT of growth will happen in 2024 – and the combination of smart, cost-controlled acquisition with brilliant retention is the one that will win the day.

Thanks for reading!

Cheers,

Sam

Related Insights