Thirteen brand bets worth watching in 2026.
A field guide for anyone building products, brands, or platforms — thirteen recent examples, each presented as a 3-slide deck. Tap any deck to open the full-screen presentation.
Compiled by Free Intelligence
Updated April 30, 2026
13 decks · 39 slides
The companies below are doing something most strategy decks describe but few firms actually attempt: betting brand equity on a category they don't yet know how to operate in. Each example sits where legacy trust meets unproven technology — exactly where most innovation either compounds or combusts.
For each, we kept the same lens: what is the company actually trying to do, what class concept does it light up, and — if you were running it — how would you move forward? Below, the headline deck first, then twelve more — including three that didn't go to plan, because the wind-downs teach more than the launches.
No. 01 — Headline Deck
Mattel × OpenAI
A 79-year-old toy company hands its most-loved IP to the most-watched AI lab in the world. The upside is massive. The blast radius is bigger.
Brand ExtensionGenerative AITrust as AssetChildren
Slide 1 · Intro
Betting the toy box on conversational AI.
Mattel + OpenAI announced a strategic partnership in mid-2025. First AI-enabled products in 2026 across Barbie, Hot Wheels, Fisher-Price, UNO.
Slide 2 · Objective & Concepts
Defending legacy IP in an AI-native market.
Value Creation · Consumer Insights · Brand Extension · Co-creation · Trust as Strategic Asset.
Slide 3 · Viewpoint
Strategically necessary, tactically dangerous.
Lead with safety. Pilot Hot Wheels first. Sell parent dashboards as core. Avoid open-ended chat in v1.
Open full deck →
No. 02
Spotify AI DJ
A synthetic voice that introduces your songs, explains why you're hearing them, and remembers what you skipped last week.
MusicVoice AIPersonalization
Slide 1 · Intro
Replacing the radio host with a model that knows you.
AI DJ pairs Spotify's curation engine with synthetic voice host. Multilingual. Request mode. Adapts as you listen.
Slide 2 · Objective & Concepts
Defending listening time against TikTok and YouTube.
Service-layer value creation · passive-listening insight · brand personality literally is the DJ voice.
Slide 3 · Viewpoint
Smartest thing since Discover Weekly.
Localize the voice. Let artists guest-host. Open the API. Counterprogram the algorithm.
Open full deck →
No. 03
Universal Music × Ethical AI
After suing Suno and Udio, UMG turned around and signed deals with KLAY, ElevenLabs, and others — choosing to be a counterparty instead of a plaintiff.
Music IndustryIP LicensingDefensive Innovation
Slide 1 · Intro
Licensing the catalogue before it gets stolen.
Two-track strategy: litigation against scrapers + licensing deals with compliant AI startups. Set the price floor early.
Slide 2 · Objective & Concepts
Existential threat → new royalty stream.
Value Capture · stakeholder marketing (artists, not listeners) · platform vs. pipe · defensive innovation.
Slide 3 · Viewpoint
Smart on licensing, weak on artist comms.
Public royalty dashboard. Tiered consent. Watermark masters. Make UMG the easiest licensor — speed becomes the moat.
Open full deck →
No. 04
Adobe Firefly
While competitors trained on the open internet and hoped for the best, Adobe trained Firefly on its own Stock library — and turned that constraint into the marketing.
Creator ToolsProvenanceB2B Brand
Slide 1 · Intro
"Ethically trained" as a feature, not a footnote.
Trained only on Adobe Stock + licensed data. Content Credentials follow assets. Enterprise gets indemnification.
Slide 2 · Objective & Concepts
The AI tool that legal departments approve.
Differentiation through constraint · B2B prices risk · provenance as product · standards as strategy.
Slide 3 · Viewpoint
Most underrated AI strategy of the year.
Push C2PA as open standard. License Firefly to platforms. Pay contributors visibly. Sell the indemnification.
Open full deck →
No. 05
Coca-Cola Y3000
A limited-edition Coke whose flavour profile was shaped by consumer sentiment data and an AI co-creation tool — packaged as "the taste of the future."
