ROAS is a vanity metric. I realise that is a provocative thing to say about the number that most DTC brands use as their primary measure of marketing success.
But after managing significant paid media spend across DTC and e-commerce brands, it is the most honest thing I can tell you about the state of how this industry measures itself.
That does not mean ROAS is useless.
It means treating ROAS as your north star — the number you use to decide whether your marketing is working — is one of the most expensive mistakes a DTC brand can make.
And guess what?
Most brands are making it every single month.
This post explains exactly why ROAS is a vanity metric, what it is hiding from you, and the five metrics that tell the truth about whether your DTC marketing is actually building a business.
Why ROAS Is a Vanity Metric: The Attribution Problem Nobody Talks About
Here is the scenario that plays out in DTC ad accounts every day.
A customer sees your Meta ad on Monday.
Clicks it, browses, leaves.
Searches for your brand on Google on Tuesday.
Clicks a search ad, adds to cart, abandons.
Gets a retargeting ad on Wednesday.
Ignores it.
Types your URL directly into their browser on Thursday and buys.
Now, let me ask you:
Which channel gets credit for that sale?
In a last-click attribution model, Google gets 100% of the credit.
In Meta's own reporting, Meta takes the credit.
In your retargeting platform, the retargeting ad claims the conversion.
You have one order worth, say, £80. Your dashboards have reported it three times across three platforms.
Your blended reported ROAS looks strong. Your bank account does not match.
This is not an edge case. This is how platform attribution works by default and it is why ROAS is a vanity metric that consistently overstates the performance of every paid channel simultaneously.
Industry analyses show platforms like Meta and Google frequently over-attribute conversions compared to independent tools like Shopify or Google Analytics. For example, Meta has recently updated its ad measurement system specifically to reduce discrepancies with third-party analytics industry analyses including recent Meta measurement updates.
The same sale, claimed multiple times, is not a reporting quirk. It is structural.
iOS privacy changes compounded the problem further. After Apple introduced App Tracking Transparency in iOS 14.5, AppsFlyer's data shows that IDFA availability dropped from 80% to 27% — meaning device-level attribution signals are simply unavailable for the majority of iOS ad interactions.
Platforms responded by filling the gaps with modelled conversions: statistically estimated outcomes rather than directly observed ones.
Your ROAS figure increasingly reflects what a platform's algorithm predicts happened, not what your Shopify backend confirms happened.
ROAS is the metric that makes your marketing look better than it is and makes it harder to see where the real problems are. A strong reported ROAS and a stagnant business are not contradictions. In DTC, they are surprisingly common partners.
There is a second problem that sits beneath the attribution issue: ROAS treats all revenue as equal.
A returning customer who was going to buy anyway, converted through a retargeting ad, counts the same as a genuinely new customer acquired through cold paid traffic.
One of those represents growth. The other is expensive email marketing. ROAS cannot tell you which is which.
Three Scenarios Where ROAS as a Vanity Metric Actively Misleads You
Let me make this concrete. Here are three scenarios where treating ROAS as the primary success metric leads brands to entirely the wrong conclusions.
Scenario 1: The Brand With a 4× ROAS That Is Losing Money
A DTC skincare brand is running Meta ads with a reported 4× ROAS.
The team is happy.
Budget is increasing.
Then the finance team runs the numbers: after product cost, fulfilment, returns, and agency fees, the brand needs at least a 5× ROAS just to break even on each acquisition.
The 4× ROAS that looked like success is actually a loss on every new customer acquired.
The problem: ROAS does not account for margin.
A 4× ROAS on a 20% margin product is not profitable.
ROAS tells you revenue returned per ad pound spent. It says nothing about what is left after you pay for the product you sold.
Scenario 2: The Brand With Rising ROAS and Falling New Customer Acquisition
A DTC brand's Meta ROAS increases from 2.8× to 3.6× over three months.
The paid media team considers this a win but total revenue is flat and new customer numbers are declining.
What has happened?
The algorithm has optimised toward the easiest conversions: retargeting existing customers who were going to buy anyway.
The ROAS went up because the brand stopped spending on cold acquisition and started spending on people already in its ecosystem.
