APP launches automated video creation tool

[coldmountain]

A few weeks ago Applovin posted on their axon blog “the first version of a fully automated, multi-agent production pipeline that generates interactives for advertisers promoting purchases on their websites.” They describe this as:

A coordinated network of AI agents handles the workflow. Each agent performs a specific task, passes output to the next stage, and validates constraints throughout the process. At a high level:

• Brand context, product imagery, and messaging constraints are gathered automatically

• Multiple structured concepts are drafted

• Creative assets are generated and iteratively refined using image and video models

• Quality checks are performed before delivery

The complexity remains hidden from our users: they receive ready-to-use outputs without having to manage production.

This was described as being only the first step:

This is the first phase. Our immediate priority is to make AI-generated interactives widely available:

• Existing advertisers will receive automatically generated interactives on a regular cadence

• New advertisers will get them as part of onboarding and on an ongoing basis

• Soon, all advertisers will be able to directly generate interactives with custom inputs for greater brand control

Further, we are expanding capabilities in two directions:

• Deeper: More templates, greater adaptability, and improved HTML formats.

• Broader: Extending automation into video generation and playables. Our goal is to automate the full creative stack.

We have already begun limited testing of fully automated video generation, producing videos up to 60 seconds for select advertisers. Early results are encouraging, with some generated assets emerging as top performers.

We anticipate expanding automated video generation to a broader group of advertisers in early Q2.

Today they posted a new note describing their tool for automating video creation for clients (addressing the Broader:…. bullet above).

[Smorgasbord1]

This world is still in flux.

Co-incidentally with APP getting into video creation, OpenAI announced they’re getting out, stopping Sora, their general purpose video creation tool. I think both within 24 hours of each other.

[coldmountain]

The App blog post is undated, but is a couple of days old, and in it the state the pilot program started a couple of weeks ago. But yeah, certainly things are moving!

Another note on the App tool is that it is very targeted and limited to the ad, which is quite different from openai’s.

[wpr101]

Some encouraging anecdotal news here that I have been getting ads for AppLovin’s Axon platform while scrolling on Reddit.

I believe this is encouraging news for both names. AppLovin views Reddit as a good place to advertise. This also likely means that AppLovin’s platform is either in general availability, or getting close to a full launch.

There’s a few different variety of ads I’ve seen for Axon. One of them had the CEO Adam Foroughi explaining how the platform scales businesses. There was an example business called “plateful” which explained their scaling journey in about a 40 second video.

Grabbed a couple screenshots of how it looks. Reddit does a clever job of making the ad seem like a Reddit post. I’m not sure if I am seeing the ads because I’ve searched for AppLovin before elsewhere, or if it’s going out to a much broader audience.

[brittlerock]

I spend a fair amount of time on YouTube and often see AppLovin ads - actually, I should say always. I even get AppLovin ads delivered to my Gmail inbox. But keep in mind, seeing an ad promoting AppLovin is not the same as seeing an AXON placed ad for some other product.

I assume I see AppLovin ads because I went to the AppLovin website a while ago and pretended to be a business owner investigating what one needed to do in order to use their service. Though I did not provide an email or any specific identifying information, they were able to capture enough information from my visit to figure out who I was with considerable detail.

As recently as a few days ago I read an article on SA that asserted that AXON was not yet GA. They are still perfecting their video generation tool, per the article with GA anticipated in 2H26 or even as late as 1H27. The careful focus is on getting it right as opposed to getting it out. I’m quite certain that AXON is still only avaialble by invitation in the ecommerce market. I don’t know how to secure an invitation, but it appears that their user base is growing.

Despite the erosion in stock price, my confidence in this company is unshaken. There is mounting competition. That’s to be expected. Meta and Google (and maybe more) have obviously seen the enormous success of this company and they are aggressively trying to make sure that AppLovin doesn’t dominate the ecommerce market. But, it’s a big market, AppLovin doesn’t have to dominate in order to gain a lot of revenue all they need to do is take share.

And further, Adam Foroughi has asserted more than once that he doesn’t see Meta as a direct competitor (I don’t know if he has said anything specific about Google). AppLovin isn’t trying to climb over the FB/Insta garden wall. He sees them as complimentary not competitive. I’m not sure I fully agree with that.

CloudX and Unity are also trying to take share away from AppLovin in the gaming market. I assume that all these competitors have high caliber engineers and data scientists on staff. The problem they face is that they don’t have the data collection capability provided by MAX, Adjust and AXON. Vast quantities of high quality data is the foundation of any AI capability, it is the raw material of intelligence. In the gaming market, AppLovin holds an incomparable database, I don’t know how much penetration they have elsewhere. At this time, I think it safe to assume that they don’t possess nearly the same depth of data as they have in the gaming market. I think it’s also safe to assume that that will change over time.

IMO, the competitive landscape boils down to ROAS and LTV. Ultimately, business owners will gravitate to wherever they receive the best return on their advertising dollars as well as wherever they can broaden their customer base. AppLovin doesn’t have to have an exclusive lock on their customers. If a business owners determines that it serves the business best to use both Meta and APP what’s wrong with that?