Most CPG founders think about Instacart wrong.
They see it as a logistics channel. A way to get product into grocery orders. Something that “just runs in the background.” And because of that, they set it up once, let it run on auto, and wonder why results are mediocre.
Here’s what the brands actually winning on Instacart understand: it’s not a delivery app. It’s a customer acquisition platform. And once you see it that way, everything changes.
Mateusz Drela has spent 17 years in performance marketing and now runs RMIQ, an AI-powered retail media platform that helps growing CPG brands generate real sales across retail media networks. Their clients include Smash Kitchen, Smearcase, Heirloom Coffee, and Kooshy Croutons.
Here's how he thinks about Instacart.
00:07 - The moment Amazon became an ad network
02:39 - A platform where size doesn’t matter
04:38 - Your first 30 days on Instacart
08:00 - More keywords, more problems
10:02 - 3 brands, 3 different ROAS
Why Instacart is the great equalizer.
Instacart doesn’t own a single store. But it sells across 1,800 national, regional, and local retail banners, covering nearly 100,000 locations in all 50 states. For a startup with distribution at three regional chains, that’s the same playing field as a Fortune 500 conglomerate.
Because Instacart has no physical shelves, it has no physical shelf politics. No slotting fees for eye-level placement. No penalties for being a small brand. The algorithm doesn’t know your revenue. It knows your bids and your relevance. That’s it.
And crucially: when a shopper types “tortilla chips” into Instacart, they’re not thinking about you. They’re browsing. That’s the moment you win them.
Up to 90% of sales generated through Instacart ads are from customers who have never bought your brand before on the platform. Of those, 30–50% become repeat buyers. This is new-to-brand acquisition at a scale that’s genuinely hard to replicate anywhere else.
The one format that actually moves product.
Instacart offers a suite of ad products: Reach, Engage, Acquire. Ignore most of them.
The format that matters is Sponsored Products (also called Sponsored Search). It’s the only format directly tied to sales velocity. Ads appear where decisions get made: search results, homepage, “frequently bought with” sections, and checkout. It’s performance-based. It’s measurable. And it’s where 70–80% of total Instacart ad spend should go.
Everything else is brand awareness. That matters eventually. But if you’re a founder trying to move units, Sponsored Products is where the work happens.
The 4 moves that separate average from excellent.
Here’s where most brands leave money on the table. They set up Instacart’s automated campaign tools, watch ROAS hover around 2.0–2.5, and call it a day. But automated campaigns are designed to be good enough for everyone. Which means they’re optimized for no one.
Moving from automated to advanced management isn’t complicated. It’s just specific.
Move 1: Go manual. Automated campaigns work by spreading budget across categories and testing keywords at scale. Instacart’s algorithm can’t favor one brand over another, so it defaults to average.
Switching to manual campaign setup puts you in control of where dollars go and why. Brands that make this switch consistently see ROAS jump significantly. One specialty coffee brand went from 2.4 to 4.9. A tortilla chip brand went from 2.2 to 4.3.
Move 2: One ad group per product. Grouping multiple SKUs into a single ad group feels efficient. It isn’t. Different products have wildly different performance profiles. One SKU might return 3x ROAS. Another returns 6x.
When they’re bundled together, you can’t bid specifically for the winner. You end up underinvesting in your best performer and subsidizing your worst one. Build one ad group per product. It takes longer to set up, but it pays off immediately.
Move 3: Bid at the keyword level, based on your data. Instacart will offer you “suggested bids.” Ignore them. Suggested bids are calibrated across all advertisers in a category, not for your margins, your ROAS targets, or your specific product.
Instead, build a product-keyword matrix. For each product, map out which keywords actually convert and bid according to your own performance data. A tortilla chip brand should bid aggressively on “tortilla chips” and much lower on “crackers” (even if crackers are technically relevant).
Your bid should reflect what you can afford to pay per click while still hitting your target ROAS. Start with a target (say, 4.0), calculate your maximum allowable cost-per-click, and build your bids around that number.
Move 4: Maintain weekly, not monthly. This isn’t a set-and-forget channel. Extract performance data weekly. Kill underperforming keywords. There’s a catch: Instacart’s UI doesn’t let you pause keywords.
The workaround is to either drop the bid to the $0.30 minimum or rebuild the ad group without the keyword. It’s annoying. It’s also the difference between wasting budget on irrelevant terms and running tight, efficient campaigns.
A note for early-stage brands.
If you’re spending under $150/day on Instacart, A/B testing is probably a waste of time. You won’t reach statistical significance fast enough to draw conclusions. Instead, run auto campaigns first. Not as your long-term strategy, but as a keyword harvesting tool.
Let the algorithm identify what’s converting. Then take those insights, build a manual campaign structure, and go from there. Scale, then optimize. In that order.
The operational details most founders miss.
Retailer-specific ads. Instacart generally runs ads across all retailers where your product is listed. But if you want to advertise specifically on Sprouts’ or Thrive Market’s website, you need a separate account — unless you’re running through a unified third-party platform.
Data feed timing. If your product isn’t showing up on Instacart, the issue is almost always the retailer’s data feed. National chains update weekly. Regional retailers often update monthly. If you just got a new account, you might be waiting. Chase your retailer to make sure they’ve included you in the specific feed Instacart pulls from.
SKU-level targeting. If you have an exclusive product at one retailer, you can effectively target only that retailer by running ads only for that SKU. It’s a useful workaround for retailer exclusivity situations.
Instacart is the virtual number three grocery retailer in the United States. It’s also the most underutilized acquisition channel in CPG for brands that are still treating it like a logistics tool.
The brands winning have made a simple mental shift: stop thinking about how your product gets delivered. Start thinking about how your brand gets discovered. That shift changes how you build campaigns, how you bid, how you measure, and ultimately, how you grow.