The problem: ROAS does not distinguish between new and returning customers.
Rising ROAS from retargeting existing buyers is not growth.
It is expensive email marketing and without tracking NC-CPA — the cost to acquire a genuinely new customer — you cannot see this happening until it is months deep.
Scenario 3: The Brand That Scaled ROAS and Destroyed Its CAC
A DTC brand optimises hard for ROAS across all campaigns.
Creative testing is entirely ROAS-driven.
Budget flows to whatever produces the highest reported return.
Six months later, ROAS is up but the cost of acquiring new customers has increased by 40%.
The brand has been optimising for short-term reported revenue at the expense of efficient acquisition and has trained the algorithm to find expensive customers who convert quickly rather than cheaper customers with long-term value.
The problem: ROAS does not track acquisition cost trend.
You can have excellent ROAS and a business model that is becoming structurally more expensive to run every single month.
ROAS Still Has a Role — Just Not the One Most Brands Give It
Before I tell you what to track instead, I want to be precise about something:
ROAS is not a useless metric. It is a misused metric.
ROAS is a useful campaign-level, tactical metric.
It tells you, within a specific ad set, whether one creative is generating more revenue per pound spent than another.
For that narrow purpose, comparing creative performance within a controlled test, it is a reasonable signal.
What ROAS is not useful for:
- Deciding whether your overall marketing is working
- Understanding whether you are acquiring new customers efficiently
- Comparing performance across channels
- Assessing business health
- Making budget allocation decisions at the brand level
- Understanding whether your marketing is building compounding value
The tell: if the primary question in your weekly marketing review is "what is our ROAS this week?" — your measurement framework is one metric wide when it needs to be five metrics deep.
You are navigating a business with one instrument.
The Five Metrics That Replace ROAS as a DTC North Star
The PLANT™ Framework uses five metrics together to give a complete picture of DTC marketing health.
These are the numbers in the Track phase and they are the metrics that, together, tell you whether your marketing is building a compounding business or simply sustaining an expensive one.
1. NC-CPA — New Customer Cost Per Acquisition
NC-CPA = Paid Media Spend ÷ New Customers Acquired
This is the metric that ROAS most aggressively hides. NC-CPA strips out returning customers entirely and asks the only acquisition question that actually matters: how much does it cost to bring in someone who has never bought from you before?
Platform-reported ROAS blends new and returning customer conversions together.
A campaign converting 80% returning customers at a high AOV will report a strong ROAS and a low blended CPA but it is not growing your customer base. NC-CPA separates these two things cleanly.
NC-CPA also solves the attribution inflation problem partially.
When you are tracking new customer numbers from Shopify (not from Meta's dashboard), you are working with data the platform cannot manipulate.
A new customer acquired is a verifiable fact. A platform-reported conversion is a claim.
Track it: Monthly. Set a ceiling i.e a maximum NC-CPA that the business can afford given your margin profile and LTV.
If NC-CPA exceeds that ceiling, the campaign is not growing your business efficiently regardless of what the ROAS dashboard says.
The benchmark question: What is the maximum you can pay to acquire a new customer and still be profitable within 90 days? That number is your NC-CPA ceiling. Most brands have never calculated it.
DOWNLOAD The PLANT™ Revenue System Guide and see how revenue compounding actually works.
2. MER — Marketing Efficiency Ratio
MER = Total Revenue ÷ Total Marketing Spend
MER is your actual return, not the platform-reported fiction.
It takes every pound of marketing spend (Meta, Google, TikTok, email platform costs, agency fees, influencer spend) and divides it into total revenue.
No attribution games. No platform claiming the same sale three times. One honest number.
Where ROAS is measured per campaign and per channel, MER is measured across the entire business.
Where ROAS fluctuates with attribution changes and platform algorithm shifts, MER is stable.
Where ROAS can look strong while the business is stagnant, MER tells you the truth immediately.
Track it: Weekly. A falling MER while platform ROAS holds steady is the clearest early warning sign that over-attribution is inflating your reported performance. A rising MER while ROAS looks modest usually means your channels are working together better than any single platform can measure.
