How to Properly Remove a Brand Character from Use
Signs That a Character Really Needs to Be Removed
The mistake of most brands is making decisions by eye, based on the team's fatigue from 'that ridiculous rabbit.' The decision should be based on data, not on the boredom of the CMO. Here are three metrics that actually signal the problem:
- Decline in spontaneous recall. That's how Gattu left Asian Paints in 2002 year — research showed that the character was remembered less often than competitors Apex and Ultima. This was a strategic decision, confirmed by data, not a whim.
- The gap between the character's audience and the brand's current target audience. If a character was created for the 40+ audience, but the brand is reorienting toward Gen Z, the conflict is inevitable — but this is a reason for evolution, not necessarily for execution.
- Direct contradiction to the new positioning. Premium repositioning with a comedic mascot on the packaging — this is a dissonance that is perceived by the consumer at a subconscious level faster than marketers can explain it.
If at least one of these three factors is confirmed by research, not feeling — it is possible to move forward. If there is no data, the decision is almost certainly political, not strategic.
Removal, phasing out, or 'sabatika': three different scenarios
Complete removal of the character from production — the riskiest step, and one that often results in reputational damage. There are softer alternatives, and the choice depends on how deeply rooted the character is in the audience's perception.
| Approach | What it means | When it works | Risks |
|---|---|---|---|
| Complete removal | The character disappears without return, usually as part of a rebranding | The character is not a visual anchor for vulnerable audience groups, low recall | Cracker Barrel: -10% traffic; possible wave of nostalgic negativity |
| Phasing (evolution 7–15%) | The character changes gradually, step by step, without sharp jumps | The character is recognizable but looks outdated — like Michelin Man or American Express | Minimal if the step 7–15% is observed; attempting to change more triggers the effect of the ‘uncanny valley’ |
| Sabbatical (‘death and resurrection’) | Temporary disappearance of the character with a pre-planned return | A surge of attention and engagement is needed; the character is strong emotionally | Requires precise timing — Mr. Peanut fell victim to the tragedy with Kobe Bryant and received the opposite effect |
Why sabbatical works better than execution
Duolingo ‘killed’ Duo for two weeks, then ‘resurrected’ it — this created a narrative that the audience discussed on its own, without additional media investment. Mr. Clean used a similar move: the announcement of ‘resignation’ garnered 2 million views in 48 hours and 66 811 search queries, then the brand brought back the character ‘with new innovations’, achieving the effect of resetting attention. The difference from Mr. Peanut’s failure: Duolingo and Mr. Clean had the return scenario written in advance, before publishing the first post.
Roadmap for character exit: where to start and how long it will take
- Month 1–2: tracking metrics. Measuring recall, Buzz Score, and associations with the character in audience segments. Without this step, the decision to exit remains a hypothesis.
- Month 2–3: risk segmentation. Check separately whether the character is a visual anchor for vulnerable groups — parents of children with autism, elderly audience, local communities. This is what Griffin's missed.
- Month 3–4: scenario selection. Full rollout, phasing on 7–15% or Sabbath — depending on the strength of the audience's emotional attachment.
- Month 4–5: test on a focus group and in a controlled media channel. Small market or social network — up to federal announcement.
- Month 5–6: communication with a retreat plan. Predefined scenario for returning in case of negative reaction — as Duolingo and Mr. Clean did.
- Month 6+: monitoring Buzz Score and sales for 90 days after launch. M&M's recorded a decline in the indicator from 19 to 10 among the conservative audience and were forced to urgently roll back the changes — early detection saved the brand from greater losses.
What the mistake costs: the price of a wrong output
| Brand | What went wrong | Consequence |
|---|---|---|
| Cracker Barrel | Attempting to remove Uncle Herschel without preparing the audience | Traffic drop by 10% — the sharpest since the rebranding |
| Griffin's | Removing the 57-year-old Cookie Bear without considering the audience with autism | Petition on Change.org, character's return through 10 months |
| Donki | Changing the mascot without testing on loyal audience | Risk of losing 93% of fan support |
| M&M's | Changing the character lineup without accounting for the conservative segment | Buzz Score drop from 19 to 10, urgent strategy rollback |
| Mr. Peanut | The character's 'death' during Super Bowl 2020 coincided with the passing of Kobe Bryant | Negative emotional tone instead of planned virality |
CMO policy or real strategy — how to tell
A new CMO almost always wants 'their' character — this is normal psychology, but a poor basis for a business decision. Distinguishing a whim from a strategy is simple: ask what data the proposal is based on. If the answer is 'he's just outdated' without recall figures, segmental analysis, and a link to business metrics, you're dealing with a political decision. If there's a decline in recognition, documented conflict with the new positioning, and an assessment of the risk to vulnerable audience groups — the decision is strategic, and it should be pursued to the end, but through phasing, not through a sudden execution of the character.
What to do if the output has already failed
The worst option is to stay silent and wait for the audience to forget. The best is to acknowledge the character's return as part of a planned scenario, even if there was no initial plan. Griffin's brought back the bear after a petition and explained this as 'listening to our customers' — it worked because the decision was swift. Mr. Clean and Duolingo show a more elegant path: turning temporary disappearance into an opportunity for renewed attention, rather than a last-minute apology. The main rule for recovery is the speed of response and honest explanation of the reason, without trying to pass off a rollback as the original intention, especially if there was a long gap between the failure and its acknowledgment.
