ASO — 22 Feb 2023
Mobile Marketing Meets AI: New Possibilities with ChatGPT
Anastasiya Starovoytova 
Thinking big is thinking disruptive. And there is nothing more disruptive than AI right now — even in mobile marketing.
Artificial Intelligence (AI) is revolutionizing the way businesses approach marketing and content creation, and mobile marketing is no exception. AI tools, such as ChatGPT, are set to improve app marketers’ daily efforts by automating tedious tasks, generating content ideas, and allowing for more personalized user experiences.
ChatGPT isn’t the only AI technology that can inspire you to knock out tasks in one go. There are other tools that will disrupt your thoughts — in a good way.
In this blog post, we’ll explore how artificial intelligence is helping app marketers reach their target audience and whether they’re worth their salt.
P.S. In case you wonder — the blog cover was created with the help of Midjourney.
Before we dive into the practical aspects of using AI-powered solutions as app marketing tools, let’s meet the hero of the hour.
ChatGPT (Chat Generative Pre-trained Transformer) is a natural language processing model that can generate human-like responses to text prompts. It uses deep learning algorithms to understand the context of a given prompt and come up with a suitable response. The tone of the reply is typically conversational and the answers are easy to digest.
No wonder that the whole interaction with the chatbot may make you feel like talking to someone you know. Or even having your personal J.A.R.V.I.S.
Just like its predecessor, GPT-3, ChatGPT was developed by OpenAI, an artificial intelligence research laboratory from the United States. In fact, our friendly chatbot was fine-tuned from a model belonging to the GPT-3.5 series.
Since its launch in November 2022, ChatGPT has been everyone’s darling. All kinds of folks — from business analytics to digital marketers — have been trying it out to generate content and pull out structured data.
Of course, mobile marketing pros threw their hat into the ring too.
ChatGPT wouldn’t have gone viral if it wasn’t useful.
If you’re just curious about ChatGPT and have no intention to use it, think again. This AI tool has a lot of convincing advantages. Here are some of them:
Sounds great, right?
ChatGPT, hands down, is impressive. However, just like everything else in the world, it isn’t perfect.
Here’s how.
While the reactions to ChatGPT’s creativity are mostly positive, marketing teams should take into consideration the following limitations of the AI tool:
No matter how brilliant a solution is, it still has certain drawbacks in its core. ChatGPT is no exception. The best course of action here is to edit and fact-check all the content generated by the chatbot.
Time to see how you can leverage ChatGPT and other AI marketing tools for mobile marketing.
The daily grind of a mobile marketer can be daunting.
You have to find a great idea for a new marketing campaign. Your target audience gets tired of old creatives, and you have to summon all your imagination to make up something new and relevant. The competition is getting bigger with new apps appearing like mushrooms after rain.
Now let’s add budget restrictions and privacy-related changes to the list, and you’ll see how easy it is to feel overwhelmed.
Mobile marketing comes with many unique challenges, and these challenges are getting even tougher when you have a small team or run several apps at the same time. Running out of steam is something many app marketers can relate to.
AI tools, however, have no such problem.
Unlike humans, ChatGPT never gets tired and can generate unique and engaging content to attract and retain users at an extremely quick pace.
This can include anything, from push notifications and responses to reviews to app descriptions and blog posts in case you’re managing a blog about your app. By using ChatGPT, app marketers can save time and resources while still creating content that resonates with their target audience.
When it comes to AI generated content and mobile marketing, ASO is the most obvious choice to try out the magic of deep learning. However, AI marketing tools can also streamline paid UA activities — like creating ad copy for custom product pages in Apple Search Ads.
During our recent App Growth Panel: Mobile UA & ASO Trends to Watch for in 2023, we dwelled on the impact AI and ChatGPT will have on mobile marketing. In the next section of this blog post, we’ll relieve these insights and see how you can leverage ChatGPT for your mobile marketing efforts and app’s content strategy.
App Store Optimization (ASO) is the process of optimizing mobile apps to rank higher in an app store’s search results. ChatGPT and other AI tools can be utilized to improve ASO in several ways, from keyword optimization to content generation, including the creatives on the product page.
AI copywriting tools can significantly reduce the amount of time you need to write short-form content and long-form content. What does it mean for ASO professionals? Well, you can speed up the process of creating app descriptions and make sure they get a high Confidence score.
We decided to play with the language model of ChatGPT a bit to check out its possibilities.
First, we opted to create short-term content aka a short app description for a non-existing game.
Quite a good job, huh?
Pay attention to the words ChatGPT picked. This is a fine example of marketing copy — all the words are set to capture the atmosphere of a chilling ghost town. If you ask your content writers to polish these descriptions a bit, you’ll be golden.
Now it’s time to do the same with the long-form content. Let’s generate a longer app description of our game.