3. CAC Trend — Customer Acquisition Cost Over Time
Blended CAC = Total Marketing Spend ÷ Total New Customers Acquired
Not blended ROAS. Not campaign ROAS. Your actual cost to acquire one new customer across all channels, tracked over time as a trend, not a snapshot.
A single CAC number tells you relatively little. A CAC trend over six months tells you a lot.
Rising CAC while ROAS holds steady means you are finding the same customers at increasing cost, often because you are optimising toward existing buyers or warming audiences rather than true cold acquisition.
Falling CAC while revenue grows is the metric combination that every DTC brand should be working toward.
Track CAC monthly, and break it out by channel where possible. Just be careful not to interpret all channel CACs the same way: Meta cold acquisition, Google branded search, and retargeting play very different roles.
4. Email Revenue as a Percentage of Total Revenue
Email Revenue % = Email Revenue ÷ Total Revenue × 100
This metric does not appear on any paid media dashboard. It does not show up in Meta Ads Manager or Google Ads. And it is one of the most important indicators of DTC marketing health available.
A DTC brand generating 30–40% of total revenue from email has built something that ROAS cannot measure and paid media cannot replicate: a compounding retention engine that generates revenue without requiring additional ad spend.
According to Klaviyo and DTC industry benchmarks, a well-functioning retention infrastructure typically sees email (especially flows) contributing toward the 30–40% range of total revenue — the hallmark of reduced paid media dependence.
Most DTC brands at the one to three million revenue level are generating under 15% from email.
That gap — between 15% and 30% — is not a creative problem or a list size problem. It is an infrastructure problem.
And it is direct, recoverable revenue that paid media spend can never compensate for.
Track it: Monthly. If this number is rising, your Nurture phase is working and your dependence on paid acquisition is decreasing. If it is static or falling while list size grows, your lifecycle flows need attention.
5. LTV by Acquisition Channel
LTV = Average Order Value × Purchase Frequency × Customer Lifespan
ROAS measures revenue from the first transaction. LTV measures revenue across the entire customer relationship.
A customer who buys once and disappears has a very different LTV from a customer who buys six times over two years and if your paid media is optimising for ROAS, it is optimising for the first transaction without any visibility into which channels produce the second, third, and fourth ones.
LTV by acquisition channel is one of the most strategically powerful metrics available to a DTC brand.
If your Meta cold traffic produces customers with a 90-day LTV of £45, and your Google branded search produces customers with a 90-day LTV of £120, the channel with the better ROAS is not necessarily the channel worth scaling.
Track it: At 30, 60, and 90 days post-acquisition, by channel. This data feeds directly back into the Plan phase of the PLANT™ Framework — informing where next quarter's budget should go based on which channels produce the most valuable customers, not just the cheapest first transactions.
What Your DTC Marketing Dashboard Should Actually Look Like
If you are currently running a weekly marketing review with ROAS as the primary metric, here is what a framework-level dashboard looks like instead, built around the five metrics above, with ROAS correctly positioned as a supporting tactical signal rather than a strategic north star.
THE PLANT™ TRACK DASHBOARD
| NC-CPA (New Customer CPA) | Paid Spend ÷ New Customers | Weekly |
| MER (Marketing Efficiency Ratio) | Total Rev ÷ Total Mktg Spend | Weekly |
| CAC Trend (blended + by channel) | Total Spend ÷ New Customers | Monthly trend |
| Email Revenue % | Email Rev ÷ Total Rev × 100 | Monthly |
| LTV by Channel (30 / 60 / 90 day) | AOV × Frequency × Lifespan | Quarterly |
| ROAS (by campaign / creative test) | Campaign Rev ÷ Campaign Spend | Weekly — tactical only |
Notice where ROAS sits in this dashboard: at the bottom, labelled as tactical only, reviewed weekly at the campaign level.
It has not been removed. It has been correctly positioned.
Also notice what sits at the top: NC-CPA — because knowing the true cost to bring in a new customer is the foundation of every other growth decision.
Why NC-CPA Is the Most Important Metric Most DTC Brands Do Not Track
Of the five metrics above, NC-CPA is the one with the most immediate, practical impact on how you manage paid media week to week and the one most consistently absent from most agency reporting.
Here is why it matters so much: blended ROAS and blended CAC both include returning customers.
