There aren't many tools available on the market to simplify the work of ASO specialists — in fact, there are almost none. All that’s left are in-house solutions and, of course, AI tools.

When combined, these tools can help free your team from tens to hundreds of hours of routine work each month (or even more if you have a large department). For an independent studio with limited resources, these tools are nothing short of essential.

Creating visuals for ad creatives with limited resources

Your goal is to attract organic traffic through “Similar” collections that Google displays on the pages of top projects. Often, the style of these projects is similar, which presents a great opportunity for indie studios without dedicated designers or resources for large-scale experiments to boost organic traffic. You’ll need ChatGPT, Midjourney, and possibly Photoshop.

• Step 1: Ask ChatGPT to generate a prompt for Midjourney. Describe what you want the image to look like.
• Step 2: Use Midjourney to create the visuals.
• Step 3: Manually review the results. If necessary, regenerate the images.
• Step 4: Use Midjourney's built-in tool to stylize the draft versions.
• Step 5: Finally, run the previously generated visuals through this stylization process to achieve a consistent style.

In some cases, you may need the latest version of Photoshop for final adjustments.

For example, if you’re short by 50 pixels in width, select the area in Photoshop, click "complete background," and the internal AI tool will extend the background to the required size. This is usually quicker and easier than cropping the image or reassigning tasks to neural networks.

How to make the right collections?

However, creatives won’t be very effective if you target irrelevant projects. Here are a couple of tips for analyzing competitors:

First, analyze the search results for your keywords and pick out relevant projects from the top results.

Next, use an analytics service to check the actual traffic volumes. Sometimes a project might look great visually but doesn’t generate much traffic, which isn’t useful for your goals.

Also, some services, like Sensor Tower, allow you to separate paid and organic traffic. This should be considered when compiling collections you want to appear in. The more paid traffic your competitor has, the less organic traffic you’ll receive if you get into their collection.

Don’t forget about related and intersecting topics. For example, if you’re developing a game in the fashion genre, you might also be interested in makeup-related projects.

Keep in mind that there’s no clear algorithm for how stores create collections. The more you test, the better your chances. We’re not shy about experimenting, and it pays off. However, for small indie studios, this can be challenging as it may dilute your focus.

For example, we also create entirely unique visuals without considering competitors. But for small studios with limited budgets, allocating resources for such experiments can be difficult.

Bonus: Automating routine tasks

Savings: Up to several hours per event.

Creatives might be one of the key factors for success in marketing and ASO, but they’re not the only time-consuming task. I know for a fact that many ASO teams still do a lot of routine work manually. This not only takes up time but also drains energy that could be better spent on more creative tasks.

For example, LiveOps often requires translating description texts into various languages — the more, the better. While ChatGPT can help, the process of copying, pasting, and editing can still take hours. If you have multiple projects that need translation into dozens of languages, this can consume entire days.

That’s why we created a Google spreadsheet with the main description in English and columns for other languages. We wrote a script via API that accesses ChatGPT. The script is too simple to explain here and requires minimal knowledge — ChatGPT can even help you write it. We mark the languages we need in the table with checkboxes and click Translate.

I also wrote a handler that distributes all the texts into the required language columns, checks the character count, and highlights any columns in red if there’s an excess. Then we manually review, make corrections, and compile the entire table into one CSV file using a Google template.

All that’s left is to upload this CSV along with the LiveOps data. Sounds simple, right? But this method saves a huge amount of work hours each month, and surprisingly, few people are using it yet.

Additional tools we use

Savings: Up to several minutes at a time.

What I’m talking about here is analyzing and managing a large number of experiments in Google Play.

Although I’m a developer with a diploma to prove it, I had never written Chrome plugins before. But with ChatGPT, I managed to create one in just a couple of evenings.

You can view test results in absolute numbers on the Google Play experiments page. For example, one banner might generate 12,000 installs, while another brings in 9,500 — but what’s the percentage difference?

That’s where a plugin that calculates these percentages becomes your best friend if you’re constantly running experiments.

Another issue that can arise with numerous experiments is tracking them. Let’s say we have 20 different tests running — we need to copy each title and Google Play Console link into Jira.

Here’s where a plugin comes in handy. It parses the Google Play page for today’s experiments and downloads a text file with links and titles. This is effective automation, especially when tests are continuous like ours.