Amazon revolutionized how brands viewed retail media advertising, but Amazon is not the only one that can leverage shopper data to build a media business anymore. In 2026, Walmart Connect, Target Roundel, and Instacart Ads have matured into separate retail media ecosystems based on various sources of competitive advantage.
Walmart Connect leverages e-commerce, offline retail, purchase history, offsite media, social, connected TV, and in-store advertising. Target Roundel connects Target’s shopper ecosystem with sponsored products, programmatic media, CTV, search, social, and influencer campaigns. Instacart Ads takes a different route, using grocery shopping intent and a network that extends across thousands of retail locations and hundreds of retail media environments.
That distinction matters because the biggest retail media network is not automatically the right network for every brand.
A beauty company expanding through Target may have a very different media problem from a packaged-food company trying to increase Walmart sales. A grocery brand may care more about basket behavior and purchase frequency than about broad display reach. A DTC brand trying to enter physical retail may need a network that can help prove demand before retailers commit more shelf space.
Hence, the appropriate question that should be asked in 2026 is not just “Is Walmart Connect better than Target Roundel better than Instacart Ads?” but rather:
Which retail media network provides the optimal balance between access to shoppers, data, media inventory, metrics, retail distribution, and commercial fit for your brand?
This guide breaks down that question in detail.
What Is a Retail Media Network?
It can be defined as a digital advertising environment created by a retailer or commerce platform that enables brands to advertise on the retailer’s digital and physical property using shopper audiences and first-party commerce data.
One of the simplest examples of RMN is the ad that appears when you search for a product on a retailer’s site.
However, modern retail media goes much further than sponsored product listings.
A mature RMN can include:
- Sponsored search
- Sponsored products
- Sponsored brands
- Display advertising
- Video
- Connected TV
- Programmatic advertising
- Social advertising
- Influencer marketing
- Search engine advertising
- In-store digital screens
- In-store audio
- Experiential advertising
- Sampling
- Off-site audience activation
- Shopper audience targeting
- Closed-loop sales measurement
However, the main difference between retail media and traditional digital advertising lies in the relationship between media consumption and commerce.
For instance, retailers can easily tell what products consumers are looking for, browsing, putting into shopping baskets, buying, rebuying, and even buying in-store. This results in the first-party data advantage that general advertising services cannot always possess.
According to EMARKETER, in 2026, U.S. advertisers will invest approximately $71.98 billion in retail media, which would indicate 18.7% year-over-year growth. Its research also points to increasing concentration among larger retail media networks rather than unlimited growth for every retailer that launches an ad business.
Why does the definition matter?
Not every RMN is really selling the same product.
One network may primarily sell high-intent sponsored search. Another may be strongest in loyalty-based audience targeting. Another may provide access to multiple grocery retailers through a single advertising platform.
That is why comparing RMNs only by audience size can produce misleading conclusions.
A brand should evaluate the entire commercial system:
Audience → Media → Availability through retail channels → Purchase → Measuring effectiveness → Optimize
If even one of those links is broken, it doesn’t matter how great the numbers for impressions may look.
Why Retail Media Is Moving Beyond Amazon?
Amazon remains an important retail media platform, but the industry’s growth is increasingly creating room for other retailers to monetize their own shopper relationships.
EMARKETER’s 2026 research describes a market in which Amazon and Walmart are capturing a large share of incremental retail media spending, while a smaller group of scaled networks is separating itself from the long tail.
That shift is happening for several reasons.
Retailers have rich first-party purchase data.
Retailers need not guess whether an individual is looking for toothpaste, laundry detergents, cosmetic goods, pet food, or infant products because they already know their shopping habits.
That data can support audience creation, campaign optimization, and sales measurement.
Physical retail still matters
Retail media is sometimes discussed as if shopping happens entirely inside websites and apps.
This statement is wrong.
Walmart, through its 2026 study, revealed that between 80 and 94 percent of the respondents had engaged in both online and in-store shopping.
That makes omnichannel measurement increasingly important.
Brands want more retail-specific measurement
Traditional advertising often struggles to connect an impression with a product purchase.
Retailers can sometimes close more of that gap because the same company operating the media environment also participates in the transaction.