CPGCo-CreationStorytelling
Slide 1 · Intro
A flavour co-created with consumers and a model.
Third entry in the Creations platform. Consumer mood data + AI tooling + AR companion app. Limited-edition by design.
Slide 2 · Objective & Concepts
Keeping the world's most distributed brand culturally relevant.
Co-creation · limited-edition risk-cap · brand-as-platform · the campaign IS the product.
Slide 3 · Viewpoint
A marketing artifact, not a product launch.
Publish the prompt data. Open Creations to artists. Localize the next drop. Stop trying to be Pepsi.
Open full deck →
No. 06
LEGO × Epic Games
A long-term partnership to build a digital play space inside Fortnite that respects parental controls and LEGO's 90-year safety reputation.
Legacy IPPlatform StrategyKid-Safe Design
Slide 1 · Intro
Building the kid-safe metaverse before someone else does.
Multi-year LEGO/Epic alliance. Chat controls, time limits, parental visibility as defaults — not opt-ins.
Slide 2 · Objective & Concepts
Extending physical play into the platform Gen Alpha already uses.
Platform choice as strategy · brand halo · adjacent innovation · partnership over build.
Slide 3 · Viewpoint
Cleanest "legacy IP enters digital" of the past two years.
License the kid-safe playbook. Tie physical SKUs to in-game items. Bring third-party builders in. Don't go to Roblox.
Open full deck →
No. 07
Duolingo AI-First
When a beloved brand restructures its workforce around generative models — and the cute green owl gets a McKinsey memo.
Operating ModelBrand AuthenticityWorkforce
Slide 1 · Intro
The cute green owl restructures around AI.
Duolingo Max ships GPT-class features. Internally: phase out contractor translators, AI-fluency in performance reviews.
Slide 2 · Objective & Concepts
Operating-model innovation hiding inside a product story.
Cut content cost · scale languages · capture value via Max · risk: efficiency erodes the qualities driving love.
Slide 3 · Viewpoint
Right strategic call, weak narrative discipline.
Reframe AI-first as more-human. Show the savings on screen. Protect the owl. Pay translators a transition fund.
Open full deck →
No. 08
Heinz × DALL-E
Heinz prompted DALL-E with "ketchup" and got back image after image of bottles that looked exactly like its own. Then it ran the screenshots as the campaign.
Brand EquityEarned MediaIconicity
Slide 1 · Intro
AI's bias as proof of brand iconicity.
Even AI thinks ketchup looks like Heinz. Rethink agency turned algorithmic bias into a billboard claim.
Slide 2 · Objective & Concepts
Iconicity, demonstrated by a third party with no incentive to lie.
Brand equity measurement · earned media engineered into creative · distinctive brand assets · cultural fluency.
Slide 3 · Viewpoint
Cleverest "AI as PR" execution to date.
Don't repeat the trick. Open-source a Brand Iconicity Index. Defend the silhouette legally. Pair with a hunger charity.
Open full deck →
No. 09
Liquid Death
Water in a tall-boy can, sold like a punk-rock energy drink. No AI. No tech. Just a $1.4B valuation built on category re-framing.
CPGBrand VoiceNo-AI Counterexample
Slide 1 · Intro
Water sold like a heavy-metal energy drink.
"Murder your thirst." Launched as a Facebook video before the product existed. $1.4B valuation by 2024.
Slide 2 · Objective & Concepts
Steal share from Red Bull and Monster — by selling water.
Category re-framing · positioning · distribution into Live Nation venues and convenience · content-as-marketing.
Slide 3 · Viewpoint
Innovation doesn't always mean AI.
Own the festival cup. Don't translate the voice — re-cast it. Launch Liquid Death Studios. Resist the IPO.
Open full deck →
No. 10
Sephora AI Beauty
Color IQ for foundation matching, Virtual Artist for AR try-on, AI Beauty Advisor in-app — Sephora has spent a decade building the digital experience layer everyone else now copies.