Signs That a Character Really Needs to Be Removed
The mistake of most brands is making decisions by eye, based on the team's fatigue from 'that ridiculous rabbit.' The decision should be based on data, not on the boredom of the CMO. Here are three metrics that actually signal the problem:
- Decline in spontaneous recall. That's how Gattu left Asian Paints in 2002 year — research showed that the character was remembered less often than competitors Apex and Ultima. This was a strategic decision, confirmed by data, not a whim.
- The gap between the character's audience and the brand's current target audience. If a character was created for the 40+ audience, but the brand is reorienting toward Gen Z, the conflict is inevitable — but this is a reason for evolution, not necessarily for execution.
- Direct contradiction to the new positioning. Premium repositioning with a comedic mascot on the packaging — this is a dissonance that is perceived by the consumer at a subconscious level faster than marketers can explain it.
If at least one of these three factors is confirmed by research, not feeling — it is possible to move forward. If there is no data, the decision is almost certainly political, not strategic.
Removal, phasing out, or 'sabatika': three different scenarios
Complete removal of the character from production — the riskiest step, and one that often results in reputational damage. There are softer alternatives, and the choice depends on how deeply rooted the character is in the audience's perception.
| Approach | What it means | When it works | Risks |
|---|---|---|---|
| Complete removal | The character disappears without return, usually as part of a rebranding | The character is not a visual anchor for vulnerable audience groups, low recall | Cracker Barrel: -10% traffic; possible wave of nostalgic negativity |
| Phasing (evolution 7–15%) | The character changes gradually, step by step, without sharp jumps | The character is recognizable but looks outdated — like Michelin Man or American Express | Minimal if the step 7–15% is observed; attempting to change more triggers the effect of the ‘uncanny valley’ |
| Sabbatical (‘death and resurrection’) | Temporary disappearance of the character with a pre-planned return | A surge of attention and engagement is needed; the character is strong emotionally | Requires precise timing — Mr. Peanut fell victim to the tragedy with Kobe Bryant and received the opposite effect |
Why sabbatical works better than execution
Duolingo ‘killed’ Duo for two weeks, then ‘resurrected’ it — this created a narrative that the audience discussed on its own, without additional media investment. Mr. Clean used a similar move: the announcement of ‘resignation’ garnered 2 million views in 48 hours and 66 811 search queries, then the brand brought back the character ‘with new innovations’, achieving the effect of resetting attention. The difference from Mr. Peanut’s failure: Duolingo and Mr. Clean had the return scenario written in advance, before publishing the first post.
Roadmap for character exit: where to start and how long it will take
- Month 1–2: tracking metrics. Measuring recall, Buzz Score, and associations with the character in audience segments. Without this step, the decision to exit remains a hypothesis.
- Month 2–3: risk segmentation. Check separately whether the character is a visual anchor for vulnerable groups — parents of children with autism, elderly audience, local communities. This is what Griffin's missed.
- Month 3–4: scenario selection. Full rollout, phasing on 7–15% or Sabbath — depending on the strength of the audience's emotional attachment.
- Month 4–5: test on a focus group and in a controlled media channel. Small market or social network — up to federal announcement.
- Month 5–6: communication with a retreat plan. Predefined scenario for returning in case of negative reaction — as Duolingo and Mr. Clean did.
- Month 6+: monitoring Buzz Score and sales for 90 days after launch. M&M's recorded a decline in the indicator from 19 to 10 among the conservative audience and were forced to urgently roll back the changes — early detection saved the brand from greater losses.
What the mistake costs: the price of a wrong output
| Brand | What went wrong | Consequence |
|---|---|---|
| Cracker Barrel | Attempting to remove Uncle Herschel without preparing the audience | Traffic drop by 10% — the sharpest since the rebranding |
| Griffin's | Removing the 57-year-old Cookie Bear without considering the audience with autism | Petition on Change.org, character's return through 10 months |
| Donki | Changing the mascot without testing on loyal audience | Risk of losing 93% of fan support |
| M&M's | Changing the character lineup without accounting for the conservative segment | Buzz Score drop from 19 to 10, urgent strategy rollback |
| Mr. Peanut | The character's 'death' during Super Bowl 2020 coincided with the passing of Kobe Bryant | Negative emotional tone instead of planned virality |
CMO policy or real strategy — how to tell
A new CMO almost always wants 'their' character — this is normal psychology, but a poor basis for a business decision. Distinguishing a whim from a strategy is simple: ask what data the proposal is based on. If the answer is 'he's just outdated' without recall figures, segmental analysis, and a link to business metrics, you're dealing with a political decision. If there's a decline in recognition, documented conflict with the new positioning, and an assessment of the risk to vulnerable audience groups — the decision is strategic, and it should be pursued to the end, but through phasing, not through a sudden execution of the character.
What to do if the output has already failed
The worst option is to stay silent and wait for the audience to forget. The best is to acknowledge the character's return as part of a planned scenario, even if there was no initial plan. Griffin's brought back the bear after a petition and explained this as 'listening to our customers' — it worked because the decision was swift. Mr. Clean and Duolingo show a more elegant path: turning temporary disappearance into an opportunity for renewed attention, rather than a last-minute apology. The main rule for recovery is the speed of response and honest explanation of the reason, without trying to pass off a rollback as the original intention, especially if there was a long gap between the failure and its acknowledgment.
Recent articles
MegaFon, BMW, VTB and more 240+ companies. See how it looks in the showcase reel:
Our branding agency is approached to order:
Branding agency services
