I am going to look at two GenAI image generation products through a product manager’s eyes. Ideogram vs. Midjourney. While Midjourney is judged by most as having the best model, my experiments this week proved that Ideogram could be ready to displace Midjourney (I will address others in future posts, this week I just compared the two.)

Ideogram is nailing the UX/UI design and making the capabilities more accessible. If you want, it is even teaching you how to write better prompts. Midjourney is still making users learn how to write a little code to get what they want. It felt like a UX battle between Apple and Microsoft, one is beautiful and easy to use, the other is clunky but easier to customize.

I have been using Midjourney for about 9 months, but it still can’t handle text, so I have been keeping my eye out for a company that can handle text in images. Ideogram has been doing it for a while, but I finally got around to testing it.

With my product sense lense on, Ideogram is clearly the winner.

Why is Ideogram (mostly) better:

• Clean interface

• Don’t have to log into Discord

• Don’t have to understand basic coding best practices

• Don’t have to look up coding tricks to get what I want

• I can see the prompt after something is created and tweak my prompt accordingly. (If I write something simple: abstract of orange slices, Ideogram shows me the more complex prompt that was really created to generate the prompt. Yes, I can do that in Midjourney, but I have to write code to ask for it.)

• I can find styles I like from the public images and find the keywords I need to describe styles I like but don’t have the vocabulary to explain.

• I can pick my favorite and try different modifications

• I can work with an existing image without crazy workarounds.

• I can take the prompt from Ideogram and copy it into Midjourney to play a little more.

What do I miss from Midjourney:

• If I can find the coding instructions, ultimately, I can have more control

• Stylistically, I find Midjourney does better with abstract

• I can modify areas of the picture with Midjourney (for example make a white man a Black woman).

If Ideogram wins, it is a win for UX.

As you can see from the list above, I took a product manager’s POV to the two products. My review wasn’t about GenAI technicalities but rather about creating a usable product that solves problems for someone who needs images to illustrate blog posts.

For those who are visual, here are some screen grabs from one experience in both platforms.