Nationwide · DTC and Shopify brands
Ecommerce marketing that growscontribution, not just revenue
Revenue is easy to buy. Keeping any of it is the hard part. We run paid social, Shopping, email and SMS as one system, measured against the only number that decides whether growth is worth having: what is left after a customer is acquired.
★★★★★ 4.9 across 312 client reviews Map your contribution by channel first
What the market looks like in 2026, before anyone touches your account
- 22%
- Median DTC contribution margin
- 1.4%
- Average Shopify conversion rate
- $39
- Median Meta cost per purchase
- 20%
- Of online orders come back
Sources: Fairview via Daymark (contribution margin, 2025 vs 2021), Littledata Shopify benchmark of roughly 2,800 stores, Triple Whale Meta benchmark for August 2025 to July 2026, National Retail Federation 2025 Retail Returns Landscape. Figures are medians and industry ranges, not a promise about your account.
The number nobody reports on
Your ROAS is fine.You are still losing money.
Almost every ecommerce dashboard is built to celebrate revenue. Shopify shows sales. Meta shows purchases and return on ad spend. Google shows conversion value. Not one of those screens knows what your product costs, what shipping costs, what your payment processor takes, or how many of those orders will be back in a box next month.
So the number an operator quotes is almost always the one furthest from the truth. A 3x return on ad spend sounds healthy right up until you subtract everything under the ad spend line, at which point it can be a loss.
Contribution is what is left after the order actually happens. It is the only number that tells you whether to spend more.
Here is the arithmetic on a single order, built entirely from published 2026 benchmarks rather than any one account. Median DTC gross margin across eleven public filings is roughly 57 percent. Median Shopify average order value sits around $85 to $95. Free shipping thresholds typically absorb $7 to $10 per order. Payment and platform fees compound to somewhere between 3 and 5 percent of revenue. Roughly one in five online orders is returned, and processing each one costs real money before the product is back on a shelf.
One $95 order, all the way down
Illustration built from published benchmarks- Order value$95.00Median AOV for a DTC Shopify store, per Daymark's 2026 benchmark set.
- Cost of goods at 57% gross margin−$40.85Median gross margin across eleven public DTC filings, per Eightx's read of the 10-Ks.
- Shipping and fulfilment−$8.00What a free-shipping threshold actually absorbs on a typical order.
- Payment and platform fees at 4%−$3.80Processing at 2.9% plus $0.30 compounds to 3 to 5% of revenue once app fees are counted.
- Returns at 20%, $15 to process each−$13.07One in five orders comes back. You keep the outbound shipping cost and add reverse logistics.
- Contribution before a cent of ad spend$29.28Break-even return on ad spend is 3.24x. Anything below that is a paid loss.
Every input above is a published median or range, and every one of them moves for your business. That is the point of the tool below, not the specific numbers here.
Now put that next to what acquisition costs. The median cost per purchase on Meta across a year of tracked ecommerce spend was just under $39. On a $95 order at ordinary margins, that customer is already underwater, and the platform will report the campaign as profitable while it happens.
It gets worse than the headline suggests, for two reasons that almost never make it into a monthly report. Platform-reported return on ad spend runs well above true incremental impact, by a factor of two to five in one analysis of 253 media-mix models covering $383M of spend. And a platform-reported purchase is not a new customer. It counts repeat buyers you would have kept anyway, which is why published customer acquisition figures for the same channel land anywhere from $68 to well over $200 depending on whose denominator you use.
None of this means paid media does not work. It means the scoreboard is wrong, and a scoreboard that reads high is far more expensive than one that reads low. Why Shopify, Meta and GA4 never agree takes that apart properly, and the contribution margin page carries the full break-even tables by margin band.
What we do instead is simple to say and unusual to actually run. We rebuild the measurement first, then buy against contribution, then push the channels that carry margin rather than the channels that carry credit.
Free, nothing to install, no email required
Map your contributionchannel by channel
Put your own economics in the left column and your last full month of spend and platform-reported revenue in the right. It works out what each channel actually contributes after cost of goods, shipping, fees and returns, and shows you which ones are quietly funded by the others.
Your economics
Defaults are 2026 published medians. Overwrite every one of them.
Last full month, by channel
Use whatever revenue figure each platform reports. The overlap between them is exactly what this is designed to expose.
What you actually keep
Updates as you type. Nothing is sent anywhere.
Want this run against your real numbers instead of a single month? A 30-minute call and read-only access to the ad accounts gets you the same map with the double-counted revenue stripped out. Book one here.
