Since the release of GPT-3, generative AI software has quickly become a popular and increasingly powerful tool for tech startups. With the newest AI, you can now potentially accelerate your content creation and streamline the process of producing content at a fraction of the cost of adding headcount.
Generative AI software apps include tools such as natural language processing (NLP), text analysis, summarization, sentiment analysis, and automated writing to help tech marketing content creators kick-start their writing process and improve drafts.
The AI floodgates have opened, and the ways in which copywriters and content writers will be able to use generative AI are growing every day. As of this writing, it feels a bit like the wild west of creative output — so in this article, I’m going to lay a foundation for startups and give you solid footing as the waves of change roll on.
Pros and Cons of Using AI Content Tools
Let’s first take a look at the why behind using AI software for copywriting. Beyond the possible financial upside of having AI create an initial draft instead of hiring a human to do it, what’s the advantage for your startup?
Potential benefits include:
- Boosting the creativity and motivation of your marketing team
- Automating tedious tasks, like initial research, summarizing, or organizing information
- Helping writers and marketers identify trends more quickly
Now that we have a good understanding of the why behind using AI to produce your startup marketing content, let’s get clear on the pros and cons of this technology.
If you ask a bot what the pros are for using AI for copywriting, it’ll tell you that AI improves content accuracy and consistency. I know this because I’ve asked five different AI apps this very question, and they’ve all produced some variation of that “pro.” Based on my actual experience using AI, however, I do not agree. If you’re looking for accurate and consistent content, you need to involve a human in the process.
That said there are pros. For many startups, there can certainly be cost savings. Generative AI tools can also give you more power to scale your content production.
The cons, however, are very real. You need to get clear on them before you pay for that AI software to start creating initial drafts of content for you.
AI is only as good as the data it’s trained on — and the data it’s trained on comes from a massive number of sources … some of which are incredibly biased or problematic. AI output can also be inaccurate or wildly outdated, so you need to make sure you have a human fact-checking everything before you publish it.
AI-generated content is also innately generic, in terms of human standards, and can also be extremely low quality. Again, a human touch is required to make AI-generated content “human ready.”
Use Cases for AI Content at a Startup
So what are the best ways to use AI for copywriting? How do you know when to use AI, and when to pay to have a human write your content?
Some quick definitions before I answer those questions:
Input: The prompts you write to tell the AI what you want it to produce for you.
Output: The text the AI produces in response to your prompt.
Now, let’s look at where generative AI shines for text production:
- Summarizing unstructured data: NLP can extract and consolidate the important points from unstructured data like meeting notes, customer feedback or reviews. Some AI tools can even identify product benefits and objections from customer reviews.
- Identifying themes and topics: AI can identify key themes and topics from the data it was trained on (for GPT-3, that’s 45TB of text data) within the prompt parameters you give it.
- Automated writing: AI uses machine learning algorithms to generate copy in seconds. (Caveat: The quality and accuracy of the output depend wholly on the input prompt and the editing your human copywriter will do.)
Based on my exploration of several popular tools, where I see the most value for startups is in rapidly generating simple first drafts for the following use cases:
- Social media posts
- Headlines and titles
- Short-form content (e.g. blog posts, articles, essays, statements and bios)
- Ad copy
- Press releases
- Product descriptions
- Landing page copy
- App and SMS notification copy
- Meta descriptions
- Video scripts
- Course outlines
- Changing the tone of an existing piece of writing
Let me reiterate, however, that all AI-generated copy and content must be fact-checked and edited. Raw output is not publishing-ready!
You’ll notice that the list above is all short or short-ish content. The most established tools on the market are already doing this well (for example, Jasper and Unbounce Smart Copy). But I’m seeing new solutions emerging for longer-form content as well (Moonbeam, for one). I anticipate more cost-effective software and additional features within existing software solutions for writing long-form will continue to surface in the next year.
Practical Steps for How to Use AI to Produce Content
1. Decide what kind of content you want to create with generative AI.
Do you want something to help you write your LinkedIn posts, an app that will write first drafts of blog posts, a solution that can produce product descriptions for you — or all of the above?
2. Choose an app that can produce that kind of content.
I’d recommend selecting 3-5 apps with free trials, testing them all, and choosing the one that is easiest for your team to use. Most AI writing software is built on the same AI platform right now, so the output won’t differ much from app to app — however, the input process does vary quite a bit.
3. Determine what information you need from your internal subject-matter experts to write effective input prompts.
Whatever AI app you use, you’ll need to give it the right input to get the output you’re looking for. Document your criteria and then ask your subject matter experts for their insight. For example: target audience, topic, tone and/or style, or any specific information to include (like reports or statistics).
4. Create a system for editing your AI content output.
Remember, raw output is not ready for human consumption! You must fact-check and edit your output before you publish it. At minimum, your team will need to fix inaccuracies and mistakes, and edit the draft for voice, structure and flow. However, if you want your content to really stand out and be taken seriously by your customers, you’ll also need your team to add what AI simply can’t: nuance, narrative and soul.
Finally, there is the ethical consideration. Generative AI is so new to the public marketplace, the ethics are still being worked out — but there is a growing consensus that if you use AI in your writing process, you should disclose this fact. I used ChatGPT AI to research this piece, and you’ll see my own disclosure notice at the bottom of the page. Feel free to adapt that for your own use!
Can You Spot the Bot?
Generative AI technology is getting frighteningly good — and it’s not going away. Startup marketing teams should not only prepare for how these tools will change their processes around copywriting and content writing, but they should find new and creative ways to use them to their advantage.
However — and I will keep ringing this bell until the whole internet hears me — you must thoroughly fact-check and edit your AI-generated copy and content. Customer trust is on the line, here. If customers can’t trust your content, they can’t trust your startup.
AI disclosure notice: The author researched, outlined, or generated this text in part with AI. Upon generating draft language, the author reviewed, edited, and revised the language to their own liking and takes ultimate responsibility for the content of this publication.