I put chatGPT to the test proofing our marketing content and here's what I found

Over the past month, I've embarked on an experiment to gauge the effectiveness of ChatGPT in proofing various types of our own marketing content. From ad copy and emails to landing pages and blog posts, I've rigorously tested this AI tool. The outcome? A blend of incredibly useful insights and some cautionary notes. Let's get into the details.

The Good

Spelling and Grammar

The accuracy of ChatGPT in spelling and grammar is nothing short of impressive. In a digital landscape where even a minor typo can tarnish a brand's image, this is invaluable.

Examples:  

• ‍‍ In one of our email campaigns, ChatGPT caught a glaring error where "customer" was mistakenly written as "costumer." A small mistake, but one that could have confused our audience.
• In a blog post about our latest product, the AI corrected "world's" to "worlds," aligning the term with the intended plural context.

Style Consistency

ChatGPT's ability to adapt to different writing styles is a significant asset, especially when consistency is key across various types of content.

Examples:  

• ‍‍For a white paper that required a formal tone, ChatGPT suggested replacing casual phrases like "you'll find out" with more formal alternatives such as "you will discover."
• In a social media post, the AI maintained a conversational style by suggesting we change "it is" to the contraction "it's," making the post feel more relaxed and engaging.

Time-Saving

Summary:  
The efficiency of ChatGPT is a game-changer for a fast-paced marketing company like ours.

Examples:

• Proofing a 2,000-word blog post took the AI less than 5 minutes (including corrections!), a task that would typically take one of our human editor at least an hour.
• For a series of ad copies, ChatGPT completed the proofing in just 2 minutes, allowing us to move on to other important tasks more quickly.

Brand Compliance

One of the most underrated yet powerful features of ChatGPT is its ability to adhere to brand compliance standards. If you've spent time and resources crafting a brand voice and style guide, you can train the AI to follow these guidelines meticulously. We found it helpful in a few different ways...

Examples:

• Tone and Voice: After loading our brand's style guide into the system, ChatGPT was able to maintain our specific tone of voice across different marketing materials. For instance, if your brand voice is "confident and professional," the AI will steer clear of casual or overly technical language.
• Terminology: We have specific terms and phrases that are integral to our brand. ChatGPT, once trained, consistently used these terms correctly. For example, if we refer to our customers as "partners," the AI will not mistakenly change it to "clients" or "customers."

By integrating your brand standards into ChatGPT, you can ensure that all your marketing content not only is free of errors but also aligns perfectly with your brand's voice and message. This is a significant advantage for maintaining brand integrity across multiple channels and types of content.

The Bad

Context Understanding

While ChatGPT excels in many areas, understanding context is not its strongest suit.

• In an email discussing our "cutting-edge technology," ChatGPT suggested changing it to "leading-edge technology." While not incorrect, the change altered the nuance we wanted to convey.
• In a landing page about cloud computing, the AI recommended changing "serverless architecture" to "server-free architecture," which, although technically correct, could mislead those familiar with the industry terminology.

Lack of Industry Jargon

ChatGPT sometimes falls short when it comes to industry-specific jargon, which can lead to unnecessary corrections or even misunderstandings.

• In a piece about "API integrations," ChatGPT suggested changing it to "application programming interface integrations," which, while accurate, is not how the term is commonly used in our industry.
• The AI also recommended changing "SEO" to "Search Engine Optimization," not recognizing that the acronym is widely understood in the tech marketing sector.

Confidentiality Concerns

While ChatGPT offers a plethora of advantages, one area that raises eyebrows is confidentiality. The question that often comes up is: "If I load my proprietary content into ChatGPT, will the AI learn from it and potentially use it elsewhere?"

Examples:

• Intellectual Property: If you're working on a groundbreaking marketing strategy or a unique product description, the last thing you want is for those ideas to be replicated. The concern here is whether the AI retains this information and uses it in future interactions with other users.
• Sensitive Information: For businesses that deal with confidential client data or sensitive company information, the stakes are even higher. The fear is that this data could be stored or, worse, leaked.

It's crucial to understand the data policies of the AI provider and how they handle user-generated content. While OpenAI, the company behind ChatGPT, states that data sent via the API is not used to improve the model, the concern about confidentiality remains a point of discussion in the industry.

Helpful Hints

During our month-long testing phase, we discovered that while ChatGPT is a powerful tool for proofing, its effectiveness can be significantly enhanced with the right approach. We experimented with various prompts, content types, and review processes to understand how to get the best out of this AI tool for proofing tasks. Based on our findings, here are some hints to optimize your experience:

1. Use Specific Prompts: To get the most accurate results, use specific prompts tailored to the type of content you're proofing. For example, if you're proofing an email, use a prompt like "Proofread this email for spelling, grammar, and style consistency."

2. Test Multiple Iterations: Don't settle for the first suggestion. Run the content through multiple iterations to get a variety of options, then choose the one that fits best with your brand voice and message.

3. Manual Review is Still Necessary: While ChatGPT is a powerful tool, it's not a complete replacement for human proofreading. Always conduct a final manual review to catch any nuances or context that the AI might have missed.

4. Train the AI on Your Brand Guidelines: If you have a specific style guide or brand voice, take the time to train ChatGPT on these guidelines. This will help the AI produce content that is not only error-free but also aligns with your brand's identity.

5. Be Cautious with Sensitive Information: Given the concerns around confidentiality, it's advisable to avoid using ChatGPT for content that contains sensitive or proprietary information unless you're certain about the data policies of the AI provider.

By following these hints, we found that we could maximize the benefits of using ChatGPT for proofing tasks while minimizing the risks and limitations of this useful tool.

What's our take on all this?

It goes without saying, ChatGPT offers a range of benefits for proofing marketing content, but it's not without its limitations. By understanding these limits and using the tool wisely, you can save a ton of time without compromising on quality. The key is to use ChatGPT as a supplement to human expertise, not as a replacement. With this balanced approach, the sky's the limit for what you can achieve in your marketing efforts.

If you want to learn more about our stance on AI at Mighty & True, read our AI policy. Also, if you like what you are hearing and want to accelerate your tech marketing programs, let's talk!