In a brand with strong retention (which is good), returning customers can make blended numbers look excellent while new customer acquisition quietly becomes more expensive. You are growing a loyal customer base but not expanding it.
Eventually that base saturates, retention softens, and the business finds it cannot acquire new customers at the cost it assumed it could.
NC-CPA forces the question early.
It separates the acquisition signal from the retention signal and makes it impossible for a high-ROAS retargeting campaign to mask an inefficient cold acquisition strategy.
When your NC-CPA rises month on month while blended ROAS holds steady, that is the early warning sign but it's one that most DTC brands only see six months after the problem started.
Set a maximum NC-CPA ceiling before the month begins. If spend exceeds that ceiling before you hit your new customer target, the constraint is structural, not creative. No amount of testing will fix it without addressing the underlying unit economics.
How to Start Transitioning Away From ROAS as Your North Star
You do not need to rebuild your entire measurement infrastructure overnight. Here are four practical steps you can take this week:
- Calculate your NC-CPA for last month. Pull your total paid media spend from Meta and Google and any other channel. Pull your new customer count from Shopify (customers whose first order date falls within the same period). Divide spend by new customers. That is your NC-CPA. Now compare it against what your LTV data says you can afford to pay. If NC-CPA exceeds your breakeven threshold, your acquisition campaigns are not profitable on new customers regardless of what the ROAS dashboard shows.
- Calculate your MER for the last 30 days. Take your total revenue for the period and divide it by your total marketing spend. Include Meta, Google, TikTok, email platform fees, creative costs, agency fees, and any other marketing spend. Some brands calculate MER using ad spend only and exclude agency or tool costs. Either approach can work, but the important thing is to stay consistent with the definition you choose. The result is your MER (Marketing Efficiency Ratio). Write it down. Calculate it again in 30 days. You now have the start of a trend.
- Pull your email revenue percentage. In Klaviyo (or whatever email platform you use), find the revenue attributed to email flows and campaigns for the last 90 days. Divide it by your total revenue for the same period. If the number is under 20%, you have identified your biggest single lever for improving DTC marketing efficiency without spending an additional pound on ads.
- Start tracking CAC as a trend, not a snapshot. Calculate your blended CAC for this month: total marketing spend divided by new customers acquired. Calendar a reminder to calculate the same number next month. Two data points is a trend. Six data points is insight.
These four steps require no new tools, no new technology, and no new budget. They require only a change in which numbers you look at first when you open your marketing dashboard on Monday morning.
Which of these five metrics is your biggest gap right now?
The free PLANT™ Revenue Audit is a 5-question diagnostic that identifies exactly where your DTC marketing system is leaking revenue — whether that is in your NC-CPA, your tracking, your retention infrastructure, your acquisition cost trend, or your email revenue percentage. It's five questions that gives you Instant clarity and done within eight minutes..
Frequently Asked Questions: ROAS as a Vanity Metric and What to Track Instead
These are the questions founders and marketing leads most often ask when ROAS stops telling the full story.
Why is ROAS considered a vanity metric in DTC marketing?
ROAS is considered a vanity metric in DTC marketing because it measures revenue returned per advertising pound spent on a specific campaign without accounting for margin, customer lifetime value, whether the conversions are new or returning customers, or the total cost of acquiring a new customer.
A brand can report strong ROAS while losing money on every new customer acquired (if margin is thin), while new customer acquisition slows unnoticed (if returning customers are driving the ROAS number), or while blended acquisition costs trend upward month on month.
ROAS looks compelling in platform dashboards precisely because every platform uses last-click attribution by default, meaning the same sale is often claimed by multiple platforms simultaneously — inflating reported ROAS across the board.
What is NC-CPA and why does it matter more than ROAS?
NC-CPA stands for New Customer Cost Per Acquisition. It is calculated by dividing paid media spend by the number of genuinely new customers acquired — customers who have never previously purchased from the brand. NC-CPA matters more than ROAS for DTC growth decisions because it separates acquisition performance from retention performance.
A campaign with strong ROAS may be converting primarily returning customers, producing a misleadingly low blended CPA and a healthy reported return while the actual cost of growing the customer base is rising unnoticed.
