Flat Budgets, Rising Scrutiny:

According to Gartner’s 2025 CMO Spend Survey, marketing budgets remain flat at 7.7% of company revenue, with 59% of CMOs reporting they don’t have enough budget to execute their strategy. The pressure to "do more with less" means every marketing dollar must be justified, and vanity metrics won’t cut it. An October 2024 InMarket survey asked U.S. brands and agencies for their top investment priorities for 2025. The results are telling:

 According to Gartner’s 2025 CMO Spend Survey, marketing budgets remain flat at 7.7% of company revenue, with 59% of CMOs reporting they don’t have enough budget to execute their strategy. The pressure to "do more with less" means every marketing dollar must be justified — and vanity metrics won’t cut it. An InMarket survey asked U.S. brands and agencies for their top investment priorities for 2025. The results are telling: attribution and measurement are the top priority (47%), reflecting the demand to prove ROI in a privacy‑tight market. 

15 Channels, 1 Question:

How do we measure what really works? The average marketer now juggles 15 media channels, Yet just 30% feel confident in their ability to measure aggregate return on investment - and 54% plan to reduce spending.

Offline Still Rules:

While most marketers are doubling down on digital, offline sales remain dominant, with over 80% of all shopping still happening in-store. But in-store environments are notoriously difficult to measure, leading to underinvestment and missed opportunities. This disconnect between digital efforts and offline conversion creates blind spots. Without attribution, marketers can't close the loop or improve strategy.

Signal Loss + Privacy Pressure:  

As third-party cookies disappear and privacy laws like GDPR/CPRA tightening, signals are further degrading. Privacy by design is now table stakes. Teams are reorganizing around first-party data, identity anonymization, and privacy-preserving tech; the industry broadly anticipates continued signal loss and new state privacy laws. The IAB’s State of Data work shows ~9 in 10 buyers changing personalization and spend mix; 71% are increasing their first‑party datasets that don’t depend on third‑party tracking.

Retail Media’s Limits: 

Retail Media Networks (RMNs) like Walmart Connect and Kroger Precision Marketing have exploded in recent years, giving brands powerful, closed-loop ad channels. McKinsey estimates RMN spend could reach $100B by 2026; 68% of marketers say RMN is more important to their strategy than a year ago. But RMNs are also walled gardens. Each platform has its own data formats, standards, and measurement models. According to the Association of National Advertisers (ANA 2024), 55% of advertisers cite lack of standardization as the top RMN challenge.
 

 Modern attribution isn’t a single tactic - it’s a system. Leaders combine methods that

1. Verified Purchase Attribution

(receipt / loyalty ID / digital barcode / card-linked)

Closed-loop attribution connects ad exposure to verified purchases, not just clicks. It replaces proxy metrics with real outcomes, i.e. who bought, what SKU, where, and when. Done well, it also generates consented first-party data you can reuse for remarketing and media calibration.

Receipt-Validated Attribution (retailer independent)

Shoppers submit a photo of their receipt after buying a product; technology like OCR reads the retailer, date/time, and line item SKUs; fraud rules dedupe and flag suspicious claims. Because it spans many retailers at once, this method delivers near-real-time readouts without waiting for each RMN’s reports. It’s ideal for national promos, seasonal pushes, and launches where breadth and speed matter - and it builds an opt-in audience you can recontact and calibrate your marketing media mix against.

• Snipp enables this with white-label reward programs where users upload receipts. For example, Reckitt’s "Schiff Rewards" program used Snipp’s receipt validation platform to confirm purchases across any channel (in-store, online, and DTC), enabling segmentation by store, basket size, and repeat rate — all without waiting for retailer data feeds.
• Recent ecosystem moves, like Circana expanding its U.S. receipt panel to 200K static panelists and NIQ unifying retail measurement with its consumer panel, are making receipt‑based reads richer and faster.
   

Loyalty‑ID Matches (retailer‑collaborative)

Brands can see new vs existing buyers, SKU detail, and basket context. Loyalty ID matching is especially powerful within a single retailer’s ecosystem and is commonly used to measure in-store media, on-site ads, and off-site campaigns that leverage the retailer’s audience. Brands can partner with RMNs that enable media exposure to be linked to loyalty card purchases — creating a ‘closed loop’ between ad spend and sales outcomes.

• Tesco (UK): Tesco’s Media & Insight Platform analyzed more than 1,000 campaigns with Clubcard usage - allowing SKU-level closed-loop measurement across digital and in-store media. When digital media was added to in‑store activity, average sales uplift was about +43.75% higher than in‑store‑only plans. This highlights the impact of blended campaigns that follow the shopper across channels — and the value of closing the loop from exposure to purchase.
• Boots (UK): Boots has started providing more transparent attribution across its digital and physical retail environments, by combining Advantage Card audience data and offline sales transaction data. Early analysis has shown a 22% uplift in ROAS when integrating Boots’ online and offline sales, compared to online only. This shows how a unified view across channels reveals value that single‑channel reports miss.