That does not mean every retail media attribution number should automatically be accepted as incremental revenue.
It means the underlying data can be much closer to the actual purchase.
The IAB’s recent guidance on incremental measurement specifically emphasizes experiments, counterfactual models, econometric methods, and other approaches for determining whether commerce media actually caused additional business outcomes.
Walmart Connect: What Brands Need to Know
Walmart Connect is Walmart’s retail media business and is designed around one of the industry’s broadest combinations of physical retail, ecommerce, first-party transaction data, and off-site media.
Its major advantage is not simply the number of people who shop at Walmart.
It is the ability to connect digital discovery, physical retail, ecommerce transactions, and advertising exposure within one retail ecosystem.
Walmart says its advertising business reaches approximately 150 million customers and uses online and in-store transaction data for audience targeting and measurement.
Walmart Connect’s Advertising Ecosystem
Walmart Connect currently spans several major areas:
- Sponsored Search
- Onsite Display
- Offsite Media
- Brand Shop and Shelf
- Store Ads
- CTV
- Social advertising
- Walmart DSP
- Audience targeting
- Closed-loop measurement
This makes Walmart particularly relevant to brands that already sell through Walmart or intend to build a Walmart retail presence.
Sponsored Search
Sponsored search puts products in front of customers while they are actively searching and browsing Walmart’s digital properties.
This is the closest equivalent to traditional retail marketplace advertising and can be particularly useful when the objective is product discovery and conversion.
The strategic advantage is intent.
Someone searching for a particular product category is already providing a useful commercial signal.
The challenge is that search inventory is limited and competitive. A brand cannot assume that increasing bids alone will solve weak product-market fit, poor pricing, low reviews, weak availability, or poor creativity.
Onsite Display
Walmart’s onsite display offering extends beyond search and gives advertisers placements across areas such as the homepage, search pages, item pages, and topic pages.
Walmart says advertisers can use contextual, behavioral, keyword, and broader site targeting, including purchase behavior from online and in-store shopping.
This makes onsite display more useful for consideration campaigns than a pure bottom-of-funnel search strategy.
Offsite Media
Walmart Connect can also take Walmart shopper audiences beyond Walmart-owned properties.
This is increasingly important because retail media networks are moving toward full-funnel advertising rather than limiting themselves to sponsored products.
In 2026, Walmart announced expanded ways for advertisers to activate Walmart audiences and measurement through external buying environments, including Yahoo DSP and VIZIO inventory through Magnite.
Additionally, Walmart has revealed a proof of concept where Walmart first-party audiences will be used within Google Display & Video 360 on YouTube campaigns, demonstrating how retail media is becoming integrated with programmatic and video advertising.
Store Advertising
This is one of Walmart Connect’s clearest differentiators.
Walmart offers digital screens, audio, and experiential placements inside stores. Its store advertising product covers touchpoints from entrance areas to aisles and checkout environments.
For brands where physical shelf presence matters, this can change the role of media from simply generating an online sale to influencing a shopping trip already underway.
Social Advertising
Walmart Connect has also been expanding beyond retailer-owned environments.
In 2026, Walmart announced self-service access to Walmart-powered social campaigns beginning with Meta, with additional platforms planned. The system is designed to allow advertisers to use Walmart audiences across social and onsite placements while managing campaigns through Walmart Connect’s Ad Center.
That matters because the distinction between “retail media” and “social media” is becoming less useful when retailer audiences can be activated outside the retailer’s own website.
Target Roundel: How Target’s Retail Media Network Works
Roundel is the name of Target’s retail media business, based on access to Target shoppers and Target’s first-party data.
While Walmart Connect relies primarily on mass retail strength and omnichannel reach, Roundel relies significantly on shopper audience, brand building, premium media, and the larger cultural and merchandising context of Target.
Target announced double-digit growth for Roundel in its Q4 and FY 2025 earnings, which shows that Roundel is now an integral part of Target’s commercial world.
Roundel’s major advertising products
Roundel currently offers:
- Target Product Ads
- Display
- Programmatic
- CTV
- Search Ads
- YouTube Ads
- Social
- Influencer marketing
- Precision Plus
This is important because Roundel is not simply a sponsored-product platform.
