Within media production - be it for brands, agencies or publishers - AI has reached a degree of maturity that’s truly exciting.

The applications of generative AI are novel enough to allow innovators to achieve groundbreaking campaigns - like cloning voices to deliver messages with 1:1 personalization. 

At the same time, AI production processes - not just tools - are developed enough to become highly versatile, repeatable and scalable.

With the revolution of the traditional content production process - particularly in audio - AI has changed not just the economics of media production - but also the ‘art of the possible’ for those who make content.

We’ll explain exactly what that means in this blog.

Shifting the paradigm for audio content production

If you work with a ‘traditional’ production process for audio, you know what the constraints are. To simply voice an ad, for example:

1. Someone manually writes copy for a script.

2. It goes through internal iteration and approval from key stakeholders. 

3. Voiceover talent needs to be sourced, rates are negotiated and the recording is waited for.

That process alone - including the production that follows - can take 5-14 days to produce a single ad - only for an end client to decide things ‘aren’t quite right’.

With an AI-powered workflow, the production process can take minutes - if not seconds - with instant revision and adjustment in a click.

It depends on a lot of individual, mature technologies, working together - AI-powered text-to-speech; noise cancellation; large-language models for script generation. The list goes on.

But rather than being embedded as individual tools in a fragmented way, the end-to-end transformation of the workflow allows the process to be streamlined, automated, repeated and optimized.

What does an AI-powered audio workflow change about the economics of current media production?

We know that AI is helping content producers and advertisers optimize workflows - not just tools - with faster, better and cheaper alternatives.

At a surface level, this changes more than a few costly processes, as:

• Sourcing and booking voice talent for ad reads, billable by the hour ➡️ Generating voiceovers on-demand with lifelike text-to-speech voices for campaigns of any scale.

• Enterprise-grade production software and the specialist expertise  ➡️ Mixing and mastering audio instantly and seamlessly after voicing into broadcast-ready ads

• Versioning content at scale with dynamic audio, highly personalized messaging - manually  ➡️ Doing the above, but pulling data from an API/spreadsheet to automatically generate thousands of new versions of an ad (e.g. with product/price/location and language changes).

And these benefits aren’t limited to agencies and advertisers. 

For publishers,  implementing a self-serve ad creation tool allows you to streamline the process for small/medium advertisers wishing to reach your platform’s audience - for example, by allowing users to create assets (that typically demand high creative effort for low return) and having an intuitive platform that connects to your ad delivery platform(s).

• See how DanAds supports brands like Bloomberg and MMS to access this elusive revenue stream with self-serve AI audio

And even the audio sales process is now massively streamlined - gone are the risks and effort of crafting audio demos for clients who ultimately won’t convert; AI-powered production can empower sales teams to create broadcast-ready demos and reach further, faster and engage them in an entirely new, interactive kind of sales process.

In these cases, these are the removal of current barriers and cost-sinks. So, without barriers, what are the opportunities?

How does AI audio change ‘the art of the possible’?

If you are in a brand creative pod, a media agency or even a specialist DCO team, you might know two things to be simultaneously true:

• Audio content has distinct advantages, both as a standalone medium and to enhance visual content.

• You can’t produce audio at the scale, pace or complexity your campaigns/activations need.

Of course, with AI audio production, the second is no longer true.

As we have talked about at length in our Omnichannel series, we can now use AI-powered audio to harness those advantages to make campaigns and always-on activations that are: hyper-targeted, hyper-responsive and hyper-relevant.

Take the example of PENNY (Rewe), a supermarket that used AI-powered audio to generate 2,800 audio spots dynamically targeting 220 products, prices and 26 locations in Germany - for an 81% reduction in cost and in days, not weeks.

• Read about PENNY’s DCO Campaign here.

But those benefits, again, aren’t just for advertisers. A great example of this workflow transformation in action is that it allows publishers to synthesize their news articles into different formats like podcasts: this can be done with AI, in real-time, giving articles greater immediacy and reach (e.g. to commuters who prefer to listen handsfree) - as well as providing a new inventory channel for audio advertising.

(And given that advertisers can experience approximately $4.9 ROAS for just a $1 ad spend within podcasts, it’s a very appealing option).

• Read or listen to a few strategies for doing that here.

So what’s left? Predicting what your edge cases care - and learning where AI can help you be the first to turn them into use cases.

Make edge cases use cases

Here are three examples of where we are seeing the earliest stages of the innovation curve in action.

1. Handling the Radio Night Shift: night shifts on broadcast radio already use automation, to the extent of pre-recorded programs and jukeboxes. With AI, hosts can clone their voices to create more engaging programming, and delivering greater variety in night-time programming. See how AudioStack is powering AI radio programming already with aireal.

2. In-car advertising: whether providing voice assistant functionality or publishing content to captive audiences with in-car apps; we are already seeing exclusive content deals between auto manufacturers and publishers to provide exclusive content. And if that content is tailored, the advertising inventory can be too.

3. Automating data-driven DCO: what can you do if you continuously update campaign content based on any data feed/spreadsheet/API? One thing’s for sure: you will never react slower than your next best sales opportunity, like a flash sale.

Fundamentally, there are two aspects in which AI-powered audio production is changing the economic outlook for brands, publishers, advertisers and beyond: removing traditional inefficiencies and costs, and from that, unlocking the revenue-generating opportunities behind them.

If you would like to learn how AI-powered audio can deliver that for you, get in touch.

Contributor Acknowledgements: Penri Jones (Arloesi) is an expert strategist and advisor in AI transformation and disruption in Content Supply chains.