Growth Fundamentals: 5 Steps to Improve Your Paid Social ROI
Your direct-to-consumer brand has survived - and maybe even thrived - in the past few years, despite the overall doom and gloom in the sector. You’ve grown to tens of millions in sales, and have all the fundamentals figured out by this point. And yet, I bet there’s some basic piece hiding in plain sight; if you could fix it, it would dramatically accelerate your profitable growth.
In the past few years, I had a chance to help a dozen consumer businesses accelerate their growth. In a series of blog posts, I’m sharing stories about the kinds of problems that, after solving, were “visible from space” for the financial outcomes of the relevant businesses.
This first post focuses on Paid Social, aka Facebook/Meta.
1. Improve New Customer CAC and LTV - not ROAS
Facebook’s job is to convince you to spend more money on their advertising channels. As such, every dashboard they have - every number they show you - is meant to make you believe that your spend is extremely effective.
Take ROAS (Return on Ad Spend), that is so front and center to Facebook reporting:
That “return” is revenue, not profit. So if you think that having ROAS of 1.0 or higher is a good thing, think again. In most cases, Facebook doesn’t even know the contribution profit of your sales, so it can’t tell whether you made money on a sale or not.
Facebook gives you little advice for structuring campaigns. Following their basic guidelines - “put everything into a single campaign!” - will result in you looking at costs of transactions intermixed between new customers and existing customers.
How do you combat this? By devising a framework for judging Facebook channel performance first, and only then thinking about tools you’ll need to get to the relevant numbers. Best practices I’d double check the team is following:
Make sure you can clearly tell how much you’re spending on new customers; versus active customers; versus churned customers. As a rule of thumb, spending advertising dollars on existing customers is not optimal. Instead, drive incremental sales from those customers using owned/free channels (ex. email and SMS).
Make sure you manage your new customer-focused spend in such a way that you can easily tell what portion of your dollars is going to prospecting (brand new audience) versus retargeting (showing ads to folks that have been to your website). If you don’t, Facebook will allocate a large portion of your new-customer-focused spend to retargeting, thus spending your ad dollars money on prospects that are likely to convert anyway - while taking credit for it.
Make sure you have more than one way of evaluating Facebook channel performance. 98% of people who see your Facebook ads do not click on them. And yet, there’s at least some long-term benefit to that ad impression, right? Find a way to capture it. Post-purchase attribution survey is a great start.
Understand the incrementality of your Facebook spend. Unless you’re specifically designing to avoid this, Facebook will take credit for transactions that would have happened anyway - without these ads. On/off tests; controlled up/down swings of spend; and geo experiments are some of the tools at your disposal.
Understand the lifetime value (LTV) of different segments. Different Facebook campaigns will produce wildly differing customer lifetime values. For example, feed ads for low-cost products will bring in the kinds of customers that will, intuitively, spend less with you in the future. So will high-initial-discount ads. Develop LTV proxy metrics and signal those to Facebook.
2. Ensure Creative Diversity
Facebook ad targeting mechanisms are quite different today than they were 18 months ago. Meta official guidelines say, “creative is the primary mechanism for targeting audiences.” Moreover, they state that homogenous / similar-looking ads will result in low scalability - and low efficacy - of the account overall. This is particularly relevant for the latest Facebook invention, Advantage Shopping Campaigns.
That is, for some prospects, seeing an influencer talk about your product will be convincing. For others, it will be the founder's story. For others, it will be an unboxing video, or your product in use, or an offer.
Examine your ads - and your competitors’ ads using the Ad Library tool - and see if you can try whole new ways of pitching your product. For example, for an apparel brand, these might be some of the angles for you to examine for your static images:
Lifestyle, professional static images vs influencer/UGC;
Feature callouts, detail shots;
Multi-shots of the same product, grids;
Close-ups vs full-length shots;
Offer focus versus new launch focus;
And that’s just for static images. On that topic: if you only use static images, or only use videos, definitely expand to the other side of the aisle.
