2025 Marketing Fails: 10 Campaigns, Brand Decisions, and Backlashes

Some marketing failures are ads. Others are pricing decisions, creator choices, product policies, or technology launches that turn into public brand problems. In 2025, all of those categories produced backlash, and the useful lesson is not simply that a campaign became controversial. It is whether the brand understood the risk, stayed consistent with its promise, and measured the business outcome after the attention arrived.

This list keeps the original ten examples but corrects an important distinction: not every item below was a campaign launched in 2025. Some are 2025 campaigns, some are brand or governance controversies that became marketing stories during the year, and a few are older examples that resurfaced in 2025 discussion. For teams planning high-visibility work, a coherent marketing strategy should connect creative, channel choice, operations, and response planning. Brand Vision treats those pieces as one system.

At a Glance

• American Eagle, e.l.f. Cosmetics, and Coca-Cola are direct 2025 campaign examples that drew public criticism or debate.
• Honey, Southwest, Unilever, and Meta are better understood as 2025 brand, policy, or governance controversies rather than comparable ad campaigns.
• Pepsi's Kendall Jenner ad and P&G's YUMI example predate 2025, while the Nestlé sources document broader industry scrutiny rather than a standalone 2025 campaign launch.

10) PayPal Honey: 2025 Fallout From an Affiliate Attribution Controversy

The Honey controversy broke in late 2024 and continued through 2025. Creators and publishers filed lawsuits alleging that PayPal's Honey browser extension diverted affiliate commissions by inserting itself at the last click. The companies involved disputed or denied key allegations, so this is best described as an attribution and trust controversy rather than a proven "scam." Google later tightened Chrome extension rules around affiliate links and user benefit, giving the wider issue more industry weight (The Verge; Washington Post; Times of India).

• What went wrong: opaque attribution mechanics created a sharp trust conflict with creators who believed they had earned the sale.
• What would have helped: clearer disclosure about when Honey could receive attribution, explicit user benefit at checkout, and independent auditing of affiliate logic.

9) Pepsi: A 2017 Campaign Still Used as a Cautionary Example

Pepsi's Kendall Jenner protest ad was released in 2017, not 2025. It remains relevant because it continues to appear in marketing education and retrospectives as an example of a celebrity-led concept colliding with a sensitive cultural context. The original backlash was strong enough that Pepsi pulled the ad and apologized (Reuters; Forbes).

• What went wrong: the creative simplified protest imagery into a celebrity moment, creating a gap between the cultural subject and the brand's treatment of it.
• What would have helped: stakeholder review beyond the campaign team, including people with direct knowledge of the communities and issues represented.

8) Coca-Cola: Share a Coke and the "Bogan" Can Debate

Coca-Cola brought "Share a Coke" back in 2025 with personalization and QR-led digital features. In Australia, the use of "Bogan" on cans revived a long-running debate over whether the term is affectionate slang or insulting. The controversy was localized rather than evidence that the global campaign failed, but it shows how personalization can create risk when a nickname carries different meanings across communities (Coca-Cola Company; News.com.au).

• What went wrong: a locally loaded term made a playful personalization mechanic feel less controlled.
• What would have helped: market-specific language review, clearer moderation standards, and an easy way to explain why particular terms are allowed or blocked.

7) P&G: A Legacy Virtual-Influencer Example in a 2025 AI Debate

SK-II's digital ambassador YUMI predates 2025, so it should not be presented as a new P&G campaign from this year. The example is still useful because AI moved deeper into advertising. Marketing Week reported in 2025 that P&G was using AI to generate ideas, test advertising, and automate media buying, while Marc Pritchard's earlier warning about preserving "creative humanity" provides useful context for the tradeoff (Marketing Week, 2025; Marketing Week).

• What went wrong: in categories such as beauty, a synthetic spokesperson can create credibility problems when audiences expect human experience and visible product proof.
• What would have helped: use AI behind the scenes or as an assistive tool, while keeping consequential claims and testimonials anchored in real people and demonstrable evidence.