RetailOmnichannelAR · AI
Slide 1 · Intro
Replicating in-store expertise online.
Color IQ scanner + Virtual Artist AR try-on + AI Beauty Advisor in-app. Omnichannel as engineering, not slogan.
Slide 2 · Objective & Concepts
Reduce returns. Raise basket size. Transfer trust to mobile.
Omnichannel experience · AI in retail · trust transfer · data network effect compounds with each match.
Slide 3 · Viewpoint
Sephora is winning beauty's digital battle quietly.
Open Color IQ as a standard. Pair AI with humans. Pay associates for digital coaching. Publish the bias audit.
Open full deck →
No. 11 — Walked Back
Klarna AI Customer Service
The first major company to publicly replace 700 customer-service agents with a chatbot — and the first to quietly start re-hiring humans twelve months later.
Operating ModelCustomer ExperienceAI Maturity Curve
Slide 1 · Intro
700 agents replaced — then partially re-hired.
2024: AI handles 2/3 of chats, parity on CSAT. 2025: messaging pivots from "AI replaces" to "AI augments."
Slide 2 · Objective & Concepts
Cost out, fast — then learn what AI alone can't do.
Operating-model innovation · value capture · CX risk · the AI maturity curve everyone hits eventually.
Slide 3 · Viewpoint
Right experiment, oversold conclusion.
Tier the funnel visibly. Publish quality on disputes. Reposition as copilot. Pay the kept agents like seniors.
Open full deck →
No. 12 — Wound Down
Nike .Swoosh / RTFKT
Nike acquired digital-sneaker studio RTFKT for ~$300M in 2021, launched .Swoosh in 2022, and quietly wound it down in 2025. The lesson is in the retreat.
Web3Hype-Cycle RiskStrategic Retreat
Slide 1 · Intro
The Web3 bet that didn't survive its hype cycle.
RTFKT acquired 2021. .Swoosh platform 2022. Nikeland in Roblox. RTFKT wound down by 2025; orderly migration paths.
Slide 2 · Objective & Concepts
Right idea, wrong moment.
Brand extension into virtual goods · hype-cycle risk · strategic retreat · brand permission.
Slide 3 · Viewpoint
Lessons in retreat — what to keep when the experiment ends.
Keep the talent, drop the wrapper. Re-platform the lessons. Honour collectors. Write the post-mortem publicly.
Open full deck →
No. 13 — Late-Mover
Yum Brands × Nvidia AI Drive-Thru
Taco Bell, KFC, and Pizza Hut roll out AI ordering after McDonald's failed in public. The drive-thru is the worst place to learn AI in public — Yum is doing it slower.
QSRAI in OperationsLate-Mover Advantage
Slide 1 · Intro
Rolling out AI drive-thru after McDonald's failed in public.
Yum × Nvidia partnership. Taco Bell hundreds of locations in 2025. McD's IBM pilot ended 2024 after viral TikTok failures.
Slide 2 · Objective & Concepts
Speed, accuracy, labour cost — at the loudest touchpoint there is.
AI in operations · CX risk · fail-publicly cost · late-mover advantage when the leader has already been laughed at.
Slide 3 · Viewpoint
Drive-thru is the third place to ship AI, not the first.
Lead with chat. Always-on human override. Train on TikTok failures. Localize voice. Publish accuracy rate.
Open full deck →
Pattern across all thirteen.
Each company is using AI or a new platform not to invent a new business, but to defend the value of an old one. Even Liquid Death, the no-AI counter-example, is doing the same job — defending water against a category that thought it owned the cooler.
The three walked-back stories — Klarna's headcount reversal, Nike's RTFKT wind-down, McDonald's drive-thru exit — are the most useful examples on this list. Successful launches can be luck. Public retreats are evidence. The companies that publish their lessons compound trust. The ones that don't repeat the same mistakes.
The work isn't building a model — it's deciding which part of your brand equity to put on the line, and which part to keep behind glass.
The hardest decision in innovation strategy isn't "what to build." It's "what to risk."