Card‑Linked Offers (bank ID matching):  

Consumers opt in and link a payment card, so qualifying purchases are recognized automatically. Card-Linked Offers are ideal in categories where traditional loyalty programs are weak or fragmented (for example, dining, quick‑service restaurants) and provide retailer‑agnostic coverage. On their own, CLOs may lack SKU detail; but paired with receipts or loyalty overlays — and matched in a clean room — they add scale and help unearth true lift (incremental sales driven by the offer).

• Ibotta uses a triangulated approach: combining card‑linked data, receipt uploads, and randomized holdout groups to measure what really works. Their method, called “Know Your Source”, has been used in over 3,600 tests. In campaigns with statistically significant results, the median incremental sales lift was about 49.9%.

2. Privacy-Safe Data Collaboration 

(Data Clean Rooms + Standard Protocols)

Clean rooms are secure spaces where brands and partner platforms can compare encrypted data to answer questions like, “How many people who saw this ad later bought the product?” No one sees the other’s raw data; only aggregated results leave the room. Clean rooms are now the standard way to connect retailer, publisher, and brand data without compromising privacy. With interoperable standards created by the IAB like DCR 1.0, PAIR, and ADMaP, brands and retailers can match datasets safely, deduplicate exposure, and standardize reports that can be audited.

• Nectar360 x Channel 4 (UK): Nectar360 (Sainsbury’s loyalty program) and Channel 4 used InfoSum clean rooms to match connected‑TV (CTV) ad exposure with loyalty purchases. Reported results showed up to 122% sales uplift for participating brands.  

3. Incrementality Experimentation & MMM Upgrades

The challenge with last-click attribution has become increasingly apparent: it’s no longer enough to show that a campaign ran successfully — you need to prove causality, i.e., did your campaign cause a lift in sales, or would those sales have happened anyway? This question is especially critical for social media platforms where consumers discover brands but don't immediately purchase. Even on platforms like TikTok Shop, most purchases of discovered brands can't be captured via last-click measurement.

That's where incrementality testing comes in. By comparing test groups to control groups, these experiments isolate the sales lift caused by your campaign — not just correlated with it. The results of these experiments are now powering modern Marketing Mix Models (MMMs). TikTok is seeing an increase in media mix modeling (MMM) as marketers move away from last-click attribution, and 61.3% of US marketers are aiming to improve their MMM in the future. Unlike older MMMs that updated once a year, modern versions are faster, follow industry standards for measuring outcomes, and rely on data that’s been validated through testing — along with controls for things like promotions and inventory.

Publishers should collaborate with brands to incorporate MMM with sales lift data, while brands should leverage sales lift results in their models to prove campaign value and secure larger budgets. MRC standards now provide frameworks for disclosing assumptions, attribution windows, and statistical confidence intervals.

• Catalina x Applegate Natural & Organic Meats: they created a custom, sequential omni-channel campaign that aimed to drive trial, bring back lapsed buyers, and enhance loyalty among its users. To make ad dollars stretch they suppressed an offer to those who already purchased an item or incentivized those who hadn’t after having been exposed to brand advertising a set number of times. The campaign’s 27MM ad impressions and in-store sequential offers drove a 42% increase in dollars per trip for existing buyers, a 25% sales lift, and an incremental increase in Return on Ad Spend (ROAS) of $2.13.

• Google x Nielsen: According to an NCS study, when Google and Nielsen added results from real lift tests from controlled experiments (e.g., geo-matched tests, household holdouts --- see Annex for information on lift tests) for 10 YouTube campaigns into their MMM, the model showed much higher ROAS -- on average +84%, with the median campaign more than doubling its effectiveness (2.6×). What this means: models get more accurate when you feed them proven test results. 

4. Commerce Media Convergence

(Shoppable Social + CTV + RMN Integration)

Commerce media collapses the traditional funnel by directly connecting audience impressions with omnichannel transactions. By linking content and commerce, brands can better serve their customers through more relevant offers and incentives, all in a privacy-protected way. While building brand relevance and favorability will always be an essential part of the marketing mix, today’s opportunity is to connect impressions to actual sales and feed these results back into true full-funnel marketing models.

The category now encompasses retail media networks (RMNs), shoppable ads across online and in-store digital screens, and "shoppable TV" where viewers buy products directly from their devices — with closed-loop sales reporting delivered back through clean rooms.

The impact is already measurable across channels. In-store digital screens help 32% of shoppers discover new products, while partnerships between RMNs and CTV platforms enable brands to combine premium video with direct attribution, creating true full-funnel campaigns.

