Download a copy of the full report here.
A baker’s dozen worth of Mars UnitedSM Commerce’s most experienced thought leaders from across our many practices examine the trends that will shape the industry in 2026 — along with the things we think should happen, but probably won’t.
And why should you care? Because we’ve gotten pretty good at this whole “prediction thing.” We’ve been making these annual forecasts since 2021, and since then have established a pretty strong track record for correctly identifying the key trends that are moving the industry forward — along with the pesky issues (like measurement standardization) that we can’t seem to solve.
So while few industry analysts ever really look back to see how well their predictions held up, we decided to do exactly that this time around. And to help us, we enlisted help from one of the most obvious trends of 2026: generative AI.
We fed our 2025 forecasts into three different AI platforms — Google’s Gemini, Microsoft’s Copilot, and Perplexity Pro — and asked them to use collective industry knowledge to determine how well we did. Here’s how they rated us:
Copilot: We were completely accurate with 75% of our predictions and “mostly” or “partially” accurate with everything else. The hope to see in-store media become more measurable and commerce activation evolve into personalization at scale were the two areas we only got “partially” correct — so you can’t fault us for trying.
Gemini: Google’s platform apparently chooses to be a little less definitive with its conclusions, although it did assess every one of our “What Will Happen” predictions as “likely successful.” As for our “What Should Happen (but probably won’t)” desires, “Many of the predictions that focused on the desirable state of the industry were less accurate, highlighting the challenges of changing entrenched practices and overcoming systemic barriers,” Gemini concluded (like we needed gen AI to tell us that).
Perplexity Pro: We’ll just let the conclusion tell the whole story here: “The Mars United Commerce document from 2024 made remarkably accurate predictions about the retail media landscape in 2025. Out of 31 distinct predictions, approximately 84% proved correct or accurately assessed, demonstrating strong industry foresight.”
And with those reviews in mind, we invite you to read our predictions for “What to Expect at Retail in 2026.”
____________
Amy Andrews
President
WHAT WILL HAPPEN:
TikTok will become a top retail media network for CPG brands.
We’ve all witnessed the growth of TikTok Shop this year. To win on traditional retail shelves, CPGs increasingly first need to win on social shelves to bring new users into their brands and categories — especially younger demographics. To win in 2026, successful brands and retailers don’t need a new playbook, they just need to meet shoppers where they are. And they’re increasingly shopping — and buying — on TikTok.
What if we treated TikTok like a retail channel as well as a social channel? Brands should consider reallocating retail media dollars from underperforming retailer networks to TikTok to drive incremental sales across all potential channels.
WHAT SHOULD HAPPEN (but probably won’t):
All shelves — physical, digital, social, and agentic — will become addressable.
At the moment, digital and social shelves are addressable and can be highly targeted through identity-based audiences. Brands and retailers are now racing to the agentic shelf. Because the physical shelf historically hasn’t been addressable, it’s been given less attention in the last year even though most shoppers still make their purchases in-store. When our industry makes the physical shelf addressable, we’ll be able to truly connect and win across every shelf.
To get there, brands and retailers should test & learn with physical, addressable shelf content that connects to digital, social, and agentic shelves. There already are opportunities to leverage existing data to make physical shelves more relevant and engaging for shoppers and to collect new data on what content at the physical shelf drives more conversion.
____________
Kandi Arrington
EVP-Retail Consultancy
WHAT WILL HAPPEN:
Retail in 2026 will be defined by the rise of agent-directed commerce.
When AI assigns the trip, allocates the basket, and determines the most efficient path to purchase across retailers on behalf of the consumer, friction is removed from the shopper journey at a scale the industry has never seen. Mission-based shopping — birthday parties, home projects, replenishment runs — will increasingly be solved end to end by agents who surface themes, curate options, optimize fulfillment, and redirect spend to retailers the shopper may have never considered. The result: a quiet, powerful redistribution of value across the retail landscape.
This will accelerate ecommerce faster than forecasts due to the emergence of “Showrooming 2.0.” Agents reduce the need to touch, test, or compare products in person and instantly give shoppers confidence in the decision. Household essentials, electronics, beauty, and multi-step shopping missions will shift online in outsized volume. At the same time, retail media networks will begin optimizing for agent behavior by building new structured content formats, evolving placements to be machine-readable, and dealing with more complex attribution as agents increasingly influence the last mile of consideration and conversion.