6) Nestlé: Ultra-Processed Food Scrutiny, Not a Documented 2025 Campaign

The sources in the original article do not establish a standalone Nestlé marketing launch that failed in 2025. Reuters reported in December 2024 on a lawsuit accusing several major food companies of marketing allegedly addictive ultra-processed foods to children, while 2025 policy discussions focused on how U.S. agencies might define "ultra-processed" foods (Reuters; FoodFix). That makes this a messaging-risk example, not a verified 2025 campaign failure.

• What went wrong: broad category scrutiny made health and convenience messaging harder to separate from debates about processing, ingredients, and children's marketing.
• What would have helped: use specific, supportable product claims and avoid relying on contested category language that regulators have not clearly defined.

5) Unilever: Sustainability Claims Under 2025 Scrutiny

Unilever faced renewed scrutiny over environmental claims rather than one discrete ad campaign. In 2025, Dutch consumer group Consumentenbond said it reviewed more than 450 products and identified 247 products with sustainability claims it considered misleading. The dispute followed earlier criticism around Unilever's environmental commitments and contributed to a broader question about how much proof should sit behind purpose-led marketing (The Guardian; Consumentenbond).

• What went wrong: broad environmental language was easier to challenge than narrow, measurable claims tied to a specific product or action.
• What would have helped: third-party substantiation, consistent claim standards across brands, and language that states exactly what has been measured.

4) Southwest Airlines: Ending "Bags Fly Free" Created a Brand-Promise Backlash

Southwest's 2025 baggage policy change was not an ad campaign, but it became a marketing problem because "Bags Fly Free" had functioned as a distinctive brand promise for years. Southwest announced that most customers booking or changing qualifying tickets on or after May 28, 2025 would pay for checked bags, with exceptions for certain status, cardholder, and fare groups. News coverage focused on the contrast between the new fees and a message customers had learned to associate with the airline (AZFamily; KKTV; Forbes; Southwest Airlines).

• What went wrong: a pricing change directly contradicted one of the brand's strongest learned benefits, so customers evaluated it as a trust issue as well as a fee.
• What would have helped: more transition runway, explicit value tradeoffs, and loyalty protections communicated before the new policy took effect.

3) Meta: AI Profiles and a Broader Governance Backlash

Meta's AI-profile controversy in early 2025 involved experimental AI-generated Facebook and Instagram profiles that had originally been introduced earlier. Renewed attention followed executive comments about AI-generated users, and Meta removed remaining profiles after criticism about their artificial identities and limitations. This was a product and governance story more than an advertising campaign (The Guardian).

Later 2025 reporting added a separate safety controversy around AI chatbots and interactions involving minors. The Guardian reported on internal policies, while Reuters and AP covered the FTC's broader inquiry into AI companion safety. Those developments belong in the same governance discussion, but they should not be described as "worst ads" (The Guardian; Reuters; AP).

• What went wrong: product experimentation moved into areas where identity, age, safety, and user expectations required stronger controls and clearer communication.
• What would have helped: stricter age safeguards, transparent labeling, adversarial testing, escalation policies, and narrower launch scopes before scale.

View this post on Instagram

A post shared by Jules Terpak (@julesterpak)

2) e.l.f. Cosmetics: Matt Rife Casting Backlash

In August 2025, e.l.f. launched a legal-parody campaign featuring comedian Matt Rife and drag performer Heidi N Closet. The casting quickly drew criticism because viewers resurfaced Rife's 2023 domestic-violence joke and his response to the earlier backlash. e.l.f.'s CMO later acknowledged that the brand had "missed the mark," while explaining that audience data had initially made the pairing appear promising (e.l.f. Beauty; Business of Fashion).