And momentum is growing: McKinsey predicts that 20% of global e-commerce sales could come from live shopping by 2026. 46% of consumers have made at least one purchase via shoppable commerce, and 38% of US ad buyers will increase focus on shoppable ads in 2025, according to December 2024 IAB data. It is also one of the major e-commerce trends in Europe for 2025, with 37% of European consumers shopping via livestream.

• Danone x Walmart Connect (U.S.): Danone partnered with Walmart Connect and agency teams to launch International Delight Cold Foam Creamers via a full-funnel omnichannel plan. Tactics spanned CTV, Walmart.com homepage and category takeovers, pre-roll video, Pinterest, TV wall and self-checkout screens, display, and sponsored search. Walmart’s first‑party data attributed 110,000 new buyers, 11.29% sales lift, and $15.52 ROAS across tactics. The takeaway: when CTV and retail media run together against the same shopper identity, you can see awareness convert into baskets.
• Instacart and Roku. With Roku’s shoppable ad formats, consumers can jump straight from their TV to their Instacart account via text message or QR code. Advertisers can also place ads on the Roku Home Screen and target consumers with Instacart’s first-party data.
• Pinterest and Instacart This partnership allows Pinterest advertisers to target based on Instacart purchase behavior and will eventually enable closed-loop attribution. With Instacart's near-instantaneous fulfillment, users can complete purchases in just a few clicks and receive items in as little as 30 minutes. This is particularly powerful given that 47% of Gen Z consumers use Pinterest as a search engine, with similarly high adoption for Millennials (39%) and Gen X (37%).

Different Retailers, Different Rules

Two networks can both claim “lift” and mean different things. Solve this by using common internal KPI set – IAB has set out an in-store retail media playbook that offers standards you can follow, like traffic and unique visitors (footfall), dwell time (time in view), a clearly defined ad impression (opportunity to see), engagement signals (e.g., attention time), and outcomes. When possible, run a retailer‑independent check (receipts or card‑linked) alongside retailer reports so you can compare apples to apples.

Slow Reporting (latency that kills agility)

NIQ research revealed that brands perceive just 23% of retailers share campaign data and analysis with advertisers in real-time while 56% deliver data only on campaign completion. Another 21% take a week or more to provide this information after a campaign ends. When speed matters, use receipt pipelines for near‑real‑time readouts and ask RMNs for weekly updates, then reconcile with the final audited report.

Fraud and Gaming

Purchase-based programs can be gamed (users may upload the same receipt twice, alter images, or submit from multiple accounts) and risk inflated claims without validation. If you run receipt validation programs, insist on image forensics, de-duping rules, bot filtering, Ai-powered image intelligence and behavioural signals as standard operating procedure -- especially when outcomes are used for pricing.

Measurement Readiness

 Many teams still lack the data science or experimental design capabilities needed for modern attribution. According to IAB’s State of Data report, nearly three-quarters of companies are experiencing and expecting more hurdles with measurement — the key to understanding performance and ROI. This is pushing ad investment to walled gardens like RMNs to power closed-loop measurement.

 

Privacy‑safe identity becomes the default

Expect broader use of clean rooms and standard ways to match data without cookies. This will make cross‑partner attribution more comparable and less custom.

In‑store measurement gets standardized

Agreed definitions for how in‑store impressions and purchases are counted will make it easier to compare results across retailers and markets. That unlocks benchmarking and better planning. New IAB/MRC standards, along with IAB Europe’s 2024 Retail Media Measurement Standards, aim to create comparable, auditable methods for SKU-level, in-store attribution.

Always‑on incrementality 

Lift testing will move from one‑off studies to a regular part of campaign execution. Retail media networks (RMNs) are already adding “incremental sales” to standard reports once minimum spend levels are met. Kroger Precision Marketing rolled out incremental sales measurement for qualified The Trade Desk buys, and Walmart Connect has publicly signaled broader rollouts (conversion lift, in-store attribution) across formats.

Attention as a bridge metric

Attention is more than just a measure of engagement — it’s becoming a useful early signal for what might drive sales. By combining things like time-in-view, eye-tracking, and machine learning, attention metrics can help brands choose better ad placements before they spend money. When these attention scores are connected to real sales data (i.e. from receipts or loyalty programs) they help identify which media and creative are likely to perform best. A study by IAS and NCSolutions found that high-attention ads led to a 157% boost in incremental sales and a 40% increase in verified sales lift. Anonymous sensors can track flow, dwell, and product interactions to improve placements and creative. While attention doesn’t replace sales-based measurement, it helps reduce wasted spend by narrowing focus early on.