Competitive hierarchies will also destabilize. Category captains long protected by shelf dominance and search visibility will be displaced when agents identify better, clearer, or more relevant options. Challenger brands and emerging subcategories will surge, driven by superior benefit signaling and cleaner structured content. In this new ecosystem, brands that organize their product data for machines while optimizing content for human inspiration will pull ahead fastest.
WHAT SHOULD HAPPEN (but probably won’t):
There will be sweeping industry alignment around the standards required for agentic commerce to reach its full potential.
Retailers should collaborate on unified product data requirements, shared taxonomies, and transparent agent-inclusive attribution. Retail media networks should overhaul their pricing and measurement models to reflect actual value creation rather than impression volume. Brands should rebuild their content systems from the ground up for machine readability rather than simply retrofit what exists. And retailers should lean into agent-driven fulfillment — even when it means losing portions of the basket.
Entrenched incentives, legacy systems, and channel protectionism will hold many players back, however, opening the door for early adopters to gain a disproportionate advantage in the agentic era.
____________
Tyrel Beutler
SVP-Ecommerce
WHAT WILL HAPPEN:
Non-public brand/retailer APIs will replace scraping as the primary source of information for AEO/GEO shopping recommendations.
ChatGPT, Perplexity, and other AI agents will rely heavily on “official” data to provide the most up-to-date information instead of relying on data of unknown provenance and quality. This will concurrently limit the influence of uncontrolled data sources like Wikipedia or Reddit to influence the agent’s suggestions and results.
To help this happen, brands and retailers should develop a plan to address catalog and data quality that goes beyond periodic updates to their PIM (product information management) systems. Agentic/generative agents will be using different information and adding, removing, and creating new attributes regularly. Expertly responding to these changes and requests will become a new discipline.
WHAT SHOULD HAPPEN (but probably won’t):
Watermarks will be used to identify AI-created digital advertising.
As AI creative becomes indistinguishable from traditional creative, a clear understanding of who — or what — created the content will become a differentiator. Human-generated content will be seen as premium, quality work, while the AI output is viewed as more commodity than craft (if it isn’t already).
Brands should listen to customer feedback about creative. There certainly are efficiencies and cost savings to be had, but if advertising all starts to look, feel, and sound the same, how will your brands stand out in a sea of AI-generated work? If everyone has access to AI creative, what will make it authentic?
____________
Willy Blesener
SVP, Media
WHAT WILL HAPPEN:
Agentic AI will have a major impact on commerce marketing efficiency.
AI — agentic, generative, and otherwise — will have a major impact on commerce marketing next year through the deployment of back-office efficiency tools that I expect to become widespread in 2026. Amazon Ads alone unveiled a number of solutions at unBoxed in November, a few of which even have agentic AI capabilities.
As just one example, creative automation is a breakthrough for the industry. Mars United’s partners at Profitero+ have built a tool that gathers intelligence from the digital shelf, identifies and prioritizes search keywords, then uses a generative AI copywriting tool to optimize content for product detail pages. The tool builds content six times faster than traditional methods, and it’s significantly improving conversion rates, too.
This doesn’t mean brands should hold off on investing in generative and agentic AI search optimization at all. In fact, we’re building a suite of tools to help our clients get ahead of the curve. But it’s important to note that success in GEO/AEO will require many of the same foundational elements that have been driving effective SEO all along — so while we’re preparing for this future, we’re continuing to strengthen current best practices.
WHAT SHOULD HAPPEN (but probably won’t):
The industry will establish a standard metric for measuring incrementality.
Brands are investing hundreds of millions of dollars in retail media across multiple retailer platforms. They need a way to more effectively compare results across those platforms as well, not just for attributed sales and ROAS — which, by the way, aren’t standardized either — but for incremental sales lift. It’s the key metric on which brands are now focused, yet retailers continue to measure and report it using various definitions and methodologies.
Mars United has a solution that does a great job standardizing results across platforms. Other third-party tech providers have tools as well. But it would be a lot easier — and far more accurate — if the industry adopted standards. This needs to happen. I’m just skeptical that it will next year.