It is positioning itself as a broader media business.
Target Product Ads
Target Product Ads provide sponsored product placements on Target.com and the Target app.
According to Roundel’s current advertising guide, Target Product Ads are self-service.
For brands already selling products on Target, this can provide a relatively direct connection between shopper intent and product visibility.
Display & Premium Publishers
Roundel’s display offering can run across Target’s digital properties as well as more than 150 premium publishers, according to Roundel.
That gives advertisers a path beyond the Target website while retaining access to Target-defined audiences.
This distinction is important for upper- and mid-funnel campaigns.
A consumer does not have to be on Target.com at the moment of exposure for the brand to use Target shopper signals.
Programmatic
Roundel’s programmatic offering allows advertisers to activate Target audiences through a DSP environment.
Roundel describes its audiences as cookie-free and designed around Target shopper signals.
For sophisticated media teams, this creates an opportunity to bring retail audience intelligence into a broader programmatic buying strategy.
CTV
Roundel also offers CTV advertising.
CTV can be useful when a brand wants to build awareness before a shopper begins actively searching for a product.
The important strategic question is not whether CTV produces immediate ROAS in the same way as a sponsored product ad.
It is whether the campaign can introduce a brand to relevant shoppers and then connect that exposure with later retail activity.
Precision Plus
One of Roundel’s more notable 2026 developments is Precision Plus, described by Roundel as an AI-powered performance engine designed for activation, optimization, and scalable performance.
Target’s 2025 annual report also identified Precision Plus as a retail media capability using data and AI learning to improve advertising outcomes.
This reflects a wider industry trend: RMNs are moving from selling inventory toward selling increasingly automated optimization systems.
Instacart Ads: The Grocery-First Retail Media Model
Instacart Ads is structurally different from Walmart Connect and Roundel.
Instacart is not simply one retailer’s advertising business.
It operates a marketplace and advertising ecosystem connected to a broad collection of grocery retailers, giving brands access to shoppers across multiple retail environments.
Instacart says its advertising ecosystem reaches 55+ million consumers and spans more than 100,000 storefronts on Instacart Marketplace and 310+ retail media networks and marketplaces.
Its measurement proposition is also heavily connected to grocery purchase behavior.
Why is Instacart different?
Think about a packaged food company that markets its products via numerous supermarket brands.
Building separate advertising strategies for every retailer can create operational complexity.
Instacart’s model can potentially provide a centralized way to reach grocery shoppers across a wider retail footprint.
That is its core strategic advantage.
High-intent grocery behavior
Grocery advertising has a distinctive characteristic: many shopping journeys are short and highly practical.
A customer may open a grocery app because they need:
- Cereal
- Milk
- Snacks
- Coffee
- Frozen food
- Cleaning products
- Personal care products
- Pet supplies
The distance between advertising and purchase can therefore be relatively short.
That makes Instacart particularly relevant for brands where purchase frequency, basket placement, substitutions, and category demand matter.
Closed-loop measurement
Instacart emphasizes closed-loop reporting tied to actual shopping activity.
Its advertising platform describes access to real-time reporting and measurement across a network of grocery retailers.
This can be highly valuable for CPG brands because the commercial question is often not simply:
“Did someone click?”
It is:
“Did advertising cause more people to buy my product instead of a competitor’s product?”
The second question is much harder.
Expansion of Ads Manager
Instacart added Ads Manager for retailers in May 2026, enabling its retail partners to use the new self-service tools and expanding its advertising platform past brand advertisements. According to Instacart, it was earning over $1 billion from ads and others in 2025.
This indicates a larger strategy, as Instacart is becoming increasingly strategic about its technology serving as infrastructure for grocery commerce media.
Walmart Connect vs. Target Roundel vs. Instacart Ads
The most useful comparison is not a simple winner-versus-loser table.