Here was our prompt:
Create a long app description for a mobile app game about ghosts. The game is a mystery. According to the plot, players have to investigate why their protagonist got stuck in a ghost city. The protagonist of the game is a girl who can see ghosts. To solve the mystery, players must collect hidden magical books and gems, solve brain teasers, join forces with friendly ghosts and stay away from evil spirits. Highlight the features: hidden objects, brain teasers, and the possibility to get to know characters from the game.
This is the text our AI copywriter created:
We didn’t ask ChatGPT to create an app title. However, it decided to come up with one, Ghost City. Some of the wording in the app description is too repetitive, but content writers won’t have any trouble editing it.
By leveraging artificial intelligence for mobile marketing, you can also create texts for screenshots or experiment with a title. ChatGPT isn’t the only solution app marketers may turn to.
Craftly, for instance, is one of the copywriting tools you can adopt to step up your content marketing game. The platform claims their AI copywriter got additional training on 10,000 hours of expert content. As for editing and rewriting, WordTune is a solid pick. Based on GPT-3, this AI assistant boosts your content production by providing alternatives to words, sentences, and even paragraphs.
We wanted to check out WordTune and put the long app description by ChatGPT into it for editing.
Most of the content writers will agree — the suggestions are great!
The new tools powered by artificial intelligence have many offerings. The AI market has been blooming with projects able to generate images, and the whole globe is in awe of the results. Admit it, you must have seen plenty of social media posts containing the artworks by Midjourney or Stable Diffusion.
The point is, you can adopt these image-related tools to work with other elements of app metadata — like, for example, icons.
To generate icons, we’ll use — surprise, surprise, — the brother of ChatGPT!
Meet DALL·E 2 — another deep learning model by OpenAI, a text-to-image system able to turn captions into artworks.
Believing that the communication should be just like with ChatGPT (detailed prompts), we approached DALL·E 2 with the following request:
Create a mobile app icon for a game about ghosts. The game is a mystery where you must collect magical books and gems to win. The icon can depict the protagonist of the game, a girl who can see ghosts. She’s trying to get out of a ghost town.
This is what we got:
It seems like the description didn’t work out. The AI tool wasn’t able to understand it properly and started creating the mobile interface/the whole phone. And the girl is, well… yikes.
DALL·E 2 also isn’t able to create captions on the image. It’s a persistent problem for artificial intelligence systems that generate visuals. Of course, these variations should never be used for an app icon.
However, we didn’t give up.
Instead of writing a detailed prompt, we settled with a short one:
Picture a cute ghost in the style of a mobile game.
And the results were much better:
Turns out simplicity is the ultimate sophistication.
When it comes to creatives, AI transformer models should be used for brainstorming. At this moment, you can’t rely on them to revamp your product page in a flash.
And, in any case, don’t forget to A/B test your creatives, AI or not.
One of the most important aspects of ASO is selecting the right keywords. An AI language model can help by suggesting the best keywords to use based on factors such as search volume, competition, and relevance to the app.
ChatGPT can be used to analyze large amounts of keyword data to identify trends and generate suggestions for optimizing app metadata, such as the title, description, and keywords. This helps to ensure that the app is visible to users searching for related terms and increases the likelihood of it being discovered.
To ease the semantic search, ask the chatbot to compose a semantic core for an app just like we did with this prompt:
We also used the chatbot to help find the most relevant keywords in the app description of Royal Match from Google Play:
The chatbot returned with the following answer:
With ChatGPT, app marketers are also able to insert the keywords into the existing app description with no sweat.
For an experiment, we asked the chatbot to create another app description of our fake game Ghost City. We choose the following keywords:
ChatGPT obeyed and put the keywords into this app marketing content:
Aside from the most obvious keyword optimization, there’s another route you can take — use artificial intelligence for localization.
During our recent App Growth Point, one of the app marketers reported that ChatGPT helped them a lot in App Store listening localization using Local Keywords. They provided them with keywords based on which Chat GPT came up with the best Localized Description. The marketer reported a 30% increase in their UA.
At the same time, marketers also highlighted that it’d be a mistake to solely rely on the results suggested by ChatGPT and the review by a subject matter expert is required to ensure success.
Another app marketer described the following situation: AI suggested a mix of irrelevant and highly relevant keywords (the marketer mentioned the results were based on the Google Play Store search suggestions only) — so it was important for the marketer to do his work on keyword research to finalize the list.
While ChatGPT is certainly fascinating, always opt for thorough target audience and keyword research.
Another critical aspect of ASO is understanding the competition.
Artificial intelligence can be used to analyze the ranking and performance of competitors, providing valuable insights into what strategies are working for them and what can be done differently.
You can use ChatGPT to compare the app’s metadata, such as keywords, titles, and descriptions, with those of competitors to identify areas for improvement. This allows for a more strategic approach to ASO and helps to ensure that the app is positioned for success.
To compare keywords between two dating apps, Tinder and Bumble, we showed their app descriptions to ChatGPT and asked for an analysis:
Compare the app descriptions of two popular dating Apps, Tinder and Bumble, and identify the common keywords within them.
