What to Do When an Influencer Collaboration Turns Into a PR Crisis

When is it a PR crisis — and when is it “just” a problem?

A PR crisis in an influencer context is different from the three other situations this Academy already covers: brand safety vetting is about preventing problems before you sign, non-delivery is about the agreed content never arriving, and a failed campaign is about numbers that don't perform even though everything was delivered as agreed. A PR crisis happens when something is already publicly visible and generating negative attention around a collaboration that has already gone live — regardless of how well the influencer was vetted beforehand, regardless of whether the content was delivered on time, and regardless of how the campaign performs on the numbers.

The defining feature is that the damage has already happened in public by the time you discover it. You can't prevent your way out of it — only respond to it in a structured way.

Three types of triggers — and why they need different responses

Not every influencer-related crisis has the same origin, and the origin decides how much of the responsibility actually lands on your brand.

Coombs' Situational Crisis Communication Theory: the academic model behind crisis response

Situational Crisis Communication Theory (SCCT), developed by communications researcher W. Timothy Coombs, is an established model for how organisations should match their crisis response to how much responsibility the public actually attributes to them. The theory groups crises into three clusters by attributed responsibility: the victim cluster (the organisation itself is targeted, e.g. rumours or sabotage — mild threat), the accidental cluster (unintentional, e.g. a technical error — moderate threat), and the intentional cluster (the organisation acted knowingly wrong or negligently — severe threat). From there, Coombs groups crisis responses into four categories: deny (reject the connection to the crisis), diminish (minimise severity or responsibility), rebuild (compensation, apology) and bolster (highlight past good works to supplement one of the other three responses).

SCCT was developed for organisations in general, not specifically for influencer partnerships — the application below of the theory's logic to a brand-influencer relationship is Make Influence's own translation of it, not part of the original theory.

The first 24 hours — a checklist

1. Pause planned and paid content immediately — including whitelisted ads and partnership ads running on the influencer's account. See Creator Whitelisting, Spark Ads and Partnership Ads Explained for how to actually pause a running flight.
2. Gather the facts before you respond publicly — what actually happened, when, and how much is confirmed versus rumour. Same discipline as the reactive step in the brand safety checklist, but here the focus is preparing a public statement, not just an internal decision.
3. Designate one internal point of contact who coordinates both the internal decision and the external communication. Two people giving different answers to the same outside question is its own crisis.
4. Decide whether the content should stay up, be hidden, or be removed — see the decision framework below. Don't remove it reflexively before you've assessed the cause.
5. Avoid a rushed public statement before you have the full timeline. A too-fast, imprecise reaction can itself become a new story on top of the original one.

Should the content stay up, be hidden, or be removed? A decision framework

Who speaks — you, the influencer, or both?

Is it the morality clause, or is it communication?

The two tracks run in parallel, but answer different questions. The morality clause, described in the brand safety checklist and as one of the 12 terms in What to Put in an Influencer Contract, decides whether and how you can end the partnership. It doesn't tell you what to say publicly while you make that decision. Conversely, good public communication doesn't automatically resolve whether the partnership should continue. Treat them as two separate decisions that inform each other but don't replace each other.

A hypothetical worked example: the cost of waiting

The example below is invented and for illustration only. It is not a real Make Influence customer.

A brand is running a 14-day whitelisted TikTok Spark Ads flight for DKK 21,000, spread evenly at DKK 1,500 per day. A crisis breaks on day 3. If the brand discovers it and pauses the flight the same day, there's no further exposure of the ad budget against the problematic partnership. If the brand instead waits a week to react while the internal decision process drags on, the flight keeps running for 7 extra days at DKK 1,500 each — DKK 10,500 of the budget spent amplifying the reach of exactly the content the brand should have paused. The maths is simple, but it's often precisely this kind of delayed decision that makes a manageable situation more expensive than it needed to be.

Common mistakes

• Confusing a PR crisis with a brand safety incident that should have been caught in vetting. Some crises happen no matter how thorough the vetting was — that isn't automatically a vetting failure.
• Removing content before you know what actually happened. Removing too fast can read as an admission.
• Letting the influencer and the brand speak independently of each other without coordinating what's said.
• Waiting for the “perfect” statement while the flight keeps running and the content stays visible with no response.
• Treating the morality-clause decision and the public communication as the same decision, when they actually require separate processes.

Make Influence's perspective

At Make Influence, we see crisis response as an extension of the same discipline behind brand safety vetting and the contract's morality clause — not a wholly separate process invented once the damage is already done. Our experience is that the single factor that most often decides how expensive a situation becomes isn't the incident itself, but how quickly the internal decision process — pausing paid distribution, establishing the facts, one spokesperson — gets moving. Brands that have thought through the sequence in advance respond markedly faster than those that first have to work out who has the authority to pause a flight while the crisis is already running.

FAQ

Is it always a good idea to remove the content immediately?

No. See the decision framework above — if you remove it before you know what happened, it can read as an admission of fault you may not actually have. Hiding it temporarily while you establish the facts is often the more defensible first step.

Should the brand always comment publicly?

Not necessarily, and not immediately. A short, factual statement about what the brand is concretely doing about the partnership is often enough — you don't need to comment on the influencer's private affairs or guess at details you can't document yourselves.

What if the crisis has nothing to do with your brief or content?

Then your direct responsibility is low, but your exposure grows the longer you wait to address the partnership's future — see the guilt-by-association row in the table above.

How is this different from just using the morality clause?

The morality clause decides whether and how you can legally end the partnership. It doesn't tell you what to say publicly while you make that decision — these are two parallel, not identical, tracks.

How is this different from a campaign that just failed?

A failed campaign is about numbers — engagement, clicks, sales — not meeting expectations, even though everything was delivered and executed as agreed. See What Should You Do When an Influencer Campaign Fails? for that diagnosis. A PR crisis is about reputation, not performance numbers, and can hit a campaign that otherwise performs perfectly well.

Can an agency or platform help with this?

An agency or platform typically already has an escalation process and can act as an intermediary between you and the influencer while the situation is resolved — the same benefit described in the non-delivery article.