If you're running Meta ads and you check your account today, there's a good chance your Dynamic Product Ads (DPAs) show the best ROAS of any campaign type you run. That number is probably lying to you, and trusting it is one of the most common ways brands accidentally destroy their own growth.
This is the DPA death spiral, and it happens quietly enough that most founders don't notice until new customer acquisition has already collapsed.
What a DPA actually is
A Dynamic Product Ad is a carousel-style ad that automatically pulls in products a user has already viewed on your website, using your pixel data. Someone looks at a pair of shoes, leaves without buying, and later sees a carousel ad showing that exact pair, sometimes with related products.
DPAs are a retargeting tool. They are built for people who are already deep in your funnel, product-aware and close to a purchase decision. That is what they are good at. The problem starts when brands treat them as a scalable growth channel instead of a narrow, bottom-funnel tool.
Why the ROAS number is misleading
Here's the mechanism. DPA ROAS looks excellent because of how attribution works, not necessarily because the ad is doing the heavy lifting.
If someone visited your site organically, saw three of your top-funnel ads over the following week, became genuinely interested, and then finally converted after seeing a DPA retargeting ad, the DPA gets full attribution credit for that sale under standard last-click reporting. The top-funnel ads that actually built the awareness, interest, and purchase intent get little to none. This is the same distortion covered in more depth around blended versus platform ROAS, just showing up specifically inside the retargeting layer of the account rather than across the account as a whole.
The fix is to look at incremental attribution, not standard attribution. Incremental attribution holds out a control group who never see the ad, then compares their conversion rate to the group who did see it. The difference between those two numbers is the ad's true incremental impact.
Run this comparison on your DPA campaigns and the "5x ROAS" you thought you had often drops to a fraction of that once you strip out people who would have converted anyway. DPAs also frequently include existing customers in their reporting unless specifically excluded, which inflates the number further, since repeat buyers convert at a much higher baseline rate regardless of the ad.
How the death spiral actually happens
Once a founder or media buyer sees DPA showing the best ROAS in the account, the natural instinct is to shift budget toward it. Here's the sequence that follows:
- Budget migrates from top-funnel creative production and spend toward DPA, since it looks like the more "efficient" channel
- Top-funnel campaigns, which are responsible for bringing new, cold audiences into the funnel, get starved of budget
- Fewer new people enter the top of the funnel each month
- With fewer people entering, there are fewer people available further down the funnel to eventually retarget
- DPA volume and spend capacity shrinks, because there's a smaller pool of warm, product-aware visitors to serve it to
- Overall account performance and new customer revenue decline, even though the DPA campaign itself may still show a strong dashboard ROAS the entire time
This is why it's called a spiral. Each stage makes the next stage worse, and the dashboard metric that triggered the original decision keeps looking healthy right up until total revenue has already started declining. This mirrors a broader pattern worth checking for across the whole account, not just DPA, where a campaign's strong dashboard number is really just capturing demand another part of the funnel already created.
Free Profitability Review
Not sure if your ROAS is actually profitable?
We review Meta, Google, GA4, and Shopify data together and tell you where the numbers are misleading you, before you spend another rupee scaling on a false signal.
Request a Paid Media Profitability ReviewThe budget rule that prevents this
DPAs should represent 10-20 percent of total ad spend, not more. The overwhelming majority of budget, 70 to 90 percent, needs to stay allocated to top-funnel creative that brings in net-new, cold audiences. The remaining share sits in mid-funnel content designed to build product awareness before the retargeting stage.
If your DPA spend is creeping past that 20 percent threshold, or if DPA has become your single largest campaign by spend, that is the earliest warning sign of the spiral starting, well before top-line revenue shows any sign of a problem.
How to check if this is already happening in your account
Run these checks in order:
Check your DPA share of total spend. If it's above 20 percent, that's your first flag.
Apply incremental attribution to your DPA campaign. In Ads Manager, use the columns and comparison settings to view incremental attribution alongside standard attribution. The gap between the two numbers tells you how inflated your dashboard ROAS actually is.
Track new customer acquisition volume over the last 3 to 6 months, independent of ROAS. If DPA ROAS has stayed flat or improved while new customer volume has declined, that's the spiral in progress. Efficiency looking fine while the customer base stops growing is the exact signature of this problem.
Audit your top-funnel creative volume. If cold, top-of-funnel ad volume has quietly dropped over the same period DPA spend increased, you've found the mechanism.
What to do about it
Cap DPA spend at 10-20 percent of total budget and hold that line even if the ROAS looks tempting to scale further. Redirect the freed-up budget back into top-funnel creative production and cold audience testing.
This will likely make your blended account ROAS look slightly worse in the short term, because top-funnel spend is inherently less efficient on a last-click basis than retargeting spend. That is expected and correct. The goal is not to maximize the dashboard number. The goal is to maximize new customer revenue and long-term account health, which requires a healthy, well-fed top of funnel.
The takeaway
DPAs are a useful, narrow tool for closing out warm traffic. They are not a growth engine, and treating them as one is one of the most common, hardest-to-notice mistakes in Meta ads accounts today. The dashboard will tell you everything is fine right up until it isn't. Check incremental attribution, hold your DPA spend cap, and keep the top of your funnel fed.
FAQ
Why does my DPA campaign show a great ROAS but my new customer growth isn't increasing? DPA ROAS is inflated by last-click attribution crediting the retargeting ad for sales that top-funnel campaigns actually built. As budget shifts toward DPA, top-funnel spend shrinks, fewer new people enter the funnel, and new customer growth stalls even while DPA's own dashboard number stays strong.
What percentage of ad spend should go to dynamic product ads? 10-20 percent of total budget. The majority, 70 to 90 percent, should stay in top-funnel creative bringing in new, cold audiences, with the remainder in mid-funnel content that builds product awareness before retargeting.
How do I check if my DPA ROAS is inflated? Apply incremental attribution instead of standard last-click attribution. Incremental attribution holds out a control group that never sees the ad and compares their conversion rate to the group that did, isolating the ad's true added impact rather than crediting it for sales that would have happened anyway.
What's the difference between standard and incremental attribution on Meta? Standard, last-click attribution credits whichever ad a customer saw right before converting, even if earlier ads did the actual work of building interest. Incremental attribution measures the ad's true added effect using a held-out control group, which is why DPA campaigns often look far less impressive once measured this way.
Is it a mistake to scale up dynamic product ads if they show the highest ROAS in my account? Yes, if that scaling pulls budget away from top-funnel creative. DPA's strong ROAS comes from serving an already-warm, already-interested audience, and that audience only exists because top-funnel campaigns keep feeding new people into the funnel. Starving that supply eventually shrinks DPA's own performance too, even as the dashboard number looks fine along the way.