Brand personality in the age of AI homogeneity

Brand voice in 2026 has converged. Almost every B2B SaaS site reads the same. Same hero structure (“Stop wasting time on X — start doing Y”), same testimonial cadence, same blog tone (helpful, slightly enthusiastic, never opinionated, never funny). The reason is structural: everyone’s using the same LLMs to write the same copy, sometimes literally word-for-word.

That homogeneity is the opportunity. Brands with actual personality stand out more in 2026 than at any point in the last decade. The bar to be memorable dropped because most competitors aren’t trying.

What “personality” means concretely

Personality in brand voice isn’t about being funny or quirky. It’s about having specific, repeatable patterns that come from a real human point of view.

Three components:

1. Specific point of view. Brands without personality have positions like “we believe in great customer service.” Brands with personality have positions like “we will not work with crypto companies regardless of the budget.” The second is specific, falsifiable, and signals what you actually think.

2. Repeatable voice patterns. Word choices, sentence rhythms, structural moves that show up in every piece of content the brand publishes. Linear’s blog has a specific dryness. Stripe’s writing has a specific patient explanatory quality. Pieter Levels’ tweets have a specific bluntness. These aren’t accidents — they’re patterns.

3. Named operators with consistent personas. The brand voice IS a specific human’s voice (or a small set of them) writing under their name. Anonymous corporate voice never has personality. It can’t, structurally.

What it looks like in practice

Brands with strong personality in 2026, with one specific pattern each:

Linear: Quietly confident, slightly bored with industry hype. Phrases like “we are not interested in [feature]” appear in product documentation. Refusal as personality.

Anthropic’s product blog: Patient, exhaustive, willing to admit uncertainty. The pattern is “here’s what we tested, here’s what we found, here’s what we don’t yet know.”

Plausible Analytics: Anti-Google Analytics. Specific opposition. Their blog publishes “why we don’t do X” posts as a recurring format.

Lenny Rachitsky: Specific operator-friend voice. “I’ve been doing this for 15 years and here’s what surprised me about X.” Always personal, always specific.

A24 (film studio): Aesthetic over explanation. They almost never explain their movies in marketing copy. The absence of explanation IS the personality.

What none of these share: a generic “About Us” page describing their mission to empower customers.

Why most brands lose personality

Three common patterns we see:

Committee writing. Voice that has to pass through legal, PR, and three rounds of stakeholder review gets sanded smooth. The unique edges that gave it personality are the first things cut.

AI without a strong human editor. Using ChatGPT to write all copy without a strong, opinionated editor passing through it produces median-of-the-internet voice. Average. Forgettable.

Wanting to be liked by everyone. Personality is divisive by definition. A brand voice that takes specific positions WILL alienate some readers. Brands that try to be universally inoffensive end up not being remembered.

The reason a strong brand voice in 2026 is more valuable than ever is that the floor for boring writing is now infinitely low. Anyone can produce indistinguishable median content with AI in five minutes. The only differentiator is specific, opinionated, human voice that takes risks AI defaults won’t take.

How to develop a real voice

The exercise we run with clients during the Brand Sprint:

Step 1: write down what you’d refuse. What kinds of clients, projects, content topics, design choices, partnerships would you turn down? The refusals are the personality.

Step 2: pick your enemies. What’s the conventional wisdom in your industry that you think is wrong? Brand voice gets sharp by having clear targets to push against. Plausible’s enemy is Google Analytics. Hey.com’s enemy is Gmail. Pick one.

Step 3: name the operator. Voice can’t be “we” forever. There needs to be a named human with a real LinkedIn whose voice we hear. For most small brands, this is the founder.

Step 4: write the voice doc. 1-2 pages, includes: banned phrases, required patterns, examples of “yes” voice and “no” voice. We share an example template in our Brand Sprint deliverables.

Step 5: enforce it ruthlessly. Every piece of public content runs through the voice doc before publish. Most teams get tired of this within 3 weeks. The ones that don’t have brand voices that compound.

Banned phrases we use

Our voice doc bans these specifically (and we apply it to every piece of content we ship):

• “Leverage” (as a verb)
• “In today’s fast-paced world”
• “It’s important to note”
• “When it comes to”
• “Whether you’re…”
• “Game-changer”
• “Supercharge”
• “Harness the power of”
• “Unlock the full potential”
• “Synergy”, “alignment”, “stakeholders” outside literal contexts
• “Cutting-edge”, “best-in-class”, “industry-leading”

If you ctrl-F any of these on this site, you’ll find zero matches. That’s the discipline.

What we do for clients

The Brand & Positioning Sprint we run ($8K, 2 weeks) is largely this work. Voice doc, refusals list, enemy identification, named-operator clarification, design choices aligned to voice.

If you want help finding your voice — or sharpening it after years of committee writing — tell us what you’re working on.

Related reading:

• How to actually build trust in your brand — adjacent: trust through specificity
• Brand-driven web design — adjacent: voice meets visual