3. Validate Technical Fundamentals
Setting up Facebook ads is a surprisingly technical topic. From setting up conversion events across the full funnel to making sure CAPI is set up right, there are gotchas that someone who hasn’t been doing Facebook every day for the past few years will miss out on. Resist the temptation of having a “digital generalist” manning all your advertising channels - they will miss at least one of these critical aspects, and your whole program will suffer by double digits.
For one of my clients, discovering that the test environment has been occasionally firing conversion events to Facebook - thus making it harder for the targeting algorithms to figure out who to show ads to - has been a key efficiency unlock. Mind you, this discovery happened after millions of dollars were already spent on Facebook ads; with brilliant people working on this for months. None of us are immune to this; validate your setup!
A few more technical gotchas that I’ve seen bite folks:
Customer list synchronization (ex. active/churned/cart abandoners). Telling Facebook about your customer segments is extremely valuable, but if those lists become stale, that’s arguably worse than not uploading those lists at all. Make sure you have a well-monitored sync pipeline (ex. Klaviyo integration works well here!)
Broken outside-of-platform attribution. Most mid-size shops do last-click attribution - that is, give credit to the advertising channel that was the source of the visit immediately before the purchase. This approach is fraught with peril: for example, tools like on-site email capture popups tend to “steal credit” from their original sources when looking at reports in Shopify and Google Analytics. Data muddying follows, preventing you from being able to evaluate the true benefit of your paid channels. Look out for a future post on marketing measurement on tips on how to handle this systematically.
4. Ensure Ad-to-Landing Page Alignment
Here are some learnings from real examples that otherwise successful, tens-of-millions-in-sales brands have fixed. Check your brand, too:
Is your ad showing a video of a beautiful jacket from the new collection, but clicking on that ad takes the user to a page where that jacket is buried three scrolls down?
Is your ad video talking about a specific health concern, but the landing page - and downstream shopping experience - says nothing about that concern at all?
Is your ad showing a best-selling product that by now has sold out, and the landing page has no real route for the prospect to convert?
Examples above are “crawl” in “crawl-walk-run.” Once you’ve taken care of those, here are a few “walk” level items to check for your landing pages:
If you have a mid-size consumer brand, and 90%+ of your paid social spend is pointing to a single page (ex. your homepage)...
If you haven’t experimented with landing pages in the past year, trying different landing pages for your top-performing ads…
If marketers can’t run A/B tests on landing page content without engineering team’s help…
… you’re likely leaving meaningful growth on the table.
5. Diversify Away from Facebook
In the past 2 years, most consumer businesses I know of have seen their Facebook ad costs rise. This may or may not be somehow related to Facebook stock price increasing.
Those that rely on Meta for 80%+ of their customer acquisition have, as a result, stared at dramatically worsening unit economics. The best ones learned their lesson, finding a way to reduce their reliance on Facebook by diversifying their media spend. There’s not a silver bullet that can easily replace it - but stitching together a half-dozen small channels to drop their Facebook reliance by 30-50% has worked for the brands I work with.
This is a very complex task - but that’s why your head of growth is paid the big bucks, isn’t she?
Some of the routes that worked for my clients:
Give a dollar to your customers instead of giving it to Zuck. Resulting conversion improvements may more than make up the difference.
Give a dollar to an influencer that “owns” an audience, instead of paying the platform for reach;
Try direct mail, podcasts, radio, affiliates, smaller social networks, partnerships. Yes, it requires meaningful effort and there’s a risk of spreading yourself thin if you aren’t careful. Find agency partners that can help you if you don’t have in-house expertise to experiment.
For larger brands, experiment with TV as a performance channel.
Finally… D2C companies that have been best able to weather the storm of rising CACs are the ones that have systematically invested into growing their customer LTVs. Lookout for the next blog posts in this series, more on that there.
I would love your feedback on this article - and on the rest of the Growth Fundamentals series. What has been most impactful for your business? What growth questions have you been embarrassed to ask? Email me, and I’ll help you with it.