____________
Victoria Cromie
Managing Director, Canada
WHAT WILL HAPPEN:
Retail media networks in Canada will expand their audience and measurement capabilities.
The impact will be significant. Brands will be able to leverage more impactful audience data in all their campaigns, not just the ones with a significant investment. It will also ensure they are reaching the right audiences. Measurement will be unlocked, too, to deliver the proper metrics and reporting for all campaigns, regardless of the investment level.
Brands should continue to push their retailers partners to provide these capabilities in all cases. The shift to retail media in Canada is moving fast, and retailers should be responding to the needs of their brand partners just as quickly.
WHAT SHOULD HAPPEN (but probably won’t):
Retail media networks will standardize their measurement practices.
It’s the only way for brands and their agency partners to properly measure results and assess the potential of each retail media network to make informed investment decisions.
It’s not happening yet, which is why brands must continue to ask “Why not?” or, perhaps more pointedly, “When?” We’ve seen significant changes happen in the marketplace over the past 24 months as a result of brands pushing for more.
____________
Steve DeVore
Chief Creative Officer
WHAT WILL HAPPEN:
AI prompt writers will become a new role embedded in agency creative and production departments.
As the demand from brand clients for “faster/cheaper” continues to ramp up, the most significant opportunity exists in automating monotonous tasks (like digital media versioning) to allow the human talent to work on the more complex tasks to be done. To do it well, though, we’ll need specialists who can write tighter prompts that reduce the back and forth required to refine the asks of AI tools.
Marketers should be patient and not fall for headlines or hype videos. Attention to detail will be needed to get these AI solutions to a place where they’re actually benefitting complex commerce ecosystems.
WHAT SHOULD HAPPEN (but probably won’t):
Scalable, real-time content will be streamed to the shelf in-store.
The ability to change in-store messaging at the click of a mouse would give brands and retailers the opportunity to leverage trends as they happen vs. catching the long tail. Shoppers are shopping creators as much as they are brands, and to pair creator endorsements with physical products at the point of purchase could fuel purchase intent like never before.
While considering the infrastructure required for a broad technology rollout, don’t be afraid to test & learn, fail & refine.
____________
Craig Elston
Group SVP-Strategic Planning & Analytics
WHAT WILL HAPPEN:
Gen Alpha will deliver the next wave of shopping culture.
Gen Alpha begins to enter full consumer independence in 2026. They are no longer children, but active participants in today’s marketplace. Therefore, understanding this generation is no longer a matter for the future but has become an immediate strategic imperative.
The first cohort to be born entirely in the 21st century, Gen Alpha will rival the spending power of both Gen Z and Millennials by 2029, numbering over 2.5 billion members globally — which will make them the largest generation in history.
They are truly the first tech-native generation, having never known a world without touchscreens, Alexa, or AI. Despite this, Gen Alpha is showing a significant preference for physical retail experiences, according to Numerator. They crave human interaction; they see retailers such as Sephora and Target as “playgrounds” or “amusement parks,” enjoy out-of-the-home shopping as family time, seek tactile experiences, and develop their own sense of independence in the process.
Being truly tech-native, they fully expect an omnichannel experience that encompasses discovering new brands via YouTube and TikTok and seamless integration with the physical world through BOPIS, augmented reality, unified shopping carts, and in-store interactive kiosks.
This generation also has a fundamentally different relationship with media and influence. The trust their favorite online influencers as highly as their parents. YouTube dominates, but TikTok is growing significantly within the cohort and Roblox is where many Gen Alpha live, play, and increasingly shop.
Brands and retailers will need to continue developing new muscles as this generation creates a fundamental reset in expectations. Along with being the most tech-immersed generation, they’re also the most values-driven. Brands must meet them where they are and resonate with shared values, because the generation that will soon inherit the world is busy reimagining it.
WHAT SHOULD HAPPEN (but probably won’t):
The failed experiment with self-checkout should be put to rest.
Retailers should return to investing in their staff to build relationships with customers. The frustrations of self-checkout, and its societal contribution to the loneliness epidemic, should compel every retailer to remember its role in local communities. Retail has always been an environment for human interaction, for helping each other, for inquiring about life, and we are poorer societies because corporations have removed one of the simplest, nicest potential interactions for people in their day.