NC-CPA forces this issue into view. It is the only acquisition metric that cannot be inflated by retargeting existing buyers.
What should DTC brands track instead of ROAS?
DTC brands should track five metrics alongside ROAS to get a complete picture of marketing health: NC-CPA (new customer cost per acquisition — the true cost of growing your customer base), MER (Marketing Efficiency Ratio — total revenue divided by total marketing spend across all channels), CAC trend over time (customer acquisition cost tracked monthly, not as a snapshot), email revenue as a percentage of total revenue (the clearest indicator of retention infrastructure health), and LTV by acquisition channel (which channels produce the most valuable customers over 30, 60, and 90 days). ROAS still has a role as a tactical, campaign-level signal for creative testing — but it should not be the primary metric for business-level marketing decisions.
What is MER and how is it different from ROAS?
MER (Marketing Efficiency Ratio) is calculated by dividing total revenue by total marketing spend — including all channels, platform fees, and agency costs, not just the spend on a specific paid campaign.
Where ROAS measures the performance of a specific ad campaign in isolation and is subject to platform attribution games, MER gives a whole-business view of how efficiently all marketing spend is converting into revenue.
MER cannot be inflated by last-click attribution because it does not come from any platform — it comes from your own revenue data and your own total spend figures. A falling MER while platform ROAS stays steady is one of the clearest early warning signs that platform attribution is overstating performance.
How do you calculate NC-CPA for a DTC brand?
NC-CPA is calculated by dividing total paid media spend for a given period by the number of new customers acquired in that same period — where "new customer" is defined as a customer whose first-ever order falls within the measurement window.
New customer counts should be pulled from Shopify (or your e-commerce platform), not from Meta or Google's dashboards, because platforms attribute conversions to themselves regardless of whether the customer was genuinely new.
Pulling new customer data from your Shopify admin under Customers, filtered by first order date, gives you a platform-independent figure that cannot be inflated by attribution overlap.
Why does ROAS look strong when revenue is flat?
ROAS can look strong while revenue is flat for two common reasons. First, the paid media algorithm may have optimised toward retargeting existing customers — people who were likely to buy anyway — rather than acquiring new customers.
Retargeting campaigns typically report high ROAS because they convert warm audiences efficiently, but they are not driving net new growth. Second, cross-channel attribution inflation means multiple platforms are each claiming credit for the same conversions, making total reported ROAS look stronger than actual business performance.
When ROAS is strong but revenue is flat, calculating MER and NC-CPA for the same period will usually reveal the discrepancy immediately.
Is ROAS still a useful metric in 2025?
Yes — at the campaign level, for creative testing and tactical optimisation, ROAS remains a useful signal. Within a structured A/B test where the only variable is the creative, ROAS can indicate which version is generating more revenue per pound spent.
The problem is not ROAS itself — it is the promotion of ROAS from a campaign-level tactical metric to a business-level strategic metric.
The brands that compound in this environment do so by measuring what ROAS cannot: NC-CPA, MER, CAC trend, email revenue percentage, and LTV by channel.
The Metric You Optimise For Shapes the Business You Build
When ROAS is your north star, you build a marketing operation optimised to produce strong ROAS numbers. You test creatives against ROAS. You allocate budget against ROAS.
You reward your team against ROAS. And over time, you build a very efficient machine for generating impressive ROAS reports and a business that is simultaneously more expensive to acquire new customers from, more dependent on retargeting existing ones, and less able to explain why growth is stalling.
When NC-CPA, MER, CAC trend, email revenue percentage, and LTV by channel are your north stars, you build something different. You build a marketing operation that knows exactly what it costs to grow.
One where paid media feeds email capture, email capture feeds retention, retention reduces acquisition pressure over time, and every cycle produces better data for the next one.
That is what the PLANT™ Framework is designed to produce — and it starts with measuring the right things.
If you want to find out which of the five metrics above represents the biggest gap in your brand right now, the free PLANT™ Revenue Audit is the next right step for you. Most people complete it in eight minutes.
If you have questions about any of the metrics in this article, or want a second opinion on your numbers, drop a comment below. I read every one and try to respond where I can.
0 Comments