How the engagement runs
Five moves,in this order
The order matters more than any single tactic. Buying media before the measurement is fixed just means spending faster in the dark, and scaling creative before the margin is known means scaling a loss.
-
01
Fix the scoreboard
Server-side events, a clean purchase signal, and a blended view that reconciles Shopify against the platforms. Then a post-purchase survey so there is one source of truth that no ad platform controls. Nothing else is trustworthy until this is done.
-
02
Set the ceiling
Your real contribution per order, by product and by category, becomes the break-even. Every campaign target is derived from that number rather than from a ROAS goal somebody picked because it sounded ambitious.
-
03
Rebuild acquisition
Paid social and Shopping restructured around the margin you actually have, with creative volume treated as the input it is. Most accounts are creative-starved, not budget-starved, and the fix is a production cadence rather than a new audience.
-
04
Own the second order
Flows before campaigns, on both email and SMS. Welcome, browse, cart, checkout, post-purchase, replenishment and win-back, built once and then improved, because this is the revenue that carries no acquisition cost at all.
-
05
Report on contribution
One dashboard, one number, reviewed weekly. What each channel contributed after everything, what changed, and what we are cutting or feeding as a result. No vanity charts, no screenshot of a platform dashboard presented as proof.
Channel mix
What each channelcosts and returns
Published 2026 medians, so you can see roughly where you sit before anyone sells you anything. Treat them as orientation. Category and margin move every one of these more than tactics do.
| Channel | Typical cost | Typical return | What it is actually for |
|---|---|---|---|
| Paid social | $15 CPM $39 per purchase |
1.9x reported ROAS 1.5% conversion rate |
New demand at volume. It is the most expensive customer you will buy and usually the only one who did not already know you existed. Roughly two thirds of tracked ecommerce budget sits here. |
| Shopping and PMax | $0.66 CPC $39 to $44 per purchase |
4x to 5x reported ROAS | Capturing demand that already exists. Cheapest qualified click in the stack because the shopper saw price and image before clicking. Feed quality, not bidding, is the lever. |
| Brand search | Low CPC | Very high reported ROAS | Mostly defence. It reads beautifully in a report and is largely revenue you would have received anyway, which is precisely why it flatters a blended number. |
| Near zero marginal cost | 4% to 5.3% conversion rate Roughly 27% of revenue |
Margin. Flows are about 5% of sends and roughly 41% of email revenue, at nearly 18 times the revenue per recipient of campaigns. Nearly half of flow revenue comes from first-time buyers. | |
| SMS | Real per-message cost | Flows are 7.6% of sends and 45% of SMS revenue |
Urgency and recovery. A separate discipline from email because of express written consent, carrier registration and a genuine send cost. Compliance is the part most agencies cannot explain. |
| Organic and direct | Slow, fixed cost | 2.7% to 3% conversion rate | Compounding, and under pressure. AI overviews now appear on a quarter of US desktop searches and zero-click behaviour keeps climbing, so treat traffic recovery as a project rather than an assumption. |
Sources: Triple Whale Meta benchmarks August 2025 to July 2026, Klaviyo 2026 ecommerce benchmark report covering more than 183,000 brands, published Google Shopping and Performance Max benchmark compilations for 2026, and cross-source conversion-by-traffic-source ranges. Reported return on ad spend is platform-attributed and consistently overstates incremental impact. Full working on what customer acquisition actually costs by category.
Where the work happens
Five channels,run as one system
Each of these can be bought on its own. They are worth considerably more when the same team owns the margin target across all of them, because that is the only way a budget can move to where it earns.
Shopify
The store is the conversion rate, and the conversion rate is the multiplier on every other line in this list. Speed, product page, cart and checkout, plus the app stack quietly taxing every order.
Platform average is 1.4%. Top decile clears 4.7%. Mobile carries most of the traffic and converts at roughly half of desktop.
Paid social
Meta first, TikTok where the category supports it. Creative volume, angle testing and a campaign structure built for the way the auction actually allocates budget now.
Median CPM around $15 and roughly two thirds of tracked ecommerce budget. The most competitive dollar you will spend.
Google Shopping & PMax
Feed work before bid work. Titles, attributes, exclusions and the structure that stops Performance Max from spending your prospecting budget on your own brand terms.
Shopping CPC around $0.66, far below search. PMax now takes about 62% of Shopping spend.
Flows built properly and then left to compound, with campaigns on top rather than instead. Segmentation, deliverability and revenue per recipient, not open rates.
Flows are about 5% of sends and 41% of email revenue. Mature programmes run 30 to 40% of total revenue on email and SMS combined.