Each network solves a different commercial problem.
| Factor | Walmart Connect | Target Roundel | Instacart Ads |
| Core strength | Omnichannel mass retail | Target shopper ecosystem | Grocery purchase intent |
| Retail footprint | Walmart stores + Walmart.com | Target stores + Target.com | Broad grocery retail ecosystem |
| First-party data | Online + in-store behavior | Target shopper behavior | Grocery shopping behavior |
| Sponsored products | Yes | Yes | Yes |
| Onsite display | Yes | Yes | Yes |
| Offsite media | Yes | Yes | Yes |
| CTV | Yes | Yes | Available through ecosystem |
| Social | Expanding self-service capabilities | Managed social offerings | Broader grocery ecosystem |
| In-store media | Strong | More limited relative to Walmart’s store network | Retailer-dependent |
| Programmatic | Walmart DSP and partners | Yes | Platform ecosystem |
| Self-service | Strong and expanding | Available for selected products | Strong |
| Best strategic fit | Mass-market and omnichannel brands | Target-focused brands and lifestyle categories | Grocery and CPG |
| Key measurement advantage | Walmart transaction data | Target shopper data | Grocery transaction data |
| Major limitation | Walmart-specific ecosystem | Target-specific ecosystem | Grocery-heavy focus |
The table should not be interpreted as a ranking.
It shows why a brand’s retail footprint matters more than the raw size of the advertising platform.
The funnel is the frame, not the pitch
Retail media started as a performance channel — sponsored products near the buy button. That’s no longer the full story.
All three networks are now selling against the entire journey:
Awareness > Discovery > Consideration > Search > Purchase > Repurchase
But they don’t do it in the same way. Each one leans on a different structural advantage. That’s the part worth paying attention to, not the feature checklist.
Walmart Connect: physical footprint is the differentiator.
Walmart’s pitch isn’t “we have more formats”. It’s “we can follow the customer into a store and still count the sale.”
What it actually offers:
- Upper-funnel video and CTV — expanded significantly through Connect Select, a curated CTV marketplace inside Walmart DSP, built on Vizio inventory (Walmart owns Vizio) plus supply from Paramount, Warner Bros. Discovery, and SSPs like Magnite and PubMatic
- Onsite display — targeted using first-party browsing and purchase history, separate from search-intent targeting
- Sponsored search and sponsored products — the original CPC engine, still the highest-volume format
- Offsite reach — via partnerships with The Trade Desk, TikTok, and Roku, not a native social product
- In-store media: self-service checkout screens, TV walls, samples, and in-store audio advertising
The key benefit: closed-loop measurement – connecting ad impressions to in-store purchases, not online purchases. That’s genuinely hard for a pure e-commerce retailer to replicate. Amazon can’t do this at Walmart’s physical scale. This is Walmart’s actual moat — not the format list.
The caveat: most of this full-funnel capability sits behind managed service or DSP complexity, not self-serve simplicity. A small or mid-size brand isn’t touching Connect Select or in-store formats without a real budget and a media partner.
Roundel: breadth through partnerships, not ownership
Target doesn’t try to build every channel itself. Roundel’s model is “use Target’s data to buy media everywhere else, better.”
What it actually offers:
- Target Product Ads — sponsored listings on Target.com and the app, self-serve via Media Studio, no DSP access fee
- CTV — via Roku and Fire TV
- Search — placements on Google, using Target’s own shopping data
- Social — Facebook, Instagram, Pinterest, Snapchat
- YouTube video
- Influencer marketing — through LTK, Target’s approved network, spanning TikTok, Instagram, and Pinterest
- Display — on Target.com plus 150+ third-party premium publishers
- In-store and DOOH — a managed-service addition, not part of the core self-serve stack
The real advantage: Target Circle first-party data applied to media it doesn’t own. Buying a Facebook ad through Roundel isn’t different from buying it directly — except the targeting and attribution are anchored to actual Target purchase behavior.
The caveat: self-serve is limited to Target Product Ads. Everything else — display, CTV, social, influencers, in-store, needs managed-service engagement, and third-party reviews note this consistently as a friction point. Compared to Amazon’s retail media scale, Target buyers describe Roundel as strong inside its own ecosystem but limited in overall reach.
Instacart: depth in one moment, width across retailers
Instacart isn’t trying to own the full funnel the way Walmart or Target are. It’s optimized for one high-value moment: the open cart, active list, decision in progress.