____________
Darren Keen
CEO, International Markets
WHAT WILL HAPPEN:
Brands and retailers will actively “shop the world,” lifting ideas from distant markets as a core input to local strategy.
Playbooks will globalise faster as leaders mine unfamiliar markets for formats, media models, and shopper experiences that will feel fresh at home. Local moats shrink as ideas travel instantly, but cultural translation and execution quality must be decisive. The winners will borrow boldly while adapting sensitively — not just copying executions.
Marketers should formalise “outside-in” market scanning, with teams accountable for sourcing examples from structurally different geographies rather than just the usual benchmarks. They should build lightweight test & learn muscles to localise ideas quickly for creative, operations, and measurement; partner with agencies and platforms that bring cross-market visibility; and reward internal champions who successfully import and adapt global ideas.
WHAT SHOULD HAPPEN (but probably won’t):
Companies will build systematic, reciprocal learning networks where markets can give and take beyond “best practice.”
If it happened, global commerce marketing would feel less extractive and more collaborative. Smaller or “non-priority” markets would shape innovation rather than just receive it. Playbooks would be richer, more inclusive, and more resilient to shocks. In reality, however, hierarchy, budget flows, and the dominance of traditional “big markets” will keep knowledge moving mostly one way.
But we can try. Organizations should design governance that treats every market as a potential source of originals, not just pilots. They can fund cross-market experiments where learnings are required to flow in both directions, and shift recognition from “who originated the idea” to “who scaled and evolved it.” Without structural change, the industry will keep watching the world — but only selectively listening.
____________
Julia Miller
GVP, Commerce Media
WHAT WILL HAPPEN:
Brands will treat answer engines as a new distribution channel.
Retailers will establish and expand direct partnerships with third-party AI platforms, demanding deeper contextual product data and richer attributes from brand partners but also unlocking opportunities for them to develop agent-specific assortments, merchandising strategies, and promotions tailored to how these answer engines guide product discovery.
Brands must upgrade their product catalogs with complete, contextual, purchase-intent attributes while also developing unique SKUs, bundles, and promotion strategies to connect with this new shopper journey. Retailers should modernize item-setup workflows and partner with brands to make this emerging channel feel distinct, focused on discovery rather than efficiency.
WHAT SHOULD HAPPEN (but probably won’t):
Retailers will reward product data quality more than media spend.
Prioritizing a brand’s data completeness over its media investment would improve shopper relevance, reduce ad clutter, and ensure that answer engines surface the most-relevant products for shoppers. But the business pressures on retail media networks make it unlikely that retailers will turn 180 degrees away from paid opportunities, even as answer engines bypass traditional ad placements to rely on richer product attributes.
In their joint business plans, brands should negotiate benefits for delivering enriched product catalogs similar to the incentives currently tied to retail media spending (incremental merchandising/displays, preferred placement, etc.). In parallel, they should ask for opportunities to pilot ads in the retailer’s answer engine, like Sparky and Rufus.
____________
Justin Peabody
VP-Creative Director
WHAT WILL HAPPEN:
In the age of AI, transparency will become mandatory.
We’ve all heard the phrase, “Seeing is believing.” But with the rapid evolution of generative AI, that’s no longer the case. Seeing alone isn’t enough. Shoppers need more reassurance than ever that the brands and products they’re purchasing are actually what they claim to be.
While authenticity has been a trend for years, radical transparency is the next step. From marketing photoshoots to product ingredients and manufacturing, brands should share real behind-the-scenes content, showcasing the real work being accomplished by real humans to create real products — all driving real brand connection.
WHAT SHOULD HAPPEN (but probably won’t):
Brands will stop chasing trends and start building communities.
Because the next trendy aesthetic is now just a prompt away, brands will have to be more intentional about who they are and who they want to reach. Breaking through the endless stream of algorithm slop will require more than a cool package design or color-coordinated Instagram grid.
Brands that create products to solve unmet shopper needs will survive far beyond the latest me-too launch. Choiceful activations and partnerships aligned with audience values and passion points will move the needle more than an empty celebrity endorsement. Brands need to find their audience’s bullseye — then be relentless about hitting it.
____________
Michele Roney
EVP-Retailer CX
WHAT WILL HAPPEN:
The fight among media networks for non-endemic advertisers will get more competitive.