SMS
Consent capture, carrier registration and a send calendar that does not burn the list. Treated as its own channel with its own economics, because it is one.
SMS flows are 7.6% of sends and 45% of SMS revenue. Express written consent is the part that carries legal risk.
Diamond Shell AI CRM
Included with the agency engagement. AI email and SMS working every abandoned cart, quote and dormant customer, with the record of what was sent and what it produced sitting in one place.
Attribution on orders, pipelines by stage, and two-way messaging from a single inbox.
Read before you hire anyone
The argumentsbehind the work
All of it free, none of it gated. If you take these five pages and run the playbook yourself, that is a legitimate outcome and it happens regularly.
- Your ROAS is fine and you are still losing moneyBreak-even return on ad spend by margin band, and the costs that never make it into the ratio.
- Why Shopify, Meta and GA4 never agreeThree attribution models, three different numbers, all of them internally correct. What to actually trust.
- What customer acquisition costs by categoryCAC and AOV ranges by vertical, and why the published figures differ by a factor of three.
- Scaling Meta ads without killing ROASBudget steps, creative volume maths, and how to tell a learning dip from a real ceiling.
- The email and SMS flows that carry revenueFlow by flow, what each one contributes and the order to build them in.
- Client resultsWhat changed, at 30, 60 and 90 days, including what we would do differently.
You've got questions
We've got
straight answers
What does an ecommerce marketing agency actually do?
In our case, four things in a fixed order. Rebuild measurement so the numbers can be trusted. Work out your true contribution per order so there is a real ceiling on acquisition cost. Run paid social and Shopping against that ceiling with enough creative to feed it. Then build the email and SMS flows that turn a first order into a second one at no acquisition cost.
What we do not do is optimise toward a return on ad spend target somebody invented, or report platform-attributed revenue as though it were profit.
What size brand do you work with?
Typically $1M to $30M in annual revenue, with margin structure that can support paid acquisition and inventory you would like to keep moving. Shopify and Shopify Plus mostly, though the work is not platform-dependent.
If you are pre-revenue, or running gross margin below about 40 percent with no repeat purchase, paid acquisition is unlikely to close and we will say so on the call rather than after the invoice.
Why do you keep talking about contribution instead of ROAS?
Because return on ad spend ignores every cost below the ad spend line. Cost of goods, shipping, payment fees and returns all come out after that ratio is calculated, and together they are usually the majority of the order.
Median DTC contribution margin has fallen from roughly 35 percent in 2021 to about 22 percent in 2025, which means break-even targets have moved for nearly everyone while the ROAS goals in most accounts have not. The full working is here, and the calculator on this page will do it against your own numbers in about a minute.
My platforms report different revenue. Which one is right?
All of them, under their own model, which is why they will never reconcile. Meta counts a view-through and a click within its attribution window. GA4 uses a last-click data-driven model. Shopify counts orders and does not care what happened before the session.
The practical answer is to stop trying to make them agree and instead run one blended number you control, sanity-checked against a post-purchase survey and against what happens to revenue when you turn a channel off. The attribution page covers how to set that up.
Do you serve one region, or nationwide?
Nationwide. Ecommerce has no map pack, no service area and no local ranking to defend, so geography does not shape the work the way it does for the trades. We are based in Southern California and run ecommerce accounts anywhere in the United States.
How quickly does anything move?
Measurement and flows are usually live inside two to three weeks, and flows alone tend to move revenue because they capture demand you have already paid for. Paid restructures need roughly 30 days to gather enough conversion data to bid on properly.
Ninety days is when a number is real. A good month can be seasonality, a promotion, or luck. Three months of contribution data is not.
Does the 90-day guarantee apply to ecommerce?
Yes, for brands already spending on Google or Meta, with the same conditions as everywhere else. We rebuild the funnel, rewrite the ads and manage the accounts until we beat your current cost per acquired customer. If we have not inside 90 days, we keep working at no fee until we do.
You cover media spend, we control the creative, and the comparison has to be like for like on offer mix and season. The full terms are linked in the footer.
Who owns the accounts and the data?
You do. Ad accounts, creative, landing pages, your email and SMS lists and your customer database. If we part ways it all walks with you.
The single exception is the Diamond Shell AI CRM, which is our own software. You can keep it on a separate subscription or take a full export of your data and go elsewhere with it.
Next step
Thirty minutes.
No pitch deck.
Bring your last full month of spend and your actual gross margin. We will map contribution by channel on the call and tell you which line to cut first, whether or not you ever hire us.
★★★★★ 4.9 across 312 reviews No obligation · Usually booked 1-2 days out