What it actually offers:
- Sponsored product listings — the core format, placed where purchase intent is already confirmed
- Display and shoppable video ads
- Off-platform extension — through Meta and other partners, letting advertisers reach Instacart audiences outside the app
- Carrot Ads — a white-label retail media infrastructure now licensed to other retailers, extending Instacart’s ad technology to over 1,800 national, regional, and local retail banners across roughly 85,000 stores in the US and Canada, including convenience chains like Vroom Delivery
The real advantage: this isn’t “Instacart plus a broader network” as a side benefit — it’s becoming retail media infrastructure for grocery and convenience retail generally. A brand can buy once and effectively reach shoppers across dozens of banners that don’t have their own ad stack.
The caveat: this is still overwhelmingly a bottom-funnel, high-intent channel. The awareness and consideration stages of the funnel are thin here compared to Walmart or Target. Instacart’s growth story is about scaling reach across retailers, not about building CTV or influencer capability to match the other two.
Audience Targeting and Shopper Intent
A major mistake is treating all retail audiences as equally valuable.
They are not.
Let us consider three potential audiences:
- Audience A: Those that have shown an interest in skin care content.
- Audience B: Those that have bought facial moisturizer in the past six months.
- Audience C: Those actively searching for a facial moisturizer.
All three audiences are valid, but they indicate varying purchase intent.
The best retail media campaigns make use of varied audiences in varied phases.
High-intent targeting
It is during the mid-funnel phase that sponsored searches and product placements usually have the most effect.
Mid-funnel targeting
Display and audience-based advertising can help influence consideration before active search.
Upper-funnel targeting
CTV, video, social, influencer, and broader programmatic campaigns can introduce products before a specific shopping need is fully formed.
The mistake is using one audience type for every objective.
A brand launching a new product may need broad discovery first.
An established product with strong organic demand may need to defend search visibility.
A mature brand may need both.
Retail Media Measurement and Attribution
Measurement is where retail media becomes complicated.
The platforms can provide highly valuable transaction-level signals, but advertisers should not assume that every attributed sale represents an incremental sale.
That distinction is critical.
Attribution is not the same as incrementality
Suppose a consumer searches for a brand they already planned to buy.
They click a sponsored product.
The retailer attributes the sale to the ad.
The sale is real.
The attribution may also be technically correct under the platform’s rules.
But the ad may not have caused the purchase.
The consumer may have purchased anyway.
This is why experienced media teams separate three questions:
- What sales were attributed?
- What sales were incremental?
- What was the true profit impact?
These are not interchangeable metrics.
Metrics to monitor
Depending on the campaign, useful metrics include:
- ROAS
- Incremental ROAS
- Cost per acquisition
- Cost per new-to-brand customer
- New-to-brand revenue
- Conversion rate
- Click-through rate
- Cost per click
- Average order value
- Units sold
- Repeat purchase
- Sales lift
- Reach
- Frequency
- Incremental revenue
- Contribution margin
The IAB’s incremental measurement guidelines recommend stronger approaches such as experiments, counterfactual methods, econometric models, and hybrid approaches when the objective is to establish true business impact.
The attribution window matters
Two networks can report different ROAS numbers while both are calculating correctly.
Why?
Because they may use different:
- Lookback windows
- View-through rules
- Click attribution
- New-to-brand definitions
- Conversion definitions
- Deduplication methods
Therefore, never compare:
Walmart ROAS = 5.0x
against:
Instacart ROAS = 6.0x
and immediately conclude that Instacart is more efficient.
Which Network Fits Different Types of Brands?
Instead of asking which network is universally best, begin with distribution.
The first question should be:
Where can consumers actually buy the product?
If a product is not meaningfully available through the retailer, advertising efficiency can become irrelevant.
CPG and grocery brands
When it comes to grocery-oriented brands, both Instacart and Walmart Connect can be very relevant, based on distribution strategy.
In situations when the objective is to communicate with grocery consumers across multiple banners, Instacart will enjoy an advantage over Walmart Connect.
However, Walmart Connect will become very relevant when Walmart is a significant distribution partner.
Beauty brands
Beauty brands should examine where their products are actually sold.
A brand with strong Target distribution may benefit from Roundel’s combination of Target audiences, product advertising, display, CTV, social, and influencer capabilities.
