The ROAS Problem No One in Your Retailer Meeting Will Mention | Left Hand Agency CPG High Five

If you've ever looked at a ROAS report from a retail media network and felt really good about it, I want to gently complicate that feeling.

ROAS is a useful metric. I'm not here to tell you to ignore it. But after years of running media for CPG brands at Left Hand Agency, and having sat on the brand side myself, I've come to think of ROAS as a directional metric, not a source of truth. It can tell you something is working. It rarely tells you the whole story of why.

For brands selling in physical retail, where the path to purchase runs across multiple touchpoints, platforms, and media types, ROAS gets misread constantly. Not because marketers aren't smart, but because the metric itself has real structural limitations that don't show up in the dashboard.

Here are five of the most common ways ROAS misleads CPG brands, and what to watch for instead.

1. ROAS Only Sees Its Own Platform

When you're running media across multiple channels simultaneously, every platform is going to take credit for the sale. That's just how attribution works. But it creates a real problem for CPG brands because nothing happens in a bubble. 

Let’s say a brand is running a campaign on a retail media network (RMN) while also investing in out-of-home (OOH) in the same markets. The RMN reports back strong ROAS and the temptation is to call that a win for retail media. But what if the OOH was doing the heavy lifting, driving awareness and purchase intent that showed up as a conversion in the RMN platform? You'd never know from the report.

This is why I always frame retail media ROAS as directional, not definitive. It can tell you the investment is moving in the right direction. It cannot tell you what else contributed to that result.

The fix isn't to stop measuring. It's to resist the urge to optimize to a single platform's numbers when other media dollars are in market at the same time. The strongest results I've seen come from brands that think holistically about their media mix and evaluate performance across the full investment, not just what one retailer's dashboard is telling them.

2. ROAS Takes Credit for People Who Were Already Going to Buy

This one comes up A LOT with sponsored search, whether that's Google PPC or a retail platform like Instacart. The ROAS on these placements often looks fantastic, and there's a reason for that. You're largely capturing shoppers who were already planning to buy your product. They searched, you showed up, they purchased. The platform counts that as a win.

And here's the thing, it is a win. Just not in the way the ROAS number implies.

Sponsored search drafts off every other media investment you've made—the OOH that built awareness, the CTV that drove consideration, the retail media that kept you visible in store. It all works together to get a shopper to the moment of decision. Sponsored search is there to make sure you don't fumble it.

I always tell brands that running media without sponsored search is like handing a portion of your ad dollars to your competitor. Because if you don't show up at that critical moment, they will.

So yes, you need to spend those dollars. But don't let the ROAS on sponsored search convince you it's doing more strategic work than it is. It's protecting your investment, not replacing the need for it.

3. Incrementality is Murkier Than Retailers Want You to Believe

If you've ever been to an industry event in the last few years, you've heard some version of this story. The retailer or agency report showed strong performance. The balance sheet told a different story. That gap is often an incrementality problem.

Incrementality is supposed to answer the question: Did my media actually drive new sales, or did it just capture purchases that were already going to happen? 

Retail media networks have gotten better at modeling this, but there's a fundamental limitation. They're walled gardens. They can only see what happens inside their own ecosystem.

That means a retailer can identify a "new customer" in their platform without knowing that shopper simply switched from a different store. They didn't grow your brand's household penetration. They just moved from one shelf to another. The RMN counts it as incremental. Your overall sales don't reflect it.

True growth measurement looks at overall sales lift, long term household penetration, and methodologies that live outside any single platform's reporting. It's simply wrong to look at ROAS as a growth metric. It's channel-level efficiency. And there's an important difference between the two.

4. Upper-Funnel Channels Get Shortchanged Before They Have a Chance to Work

This is honestly the hardest conversation we have with clients. And I'll be straight with you: we don't always win it.

Here's what typically happens. A CPG brand invests in CTV or streaming audio to drive broader awareness and reach new households. The ROAS doesn't materialize quickly because these channels live in the upper and mid funnel. They're building the conditions for a future sale, not triggering an immediate one. So the budget gets cut, the channel gets labeled as underperforming, and the brand wonders why velocity isn't growing.

The real issue is that it's often not the CMO making that call. It's the CFO who needs short-term numbers to report against. And the person in the room saying "you're measuring this wrong" isn't always the most popular person at the table.

The best case for brand investment has already been made, and made rigorously, by researchers like Byron Sharp, Les Binet, Peter Field, and Mark Ritson. The data is there. Sharp's How Brands Grow was practically written with the CFO audience in mind, which makes it a bit of a dense read, but the numbers are ironclad. We've been known to send clients a copy.

The takeaway for brand-side marketers is this: if your attribution window can't see past 30 days, it will never give you an accurate read on the channels doing the most important long-term work. ROAS will always make performance media look like the hero. That doesn't mean it is.

5. ROAS Swings Are Almost Never About ROAS

Here's something I've seen play out more times than I can count. 

A brand's ROAS on their retail media or paid search suddenly gets really efficient. Everyone's excited. The temptation is to credit the media plan. But dig a little deeper and you find out a brand awareness campaign just launched, or a TikTok video went viral, or a competitor went on backorder and cleared off the shelf for two weeks.

The opposite happens too. A competitor runs a BOGO and your ROAS tanks. Trade spend shifts. A rival brand has their own viral moment. Suddenly your numbers look soft and the knee-jerk reaction is to question the media investment.

ROAS rarely swings without a secondary reason. The metric itself doesn't tell you what caused the movement. It just reflects the outcome.

This is why pulling up to the 30,000 foot view matters so much. What else was in-market when your numbers shifted? What was your competitor doing? Was there any earned media, trade activity, or organic momentum at play? Performance media is porous. It absorbs the impact of everything happening around it and reports it back as its own.

The brands that get this right treat ROAS as one signal inside a broader picture, not a standalone verdict on whether their media is working.

---

How To Deal With ROAS Problems?

ROAS isn't a bad metric. But ROAS problems are also real. Because it's a limited metric. And the retailers and platforms reporting it back to you have no financial incentive to tell you that.

The marketers who get the most out of their media budgets are the ones who use ROAS as a starting point for a conversation, not a final answer. They ask what else was in-market, pressure test incrementality claims, protect upper funnel investment even when the CFO is asking hard questions, and know that a number that looks great on a dashboard doesn't always show up on the balance sheet.

Next week we're going to go a layer deeper with five more ways ROAS creates blind spots for CPG brands, including why it fails you on customer lifetime value, why margin and price point make benchmarks nearly meaningless, and how seasonality gets mistaken for media efficiency.

---

We are Left Hand Agency, a boutique media buying agency built for CPG brands. We help brands make smarter, more holistic media investments across retail media networks, out of home, CTV, paid search, and streaming audio. If your ROAS reports are raising more questions than answers, we should talk.