Not having the “perfect” brand voice shouldn’t stop you from creating content — brand voices change and evolve over time, whether naturally or as the result of a strategic planning session. But it’s important to be consistent (ish) with your brand voice when creating SEO content. If your marketing style changes, be consistent with that new brand tone, and so forth.
A consistent brand voice:
- Builds trust and brand recognition with your target audience
- Makes people feel like they ‘know’ you or your brand
- Shows off your unique brand personality
- Creates consistent brand messaging across different social media platforms, email, and your website
- Supports a strong visual identity for brands and content creators who use visual elements like graphics and videos
You may not realize it right away, but when you offload a lot of your writing tasks to an AI tool — and I’m specifically talking about content that is AI generated and published to your website or other channels — your brand voice gets diluted.
Why AI dilutes your brand identity and voice
“But wait!” You might be saying. “I’ve trained ChatGPT on my voice!”
While it’s true that adding brand messaging guidelines, tone of voice rules, and even style guides into an AI tool will improve the outcome, the content that you receive is ultimately made up of predictions. It’s strings of words that are statistically most likely to appear next to each other in a sentence if said by a human.
What human, you ask? Why, the generic average human that we’d get if we cobbled together both the finest literature and the most degenerate Reddit posts into one. Because that’s what makes up AI training data sets.
There is SO MUCH MORE of that training content in the AI than whatever you uploaded in terms of a brand voice guide or marketing strategy framework.
You would need 5,125 Kindle Paperwhite e-readers if you wanted to collect and store all of the books that Meta fed into its large language model training data. 5,125!
Ultimately, no matter how well you structure your prompt or how thorough your brand guide is, you run the risk of producing brand content that sounds less like you and more like an approximation of you. (And if you’re part of the 42% of people who don’t edit and tweak AI outputs before publishing on your website, you’re losing an opportunity to make the AI content better match your brand.)
AI dilution is a problem with content creation at all levels
The problem of AI diluting brand voice in marketing materials isn’t just a small business or solopreneur problem. There are some big organizations that have lost the impact of a strong brand voice due to overuse of AI.
I wrote about one such situation back in 2024 when I noticed that the Professional Women’s Hockey League (PWHL) used generic, AI-generated descriptions to announce each new team name. These team names were something league fans had been waiting over a year for, as the league’s first season just used the names of cities. (“Boston,” “Vancouver,” etc.)
They lost possession of the puck hard with that one, if you will.
But negative reactions to AI-generated brand assets have been seen across the board, from people wondering why their favorite influencer sounds “off” to entire families getting the ick from AI-produced Coca-Cola ads over the holidays.
What’s the solution?
If you’re dead-set on using AI as part of your content generation process, then at a minimum, you must have:
- Clear guidelines about your brand’s values, messaging, and experience
- Further brand guidelines about how your content should be structured and sound on different platforms (i.e. TikTok scripts vs. blog posts)
- A robust style guide including on-brand punctuation guidelines
- Information about your customer personas, if you use those in your marketing strategy
Then, when you do generate content with AI, you’ll want to:
- Use a dedicated GPT or workspace for different kinds of content
- Do a complete edit of the entire draft produced by AI
To be truthful, though, you’re still going to struggle to get a strong brand voice out of this process. You’d be better served by writing content yourself (or hiring a writer), then using AI to give a quick check to assess style guide adherence, if anything.
I know that the idea of drafting content from scratch can feel daunting and slow, but honestly, you might not be saving that much time by using AI. (My recent study found that almost half of all AI users work at the same speed or slower when using AI vs. without.)
The process of writing content on your own can actually be faster than using AI if you:
- Have the right keywords and topics selected for your audience
- Know what kind of content to write and how to structure it
Want to make this process even easier? Get the system that I and other writers have used to create 1500 words of great SEO content in under two hours.