A mass-market beauty brand with significant Walmart distribution may find Walmart Connect’s broader omnichannel ecosystem more relevant.
The decision should follow retail distribution, not category stereotypes.
Home and household products
Household products can benefit from Walmart’s scale and physical-store presence, especially when purchase decisions occur during routine shopping trips.
However, if the product has strong Target distribution and the brand depends heavily on lifestyle positioning, Roundel’s broader media options may deserve testing.
DTC brands entering retail
This is one of the most interesting use cases.
A DTC brand moving into Walmart or Target should not treat retail media simply as a sales channel.
It can also function as a retail expansion testing system.
For example, a brand could:
- Launch products through an eligible retail marketplace.
- Use retail media to generate demand.
- Measure new-to-brand customers.
- Study geographic demand.
- Analyze repeat purchase.
- Identify strong categories or SKUs.
- Use those signals to support retail expansion conversations.
How to Choose a Retail Media Network in 2026?
Most brands don’t lose money on retail media by picking the wrong network. They lose money by picking a network for the wrong reason — because a competitor is on it, because a sales rep made a compelling pitch, or because the platform is the one getting written about. This framework exists to slow that down and force a decision based on where the brand actually sells, not where the industry is talking.
Use it before allocating budget, and use it again every two quarters, because the right answer changes as distribution and objectives shift.
Step 1: Map your retail distribution honestly
Before touching ad formats, get the commercial facts straight. Build a simple table.
| Retailer | Revenue Share | Store Availability | Ecommerce Availability | Growth Potential |
|---|---|---|---|---|
| Walmart | % | Yes/No | Yes/No | High/Medium/Low |
| Target | % | Yes/No | Yes/No | High/Medium/Low |
| Grocery partners | % | Yes/No | Yes/No | High/Medium/Low |
Where this needs more precision than the original version gives it:
- “Revenue share” should mean share of total retail sales through that banner, not share of media budget. If you’re spending 40% of your ad budget on a retailer that drives 5% of your sales, that’s the finding — not something to smooth over.
- “Growth Potential” shouldn’t be a gut-feel label. Use something tangible here: an expansion of distribution that has already been agreed upon, the rate of growth in the category at this retailer, or your sell-through rate for the past two quarters. Without a specific number or signed agreement, leave it marked as “unknown” and not “high.”
- Add a row for contractual exclusivity or MDF (market development fund) obligations. Some retailers tie shelf placement or co-op funding to minimum ad spend. If that’s true for you, the retailer’s ad platform isn’t really optional — it’s a cost of doing business, and it shouldn’t compete for budget against the others in this table as if it were discretionary.
This table has one job: telling you which retail media networks are even eligible for a serious budget, based on where you actually sell a product today not where you’d like to sell it, and not where competitors are visibly advertising.
A brand with 60% of sales through Kroger and Albertsons has no real reason to prioritize Roundel over a grocery-focused network, no matter how good Target’s influencer program looks in a case study deck.
Step 2: Identify your primary objective, but be honest if you have two
The original framework asks you to pick one objective. In practice, most brands are chasing two at once, and pretending otherwise leads to budgets getting split evenly across formats that don’t reinforce each other.
Objective A: Increase immediate sales
Prioritize high-intent placements — sponsored product and sponsored search, wherever your buyer already searches.
Objective B: Acquire new customers
Prioritize audience targeting, off-site media extension, and new-to-brand measurement — meaning the retailer can actually show you a customer who hadn’t bought your product (or category) from them before the campaign ran, not a proxy metric dressed up to look like one.
Objective C: Build awareness
Use display, video, CTV, social, and influencer formats, aimed at broader audience activation rather than people already searching for you.
Objective D: Increase physical retail sales
Prioritize networks with real store-level measurement — not just “the retailer has stores,” but a platform that can tie a digital ad exposure to an in-store transaction using loyalty or payment data.
Objective E: Expand distribution
Use retail media performance as a demand-validation tool — proof to a retailer’s buying team that a product sells, ahead of a listing or shelf-space negotiation.