• What went wrong: the spokesperson's public history conflicted with how parts of e.l.f.'s audience understood the brand's inclusive positioning.
• What would have helped: stronger talent-risk review, including older clips and controversies, plus alternate casting plans before production. The same principle applies to influencer and creator partnerships.

1) American Eagle: "Great Jeans" Became a Polarizing Marketing Story

American Eagle's 2025 campaign with Sydney Sweeney used the line "Sydney Sweeney Has Great Jeans," deliberately playing on "jeans" and "genes." The wordplay generated immediate attention and criticism, including arguments that the execution carried exclusionary or eugenics-adjacent connotations. The launch became one of the year's most polarizing advertising stories (Axios; Forbes; Reuters).

The important nuance is that controversy is not the same as commercial failure. The campaign also produced an immediate stock-market reaction and sustained attention, so the "worst campaign" label is not a clean commercial verdict. This is better treated as a case study in narrative risk and brand interpretation than as an ad that simply failed.

• What went wrong: a double meaning left enough room for audiences to attach a much heavier interpretation than the brand intended.
• What would have helped: pretesting the line across different audiences and evaluating not only whether the pun was memorable, but what associations it could trigger at scale. Compare the Gap x KATSEYE and American Eagle denim campaigns for a fuller performance analysis.

The Final Word

The useful lesson from 2025 is not "avoid controversy at all costs." It is to distinguish reputational backlash, governance failure, and commercial underperformance instead of treating them as the same thing. Strong branding sets expectations, and campaign planning should define which risks are acceptable, what evidence will determine success, and how the organization will respond when the audience interprets the work differently than expected.

FAQ

What counts as a marketing fail in this list?

This article uses the term broadly for campaigns, policy changes, creator choices, and technology or governance decisions that created avoidable brand risk in 2025. It now distinguishes those categories rather than implying all ten were comparable ad campaigns.

Which examples were actual 2025 campaigns?

American Eagle and e.l.f. were clear 2025 campaigns that drew substantial backlash. Coca-Cola's revived Share a Coke campaign also created a localized debate in Australia. Other examples in the list are policy, governance, or legacy cases included for their marketing implications.

Why do AI-related brand problems become controversial so quickly?

AI products can scale before identity, safety, disclosure, and moderation systems are mature. When that happens, the public story can move from innovation to governance faster than the brand can respond.

Why did American Eagle's Sydney Sweeney campaign spark backlash?

The "Great Jeans" concept used a genes-and-jeans double meaning that some audiences interpreted as exclusionary or eugenics-adjacent. The campaign also produced significant reach and product demand, so it is more accurate to call it polarizing than simply unsuccessful.

What Business Owners Can Learn From These Marketing Fails in 2026

Most marketing failures do not begin with bad intent. They begin with a gap between the promise, the execution, and what the audience actually experiences. Treat campaign planning as both creative development and risk management.

• Pressure-test the promise before launch. If a campaign touches pricing, safety, identity, health, or a long-standing customer benefit, a clear brand strategy should define what the company can credibly say and what must remain consistent.
• Build guardrails for personalization and AI. Filters, age controls, disclosure, and red-team testing should be designed before scale rather than added after backlash.
• Audit the full funnel, not only the ad. Attribution, fulfillment, moderation, policy, and customer support can create the reputational problem even when the creative itself is sound. A marketing consultation can help map these weak points before launch.
• Measure business outcomes separately from online outrage. Use a consistent set of marketing metrics so reach, sentiment, acquisition, conversion, and retention are not collapsed into one verdict.
• Design for comprehension rather than cleverness. Strong UI and UX design reduces ambiguity across landing pages, forms, and support paths after the campaign earns attention.
• Make sure the destination matches the message. Web design, performance, and mobile clarity determine whether a spike in attention becomes a usable customer journey.

If you are planning a high-visibility campaign, Brand Vision can help pressure-test the strategy, measurement plan, and customer journey before launch. Talk with Brand Vision about your campaign.