This is what ChatGPT was able to identify:
Please remember that you should insert some of the data above yourself for the chatbot to process it. Since the event data of ChatGPT is restricted to September 2021, it won’t be able to provide you with an up-to-date analysis on its own.
User reviews and ratings play a significant role in app ranking and visibility.
Natural language processing systems can analyze reviews and provide insights into what users like and dislike about the app.
For example, ChatGPT can identify patterns in the reviews, such as recurring themes or issues, and provide recommendations for improvement.
This information can be used to optimize the app and improve the user experience, leading to higher rankings, more downloads, and increased revenue.
To check ChatGPT in action, we asked the bot to analyze the reviews about a widely popular mobile app for meditation and mental health — Calm.
ChatGPT came up with a clear and concise summary of the reviews we put into it:
You can benefit from this language model solution in one more way.
Save time by writing content to reply to the reviews your app gets. You can accurately respond to all the comments, win over the audience, and focus on more important tasks.
Here’s an example of such a response generated by ChatGPT when analyzing the review about Calm:
The verdict: The response sounds nice, polite, and reassuring. Just what we aimed for.
Paid User Acquisition (UA) is the process of acquiring new users through paid channels, such as social media advertising or in-app advertising.
Let’s investigate how the AI-driven digital transformation can help to optimize UA campaigns.
One of the most important aspects of UA is targeting the right audience.
AI can be used to predict which ad campaigns will perform best and adjust your campaigns accordingly.
For example, AI algorithms can analyze behavior, such as demographics, interests, and purchase history, the most effective times of day and locations to display ads for your app and identify the best audience for it.
ChatGPT can be used to provide insights into the performance of campaigns and make recommendations for improving targeting. This helps to ensure that campaigns reach the right audience, leading to higher conversion rates and lower acquisition costs.
You can also step up your Apple Search Ads efforts by developing effective custom product pages and create content for them (at least its text component) with ChatGPT.
Designing the right campaign structure (e.g. – Apple Search Ads campaigns) is one of the most time-consuming aspects of paid UA.
Hypothetically, AI can speed up not only the content production, but also the process of structurizing your campaigns.
SplitMetrics Agency leveraged AI to group keywords based on their semantics, and ChatGPT showed an ability to create quite a nice structure that can be applied. The team carried out just a few iterations of such an experiment and didn’t try it out for a real case, so the potential remains to be seen.
When it comes to ad creation, app marketers usually work with motion designers, who don’t necessarily create unique pictures and instead use ready ones from different stocks. And it takes time, effort and some degree of luck to find a proper photo that would fit. So AI creating unique images for motion designers can greatly improve the working with creatives.
We’ve already listed some of the AI-backed systems that can create a design asset from scratch. Some of them — like Mirage — are even able to create 3D objects, which can prove helpful for motion designers. One more option is to search a library of curated AI images like Lexica, the Stable Diffusion search engine.
Another critical aspect of UA is ad optimization.
Artificial intelligence can analyze the performance of ads and make recommendations for improvement. ChatGPT can be used to identify trends in the data, such as which ad creatives are performing best. This helps to ensure that campaigns are effective and cost-efficient, leading to higher return on investment.
Managing UA budgets can be a complex task, and it is important to make the most of every dollar spent.
AI can be used to analyze spending patterns, identify top performing campaigns, and re-allocate campaign budgets to them — to make the most out of your paid UA efforts.
There are also solutions that will ease one of the most common pains of the marketing industry — analyzing data in spreadsheets. As an example, SheetAI can help you automate data analysis in Google Sheets.
The use cases of leveraging AI for mobile marketing go beyond ASO and paid UA — AI tools can help optimize various aspects of your app, from user experience to revenue.
For example, machine learning algorithms can be used to analyze user behavior and optimize your app’s features and design to improve user engagement and retention. AI algorithms can also predict which in-app purchases will be most popular and make personalized recommendations to users, helping increase your revenue.
With AI-based solutions, mobile marketers can unlock the true power of user acquisition channels like Apple Search Ads. SplitMetrics Acquire allows you to hit ROAS with AI algorithms and provides you with smart suggestions that enable you to scale at a quicker pace.
App marketers can also see a big benefit of using AI to help automate routine tasks. Typically, a UA manager spends a huge portion of time making reports from Google Ads. Instead, they can ask ChatGPT to create Google Ads scripts for reporting. Obviously, this is just one of the myriads of examples.
AI, including ChatGPT, has the potential to revolutionize the mobile marketing landscape. AI algorithms can be used to analyze and optimize various aspects of your app, from ASO and paid UA management to optimization.
By using these new tools, you can stay ahead of the competition, strengthen your content strategy, reach your target audience, and improve user engagement and revenue. As AI continues to evolve, the possibilities for mobile marketing are endless, and it’s up to you to see them.
After all, AI is just a tool, not a prophet.
Unlike humans.
Unlike you.