The fix the original framework needs: if you genuinely have two objectives — say, immediate sales on your top retailer and awareness-building on a newer one — split the budget explicitly by retailer or by time horizon, and score each allocation separately using Step 4. Don’t force one weighted model to serve two different jobs. That’s how a scorecard ends up recommending a compromise that satisfies neither goal.
Also worth naming directly: Objective E is different in kind from the other four. It’s not a media objective, it’s a negotiating tool. Don’t run it through the same measurement lens as a sales or awareness campaign, the win condition is a buyer meeting, not a ROAS number.
Step 3: Assess your measurement maturity and don’t buy more than the decision needs
A small brand needs simple reporting: impressions, clicks, conversions, return on ad spend. That’s often genuinely enough.
A large enterprise brand may eventually need:
- Incrementality testing
- Marketing mix modeling (MMM)
- Clean-room data analysis
- Retailer data collaboration agreements
- Cross-channel attribution
- Formal experiment design (holdout groups, geo-testing)
The rule the original draft states correctly but doesn’t operationalize: don’t buy sophisticated measurement because it exists. Buy it when the size of the decision in front of you justifies the cost of getting the answer wrong.
Here’s a rough way to apply that instead of leaving it as a slogan:
- If a single retailer’s budget is under roughly 250K–500K annually, basic platform reporting is almost always sufficient. The cost of an incrementality study or a clean-room integration will exceed the value of the precision it buys you.
- If you’re deciding whether to shift seven figures from one retailer’s platform to another, or whether a national TV-adjacent CTV buy is actually driving incremental sales versus just reaching people who’d have bought anyway, that’s when incrementality testing and MMM earn their cost.
- Clean-room analysis and retailer data collaboration make sense mainly when you’re trying to build a durable, retailer-specific audience strategy over multiple quarters — not for a single campaign decision.
Skipping this discipline is how brands end up paying for enterprise measurement tooling that produces reports nobody in the room can act on, because the underlying decision was never big enough to need that level of certainty.
Step 4: Score commercial fit, not platform popularity
Build a weighted evaluation model instead of ranking platforms by reputation.
| Factor | Suggested Weight |
|---|---|
| Retail distribution | 25% |
| Audience relevance | 20% |
| Purchase intent | 15% |
| Measurement quality | 15% |
| Media inventory | 10% |
| Operational ease | 10% |
| Expansion potential | 5% |
These weights aren’t fixed. They should move with the business model:
- A grocery CPG brand should probably push purchase intent and retail distribution higher, since the buying decision is fast and repeat-driven.
- An emerging lifestyle or DTC-related brand entering into retail for the very first time should place more emphasis on audience relevance and upper-funnel media, because the objective is to generate demand prior to the sale and not capture intent that doesn’t exist.
What the original draft fails to mention altogether, and shouldn’t:
- Minimum spend thresholds. Several networks — Roundel and Walmart Connect included — gate their more advanced formats (CTV, influencer, in-store) behind managed-service minimums, sometimes in the tens of thousands of dollars per quarter. A scoring model that ignores minimum commitment size will recommend a network a brand can’t actually afford to activate properly.
- A decision threshold. A weighted score is meaningless without a cutoff. Decide in advance: does a network need to score above a certain number to get funded at all? Does the gap between the top two scores need to exceed some margin before you shift budget, so you’re not reallocating spend over noise?
- A review cadence and exit criteria. This is the biggest gap. The framework should specify: recheck this scorecard every two quarters, and set a rule for when a network gets cut — for example, if it hasn’t hit target ROAS or new-to-brand share for two consecutive quarters despite adequate budget and creative investment. Without this, a network that scored well eighteen months ago keeps getting funded on inertia rather than current performance.
- A test-before-commit step. Before assigning a full budget based on the scorecard, run a capped pilot — four to six weeks, fixed budget — on any network scoring in the top two but with limited track record. The scorecard tells you where to look first. It shouldn’t replace an actual read on performance before real money moves.
A Practical Retail Media Budget Framework
There is no responsible universal minimum budget for Walmart Connect, Roundel, or Instacart Ads because media access, category competition, distribution, campaign objective, and buying model can all affect the economics.
Instead, start with the amount of money required to generate a useful learning cycle.
Phase 1: Test
The goal is not maximum ROAS.
