What Starbucks Korea Got Wrong (And Why It Keeps Happening)

By now you've probably heard about Starbucks Korea's "Tank Day" promotion. On May 18, 2026 — the anniversary of South Korea's Gwangju Uprising, in which hundreds of pro-democracy protesters were killed during a military crackdown involving troops and tanks — Starbucks ran a campaign promoting a large tumbler it had nicknamed "the tank." The slogan and framing were widely criticized as evoking state violence and the death of a student activist. The company's CEO was removed within hours. The chairman of Shinsegae Group, which operates Starbucks in Korea, later appeared on live television to bow and apologize to the families of democracy activists.

It is a spectacular failure. It is also, if you've spent any time working in communications or watching how organizations engage with the communities they claim to serve, a deeply familiar one.

The pattern has a long history.

In 2015, Starbucks CEO Howard Schultz launched the "Race Together" initiative, encouraging baristas to write the phrase on cups in order to spark conversations about race in the United States. Schultz had consulted employees, 40 percent of whom were people of color. The intention was widely understood to be genuine. The execution was widely criticized as superficial, shifting the burden of discussing structural racism onto underpaid service workers during routine transactions.

Sony's 2006 PSP advertising featured imagery critics described as evoking racial hierarchy. Intel's 2007 campaign showed Black men in submissive athletic poses surrounding a white executive under the tagline "maximize the power of your employees." The City of Toronto released a community guide criticized for superficial and poorly executed representation. The "Asian Republican Coalition" launched without meaningful Asian leadership, relying instead on loosely defined affiliations.

The pattern is not ideological. It crosses corporate, governmental, and political contexts. The Biden campaign's "Latinos con Biden-Harris" launch in Miami featured non-Latino surrogates as its public-facing representation — a misstep that later analysis within the Democratic Party cited as emblematic of a broader failure in Latino outreach strategy. Neither major party, nor the corporate world more broadly, is exempt.

There are four recurring reasons this kind of failure happens — often reinforcing one another, and frequently in this order.

Structural absence. The root condition. The people who could have prevented the failure were not meaningfully present in the process — not because they were excluded from a meeting, but because the organization was never in sustained relationship with the communities it was attempting to represent. Consultation cannot substitute for connection. And connection cannot be built only when it becomes necessary. When this structure is missing, every downstream safeguard weakens.

Ignorance. Without real relationship structures, there is no reliable feedback loop. People genuinely do not perceive problems that are obvious to those with lived experience, because no one with that experience is close enough to shape the work before it ships. This is often more understandable than the failures that follow, but no less consequential.

Laziness. Once ignorance becomes normalized, shortcuts follow. Leadership does not open the attachment. The right stakeholders are not consulted. A calendar mismatch goes unchecked. These failures resemble process problems, and organizations tend to treat them as such — add a checklist, assign accountability, move on — because process fixes are easier than structural reform.

Not caring enough to check. This is the hardest failure to name, and often the most visible to affected communities. It takes two forms. The first is active disregard: a decision, explicit or implicit, not to prioritize the question. The second is structural indifference: the issue never needed to be prioritized because no feedback pressure existed to make its absence felt. In both cases, the result is the same. The communities affected are not present enough in the organization's decision-making environment for their absence to register as a risk. The failure is not a gap in knowledge. It is a gap in regard.

The downstream responses (executive firings, public apologies, televised gestures of contrition, sensitivity training) often appear only after reputational or financial consequences materialize. This pattern is difficult to ignore, because it suggests that institutional responsiveness is frequently tied more to shareholder pressure than to the communities harmed. If genuine regard had existed earlier in the process, the failure would not have occurred in the first place.

Early in my communications career, I was asked more than once to make marketing materials "look more diverse" — meaning, to add more faces of color to existing imagery. Not to change who was involved in creating the work, not to build relationships with the communities being represented. Simply to adjust the visual output so it appeared more inclusive.

I've also watched organizations defend campaigns after concerns were raised by affected communities, not because those concerns were addressed, but because the work had already been approved. The budget had been spent. The timeline was fixed. The objection was noted and set aside.

This is tokenization — not as an isolated mistake, but as a predictable outcome of treating representation as aesthetic rather than structural. Diversity becomes a layer applied after the fact, rather than a condition built into the work itself.

The USPTO is a useful institutional parallel. For years, it made determinations about potentially offensive terms and names without meaningful consultation with the communities affected — including Asian American communities. That is structural absence at the institutional level: decisions made about a group without sustained relationship with that group. When objections were raised, they were sometimes acknowledged but ultimately dismissed as "laudable but not persuasive." This reflects the regard gap. Feedback existed, but it did not meaningfully influence outcomes. The logic is consistent with what we see in marketing: authority is centralized, consultation is procedural, and affected communities are positioned as inputs rather than participants.

Real relationships — in organizing, in communications, in institutions of any kind — are built before they are needed. They are not assembled under pressure. They involve shared risk, mutual accountability, and the expectation that mistakes will be addressed within an ongoing relationship rather than managed as isolated incidents. They are also the only conditions under which feedback becomes reliable rather than performative.

The musician metaphor fits cleanly here. Structural absence is having no way to hear the room at all. Ignorance is misreading the audience you do have. Laziness is repeating the same set regardless of context. And disregard — active or structural — is continuing the same performance after feedback has already been given and set aside.

Institutions do this constantly. They launch campaigns, roll out initiatives, and declare engagement. But engagement without relationship is just projection. The audience can usually tell the difference.

I speak on civil liberties, identity, and what authentic engagement with communities actually requires at universities, civic organizations, and corporate events. If these questions connect to work you're doing, simontam.org/speaking is the place to start. For civic engagement programming specifically, simontam.org/civic-engagement-democracy has more.