AI for smarter insights

AI doesn't replace attribution; it amplifies it when fed verified outcomes. Expect models that spot under-performing pockets early, flag budget re-allocation opportunities, and predict incremental response by audience or placement. Industry research shows the pivot is already underway — IAB State of Data 2024/2025 notes rapid AI adoption precisely to cope with privacy-driven signal loss and measurement complexity.

• Amazon Brand+ exemplifies this transformation. The new AI-powered solution optimizes CTV and online video ads by leveraging trillions of shopping, browsing and streaming signals across Amazon channels (Prime Video, Twitch) and third-party publishers (BuzzFeed, Dotdash Meredith). Beta testers report over 10% sales increases and 70%+ website traffic lifts, positioning Brand+ alongside Google’s Performance Max and Meta’s Advantage+ as major AI-driven attribution tools. Most significantly, it demonstrates how AI enables retail media networks to unify upper-funnel video with verified purchase outcomes.

Going forward, AI will enable three key capabilities:

Predictive Attribution: Machine learning models forecast likely outcomes based on historical and real-time data

Dynamic Budget Allocation: Attribution signals feed into planning tools that reallocate spend across channels and tactics in real-time

Creative and Offer Optimization: Attribution insights will influence not just where ads appear, but what they say and offer

The shift from descriptive analytics to prescriptive decisioning is what turns attribution into a revenue driver. AI will optimize campaigns in-flight, not just after the fact, using verified sales signals to build smarter targeting and creative.

Start simple, prove lift: Run a receipt-validated campaign to get fast, retailer-independent proof of purchase.

Choose 2–3 retailer partners: Prioritize those offering loyalty-ID or clean-room access for richer insights. Agree on reporting cadence and what counts as success before launch.

Standardize success metrics: Use one scoreboard everywhere: ROAS, iROAS (incremental sales ÷ spend), sales lift, new-to-brand rate, repeat rate, and basket lift. Build in fraud checks (duplicate receipts), windows (e.g., 3/14/30 days) and control for promos and stock-outs so results aren’t distorted.

Speed up the feedback loop: Push for faster attribution from RMNs. Ask for weekly directional readouts and plan check-ins to rebalance spend. Close with a fully audited report a few weeks after the campaign ends.

Scale what works (with causality): Feed verified outcomes into AI models and MMMs to optimize future spend. Re-run winners via geo-matched tests or sequential rollouts to confirm lift. Expand retailer coverage only when your scoreboard stays consistent.

Protect privacy by design: Work within protocol-adhering clean rooms; use hashed IDs; agree on data retention and access rules. This speeds approvals and makes cross-partner comparisons easier.

Shoppers no longer follow a linear path to purchase. A single basket might be shaped by a mobile ad during breakfast, a shelf display at lunch, and an app offer at checkout. In a connected shopping universe where online and in-store are part of one continuous journey, the goal should shift from slow, fragmented vanity metrics to fast, verifiable sales outcomes. Even if those outcomes are triangulated through ‘good-enough’ methods.

The pursuit of “attribution perfection” can slow growth.

Instead of chasing a perfect system, brands should agree on a simple, shared scorecard and start with methods reliable enough to guide decisions. The biggest barrier is often organizational, not technical: aligning finance, media, and shopper teams on attribution windows, standardized metrics, promotion controls, stock-out visibility, and reporting cadence. That alignment — not technological sophistication — creates momentum.

Marketers can't expect e-commerce-level precision from in-store measurement right away. Other markets demonstrate that starting with practical approaches pays off. Europe offers a helpful model: retail media there grew in-store first, not digital-first, which led to grounded, outcomes-based approaches. That foundation is now scaling: RMN spend in the UK is projected to rise from ~$4.66B in 2024 to ~$9.57B by 2029, with Germany ($2.71B → $5.63B) and France ($1.37B → $3.0B) also accelerating — underscoring that in-store attribution is becoming standard practice, not an experiment.

Cadbury’s Easter campaign with Tesco illustrates the power of blended execution. To celebrate its 200th anniversary, Cadbury activated across physical and digital touchpoints: branded power aisles, in-store digital signage, scan-as-you-shop ads, and shelf talkers with recipes. The result? Cadbury’s best-ever Easter sales at Tesco. The lesson is clear: when digital and physical work together, performance improves.

Attribution doesn't need to be perfect; it needs to be reliable, directional, and fast enough to influence the next budget decision. As privacy rules tighten, budgets flatten, and journeys grow more complex, the winners will be brands that stop waiting for perfect clarity and start acting with verified confidence. The four pillars outlined here — verified purchase attribution, privacy-safe collaboration, proven incrementality, and commerce media convergence — provide that foundation.

Together, they offer a realistic path from guesswork to SKU-level outcomes, fast enough to influence this quarter's plan.

Download the full report here