The goal is to determine:
- Which products respond?
- Which search terms matter?
- Which audiences convert?
- Which creative works?
- Which retailers produce the strongest economics?
- How much incremental demand exists?
Phase 2: Validate
Once early signals appear, test:
- Audience expansion
- New creative
- Broader placements
- Off-site campaigns
- New-to-brand acquisition
- Geographic expansion
Phase 3: Scale
Only after the economics are understood should the team consider significantly increasing spend.
Scaling a bad campaign simply makes the problem larger.
Build a retail media P&L
Do not stop at ROAS.
A more useful model is:
Revenue – COGS – Retail Fees – Media Cost – Fulfillment – Discounts – Agency/Platform Costs = Contribution
This is particularly important for low-margin products.
A campaign producing 5x ROAS can still be commercially weak if the product margin is too low.
Frequently Asked Questions
What’s the actual difference between Walmart Connect, Target Roundel, and Instacart Ads?
Walmart Connect covers the most ground. It runs across Walmart’s stores and website, using purchase data from both, and stretches into CTV, social, and in-store screens.
Roundel is Target’s version of the same idea but built more around partnerships than ownership. It uses Target’s shopper data to buy smarter media on other platforms, Google, YouTube, Facebook, and Pinterest alongside its own product ads and display inventory.
Instacart plays a different game entirely. It’s grocery-first, and its real strength isn’t just its own app, it’s the network of grocery retailers running on Instacart’s ad technology behind the scenes.
Do you have to sell products at Walmart to use Walmart Connect?
No. Sellers on Walmart Marketplace and eligible suppliers can advertise there, and Walmart also opens up some of its inventory to non-endemic brands — companies that don’t sell products through Walmart at all. Consider companies like insurance firms, streaming services, or financial institutions investing in a self-service screen or audio ad inside Walmart.
What the platform can provide access to depends on what kind of product is being purchased. It would be wise to check the eligibility for each respective format.
Does a brand need to sell at Target to get value from Roundel?
Selling at Target definitely helps a lot of what Roundel does is built around Target’s own shopper data, so the products tied to Target.com and the app work best if you’re actually on those shelves.
But Roundel also sells media that has nothing to do with being a Target vendor: programmatic display, CTV, search ads on Google, social campaigns. Those run on Target’s targeting data but land on other platforms. So don’t write off Roundel just because you’re not sold at Target — look at the specific product before deciding.
Is Instacart only worth it for grocery brands?
Grocery and everyday CPG products are where Instacart makes the most sense, because that’s the shopping behavior the whole platform is built around.
That said, other categories can work too, as long as they’re actually distributed through the retailers Instacart’s network reaches. The real question isn’t “am I a grocery brand”, it’s “can shoppers actually find and buy my product through the stores Instacart connects to.” If the distribution isn’t there, the ad spend won’t do much.
Which network has the best first-party data?
Wrong question, honestly. More data doesn’t mean more useful data it depends on what you’re trying to learn.
Walmart’s advantage is connecting online and in-store behavior, which nobody else does at that scale. Target’s data is strong specifically for understanding its own shoppers. Instacart’s strength is grocery purchase patterns across a wide set of retailers.
Select the solution based on what the specific behavior is that needs to be changed, not on what the firm has inflated figures for.
Is there any reason why smaller brands should pursue retail media at all?
Absolutely, but do so selectively. The biggest issue that smaller brands face is trying to spread out a small budget across three to four different solutions, and therefore none of them are being funded sufficiently to provide meaningful insight.
Select a single solution, have a single clearly defined objective, and fund it sufficiently to deliver the result.
Besides ROAS, what else should brands actually track?
ROAS tells you if the ad paid for itself. It doesn’t tell you much else. Also keep an eye on:
- Incremental revenue — sales you wouldn’t have gotten anyway
- Contribution margin — not just revenue, but what’s left after costs
- New-to-brand customers — are you growing or just re-selling to existing buyers
- Customer acquisition cost
- Repeat purchase rate
- Sales lift during the campaign window
- Units sold
- Average order value
- Retail availability — ads don’t help if the product’s out of stock
- Organic sales — sometimes ads pull sales that would’ve happened anyway, and this shows you that