The retail media marketplace will continue to become more competitive, especially in the non-endemic space, due to the accelerated entrance of non-traditional “commerce media networks” — non-retailers that have strong shopping and conversion data from a high-quality, niche audience. I expect 2026 to bring an explosion in the launch of networks in the travel, financial, telecommunications, and transportation arenas, among others.
With Amazon and Walmart still dominating the space, the other 200-plus traditional media networks already facing stiff competition will only experience more pressure. Their endemic brands — whether CPG, electronics, pet, apparel — will increasingly find non-endemic opportunities on newer commerce media networks.
What will make a network stand out in fiercer competition? Networks will have to broaden, deepen, and scale their data and analytics offering to demonstrate how their audiences drive tangible business impact for brands. Go-to-market strategy and demand generation will also play a pivotal role because networks will need to bring strong propositions to potential advertisers.
WHAT SHOULD HAPPEN (but probably won’t):
Retail media networks will level up their analytics and measurement capabilities.
Advertisers will continue to grow their investments in retail media, but only with networks that prove out a tangible impact on their business. Data, analytics, and measurement are crucial, but I don’t expect more than a handful of those 200-plus networks to bring more sophisticated offerings to the market in 2026. Yet the networks that fail to innovate and build analytics as core competency will start falling behind next year.
____________
Prabhpreet Sidhu
VP-Analytics
WHAT WILL HAPPEN:
The retail media marketplace will contract as smaller networks merge or form federations to survive.
The “long tail” of retail media will start to collapse in 2026 as brands — who currently lose roughly 20% of revenue to fragmentation inefficiencies — pull spending from niche networks. This is forcing a shift from direct sales to unified, programmatic buying, with budgets concentrated at dominant players and smaller retailers becoming inventory sources rather than standalone platforms. And that will lead to a decline in the RMN ecosystem.
Brands need to consolidate their spending into strategic partners and demand programmatic access to the tail. Smaller retailers should stop building proprietary walled gardens and join federated cooperatives. This will allow them to pool data and inventory, thereby offering brands the scale and ease of buying they require to continue investing.
WHAT SHOULD HAPPEN (but probably won’t):
The industry will establish a universal, cross-retailer measurement standard across all retail media networks.
Instead of transparent, apples-to-apples performance comparisons, fragmentation will persist, costing brands that 20% in revenue. “Vanity metrics” will obscure true business value, making it hard for commerce marketers to prove incrementality to their CFOs and finance teams. And smaller networks may lose investments as brands consolidate spending with dominant players offering “good enough” proprietary data.
Brands should not wait for a mandate; they must invest in their own solutions — independent measurement providers, internal dashboards and solutions — to unify the data. Retailers should at least voluntarily embrace partial standardization via clean rooms to differentiate themselves through transparency and earn brand trust.
____________
Kelly Wearmouth
Managing Director, Australia-New Zealand
WHAT WILL HAPPEN:
The number of retail media networks in ANZ will increase as the commercial opportunities offered by existing platforms multiply.
The marketplace fragmentation we are already experiencing will proliferate, bringing with it calls for standardisation and, where practical, consolidation. It will become increasingly complex — and unviable financially, in some cases — for brands and their agency partners to create well-aligned trade and retail media plans as the number of networks grows past 15.
To manage this, brands should absolutely remain curious and investigate which networks and what offerings provide an additive consumer experience while delivering measurable business impact. Following those learnings, they should be prepared to make tough choices about which networks to keep in their plan.
Retailers should continue to consider and build upon what makes their offering unique to both brands and their shoppers. They also should think deeply about the infrastructure behind their offering, building something that can flex as their network matures and the needs of brand clients evolve.
WHAT SHOULD HAPPEN (but probably won’t):
The silos that we continue to work around will come down, providing seamless planning and booking across all brand investments.
This will enable marketers to tackle some of the huge changes that are coming at us — one example being the influence that large language models will have on the brands consumers choose and the ones AI agents preference. Navigating this new playing field requires everyone to be at the planning table together rather than working exclusively in their “silos.”
Brands should build teams, structures, and processes backwards from what the landscape looks like now — but be prepared for the plan to become obsolete in a year’s time. They should find partners that can flex along with them and double down on the constant curiosity needed to keep answering the question, “Is there a better way?”
