Most DTC plateaus start months before revenue flattens. The usual causes: new-customer acquisition weakening while returning customers hide it, a best seller that wins first orders but doesn't create repeat buyers, discounts pulling demand forward, ad platforms and email all taking credit for the same sales, Amazon and your site trading the same customers, or margin leaking after the sale. A useful growth roadmap starts by finding which of these applies to you.
Six reasons DTC growth stalls
1. Acquisition is weakening behind a steady MER
Total revenue over total marketing spend can hold flat for a year while revenue from new customers falls. Returning customers cover the gap until they don't.
Read: MER vs ROAS vs aMER →2. Your best seller is a dead-end entry product
The product that wins the most first orders is often not the one that creates repeat customers. Budget follows the first order and the customer base gets weaker.
Read: How to find your gateway product →3. Discounts are doing the selling
If most repeat orders only happen during promotions, retention is being paid for with margin and full-price demand is shrinking.
Read: Measuring discount dependency →4. Every platform claims the same order
Meta, Google and Klaviyo can each take credit for one sale. Added up, attributed revenue can exceed what Shopify actually took, so the plan scales the wrong things.
Read: Attributed vs incremental revenue →5. Amazon and your site are trading customers
Amazon growth can be new demand or your own customers switching channels. The two need very different decisions.
Read: Amazon or DTC: where the next dollar goes →6. Margin leaks after the sale
Returns, shipping subsidies and landed cost changes can turn a product that looks profitable at checkout into one that loses money.
Read: True margin after returns by SKU →How we turn the cause into a roadmap
- ConnectAds, Shopify, Amazon, email and product data in one place, reconciled so the store, not the ad platform, is the source of truth.
- Find the causeCheck each of the six patterns against your numbers. Usually one or two explain most of the slowdown.
- DecideEach finding becomes a decision: scale, fix, test or stop, with the number that supports it.
- PlanRank by impact and effort. A 30-day plan in the Snapshot, 90 days in the Growth Decision Analysis, each step with an owner.
- ReviewWith Growth Copilot, check results against the plan every month and adjust.
What a 30-day plan can look like
- Scale: move a share of prospecting spend to the product whose buyers reorder most, and judge it on new customer ROAS.
- Fix: rebuild the post-purchase flow for best-seller buyers, who mostly never order again.
- Test: swap the 20% welcome code for a gift with purchase on half of traffic and watch 60-day repeat.
- Stop: retargeting existing subscribers with acquisition ads.
An illustrative example for a brand doing about $20k a month. Your plan depends on what your data shows.
Questions founders ask
Why has my Shopify store stopped growing?
The most common reasons are weakening new-customer acquisition hidden by returning customers, budget concentrated on a product that doesn't create repeat buyers, discount dependency, double-counted attribution, Amazon and DTC trading customers, and margin lost to returns or shipping. Splitting new and returning revenue is the fastest first check.
Should I change my agency?
Not necessarily. A plateau is often about which products, customers or channels the budget points at, which sits above an agency's brief. Find the cause first, then judge the agency on numbers like new customer ROAS and cohort payback.
How long does it take to get out of a plateau?
It depends on the cause. Some fixes, like moving spend to a better entry product, show up in weeks. Others, like reducing discount dependency, take a few months. We won't promise a timeline before seeing the data.
Can I build a growth roadmap myself?
Yes. Start by splitting revenue into new and returning customers by month, then look at repeat rate by first product. Our blog walks through each check. The Snapshot is there if you'd rather have an operator do it.