You can’t have missed the talk about how AI is going to revolutionise service businesses. From app marketing to coding, the launch of tools like OpenAI’s ChatGPT has everyone considering whether our new robot overlords are here to make our lives easier or steal our jobs.
ChatGPT is the big name due to first mover advantage but it isn’t the only AI chatbot on the market. There are the freebies – YouChat is a slimmed down version of ChatGPT, while Google has launched Bard by invitation only, and Microsoft has reimagined Bing as an AI tool. Services like Jaspar and Chatsonic charge a fee, with the promise of a better end result and more up-time (something that ChatGPT struggles with).
Of the free services, you’ll find they all work in a very similar way: providing you with an area to enter your text prompt, followed by a response within that thread. As you can see from the examples below, they also share a similar style of output:
OpenAI’s ChatGPT
Much has been written about ChatGPT et al’s potential to provide content and copy for over-worked marketers, but it’s quickly become apparent that (currently) the tools don’t have the eye for detail – or indeed desire to get their facts straight – that a human does, so there is still a need for the human hand.
However, if you are prepared to play at editor then there are many other ways that the current crop of AI chatbots can lend a helping hand to your marketing efforts.
- Keyword research
Generating keywords for your ASO through an AI chatbot is far more efficient for app marketing than manually researching and brainstorming. Plus, the tool has access to data regarding competition that you simply don’t, unless you subscribe to a keyword tool.
The first, and most obvious, step is to ask ChatGPT to collate a primary list of keywords.
Prompt: Give me a list of keywords for my dating app
But a more sophisticated enquiry is to ask it to generate a list of keywords that have a high volume but low competition – the holy grail of keywords.
Prompt: Give me a list of high volume low competition keywords for my dating app
Once you’ve got a list of short tail keywords, long tail keywords (those with fewer searches, but higher intent) are the obvious progression.
Prompt: Give me a list of long tail keywords for my dating app
Overall, AI chatbot-generated information like this doesn’t replace paid-for tools like AppTweak, or MobileAction who not only have better data but can present it in a more useful way. But it can provide a fast and free starting point, as well as open up some surprising avenues worth further exploration.
- Competitor research
Understanding your competitors is an important part of any app marketing strategy, but it can be difficult to know where to start. Enter: ChatGPT.
Prompt: Give me an overview of the top-performing apps in dating apps on Google Play and App Store. Analyse their app titles, descriptions, keywords, and ratings to identify trends and areas for improvement.
The information gleaned isn’t necessarily hugely insightful, but it took about 30 seconds to pull together, making it an extremely efficient way of beginning competitor research.
- Market research
There are many ways to define your audience, including interviews, A/B testing of marketing collateral and mobile analytics tools. Now you can add AI to that list.
Prompt: Who is the target audience for a dating app. What’s the demographic profile of the most likely users?
Again, this information alone isn’t enough to base your marketing strategy on, but it works as a starting point from which to build on.
- UX research
Currently, ChatGPT can’t access specific reviews on specific product pages, but its ability to provide a general analysis of feature requests and feedback can be invaluable in suggesting ways in which you can improve your app.
Prompt: What are the most requested features in dating app reviews?
- Creative assets?
Chatbots are text-based, which means if you want to move into imagery, you need to find another source. Midjourney and DALL-E are two such services which generate imagery based on a text-based prompt, but you now need to pay for both. They’re controversial, in that they learn through scraping the web for other artists’ work, creating a myriad of plagiarism issues.
In short: don’t get rid of your creative team any time soon.
Conclusion
You may already have figured out the strengths and weaknesses of ChatGPT and its ilk when it comes to marketing: they can provide you with fairly general information, but they don’t have the deep knowledge that you as an app marketer possess about your particular app.
Getting an answer to a query feels like finding the right blog post to answer your question, but not necessarily like it was written for you. There’s certainly an element of timesaving in getting this information quickly, but it’s not going to do your job for you.
Plus, they don’t get it right every time. You might have spotted the obvious error regarding high volume and low competition keywords: best online dating app doesn’t sound like it would be a rarely used search term. In fact, you might think it’s something most dating apps would try to rank for.
And you’d be right.
Which leads to the only possible conclusion: AI chatbots are in their infancy, and while they do a good job of looking like they’re giving you correct information, that’s not always the case. At this stage of sophistication, they’re best placed doing the time consuming and repetitive work on your To-Do list, before you pass it over to a human to examine with a fine toothed comb.
