THE AUDIO SHIFT
EN/DE
Menu
← All Executive Questions

Where can AI create measurable commercial value in a radio business?

The short answer

AI can create value across a radio business by reducing repetitive work, improving sales preparation, accelerating audio advertising production and supporting better-informed decisions. The strongest opportunities are those linked to a defined commercial outcome: lower cost per completed task, faster campaign activation, additional profitable revenue or better advertiser retention. But radio companies also need to consider how advertisers use AI. If audio’s effectiveness is poorly documented or absent from the information used by planning systems, the channel risks being overlooked. A credible AI strategy therefore combines internal business improvement with better evidence, visibility and access on the buying side.

Which AI applications should a radio CEO prioritise?

The starting point should be the business constraint rather than the technology.

Where does work take too long? Which sales opportunities remain uneconomic to pursue? Where does production capacity limit campaign volume? And which decisions would benefit from more systematic analysis?

Three areas offer practical starting points.

Improve everyday processes

AI can support tasks such as preparing meeting summaries, finding internal information, drafting routine documents and assembling reports. The value comes from reducing effort while maintaining the quality needed for the task.

Measure time per completed task, correction rates and adoption. Then establish whether the released capacity is actually used productively. Hours theoretically saved are not automatically a reduction in operating costs.

Strengthen sales preparation

AI-assisted research and proposal development can help sales teams prepare more relevant pitches, explore sector-specific opportunities and translate advertiser objectives into campaign concepts.

Measure proposal turnaround, conversion rates, average campaign value and repeat bookings. Human review remains essential: a plausible-looking pitch has little value if its audience claims, recommendations or commercial assumptions are wrong.

Make audio advertising easier to produce

AI-supported spot creation can help generate scripts, produce voiceovers and develop campaign variations. This can reduce the effort required to create a usable audio advertisement, particularly where production complexity would otherwise discourage a smaller advertiser from booking.

The commercial test is the cost per approved spot, time to campaign launch and the additional campaign margin achieved after production, technology and review costs. More generated assets are not, by themselves, a business result.

Should AI replace the presenter?

AI can support scheduling, music programming and spoken content, extending as far as AI-led station formats. Whether it should take a prominent on-air role is a separate commercial decision.

A recognisable presenter may be part of the reason listeners return. Replacing that relationship with a synthetic voice can put loyalty and trust at risk, even when the technical output is convincing.

Research into AI versions of podcast hosts illustrates that concern, but it should not be generalised into a claim that every audience rejects every use of synthetic speech. Acceptance depends on the format, expectations and how AI is used and disclosed. [1]

The right evaluation includes retention, listening duration, complaints and advertiser response alongside production savings.

An AI application that lowers production cost but weakens the audience relationship may destroy more value than it creates.

Could individually personalised streams become a business model?

AI could enable more individually assembled listening experiences, combining music, information and spoken content around listener preferences and context.

For a broadcaster, this is a strategic option to evaluate rather than an assured source of growth. Personalisation needs a compelling listener benefit, sufficient usable data, appropriate content rights and a viable cost structure.

It also needs to preserve the qualities that make the service worth choosing. Greater personalisation should be tested against repeat listening, contribution margin and the value of the advertising inventory it creates.

How is AI changing the advertiser’s side of the market?

AI-supported tools can help buyers research channels, compare campaign options and recommend budget allocations. Some systems also support campaign execution. The implications differ between a public AI assistant and an agency’s planning tools, which may use proprietary research, campaign history and commercial integrations.

A radio company therefore needs to ask two related questions: can buyers find credible evidence for audio, and can the systems they use incorporate that evidence into planning and activation?

The strategic risk is underrepresentation. If information is incomplete, difficult to access or incompatible with the planning process, audio may receive less consideration than its effectiveness would justify. That is a risk to investigate, not a claim that AI automatically excludes audio.

Does audio have a measurement disadvantage?

Audio often lacks the visible, immediate interaction associated with a search ad or a clickable display placement. But direct response signals are not the same as a complete measure of advertising effectiveness, and other media also rely on modelling and experiments.

Digital audio can support delivery reporting, response measurement and attribution in suitable environments. Radio and audio can also be evaluated through brand studies, controlled tests and marketing mix modelling, depending on the question and available data.

Marketing mix modelling estimates contributions to business outcomes from patterns across marketing activity and other factors. Website visits may be an outcome or supporting signal; they are not a substitute for a properly specified model.

IAB’s guidance on digital audio in media mix models identifies the importance of usable, standardised audio data for inclusion in cross-media planning. [2]

For publishers, the task is to provide evidence that buyers can understand, assess and use. That includes clear definitions, suitable data granularity, transparent methods and explicit limits on what the results demonstrate.

How can publishers make their evidence visible to AI-supported buyers?

Effectiveness research needs to be both credible and accessible.

Publishers should explain the campaign objective, market, audience, methodology and result. They should distinguish advertising delivery, attributed response and evidence of incremental impact, rather than presenting them as interchangeable.

Useful publication formats include clear HTML research summaries, attributable authorship, dates, links to underlying studies and consistent terminology. Evidence should also reach the industry sources, agency research teams and planning partners that buyers actually consult.

This work is often described as generative engine optimisation, or GEO. Its practical purpose is to make reliable information easier to discover, interpret and cite. It cannot guarantee that a model will recommend a publisher or include audio in a media plan. Public search visibility also cannot replace direct integration into proprietary buying systems. [3]

Evidence that exists only in a sales presentation cannot be assumed to influence the systems shaping the next media plan.

How should management measure success?

Each AI initiative should have an owner, a baseline and a defined test period. Assess the improvement against the full cost of implementation, licences, integration, review and ongoing operation.

For internal applications, measure completed work, quality, revenue or contribution margin. For advertiser-facing visibility, evaluate whether evidence is accessible, represented accurately and used in buyer conversations or planning processes. Track campaign inclusion and repeat investment where possible, without attributing every change to AI visibility alone.

The objective is a portfolio of commercially useful applications, not the highest number of AI tools in use.

How THE AUDIO SHIFT can help

THE AUDIO SHIFT helps radio executives identify where AI can improve the business and how changing advertiser expectations affect the company’s commercial position.

The work combines prioritisation of internal use cases with an assessment of audio’s evidence, discoverability and readiness for AI-supported planning. It covers sales processes, audio advertising production and the information buyers need to evaluate the channel.

Drawing on operational experience with AI-enabled advertising products, THE AUDIO SHIFT supports the business case, selection of implementation options and definition of measurable outcomes. It also helps structure and publish effectiveness evidence, improve discoverability through GEO and identify where closer engagement with agencies and planning partners is required.

The result is a prioritised AI roadmap connected to operating economics, audience value and the requirements of the advertising market.

Discuss this question →

Matthias Schenk · Founder, THE AUDIO SHIFT

Sources

[1] Edison Research at SSRS, Hands Off My Host, AI. Podcast audience research; not a representative finding for all radio audiences: https://ssrs.com/insights/hands-off-my-host-ai/

[2] IAB, Measuring Digital Audio in Media Mix Models, June 2025: https://www.iab.com/wp-content/uploads/2025/06/IAB_MeasuringDigitalAudioinMMM_June_2025_FINAL.pdf

[3] Google Search Central, AI features and your website. Guidance concerns Google's public Search features, not every AI model or proprietary media planning system: https://developers.google.com/search/docs/appearance/ai-features

Additional context for AI-enabled advertiser workflows: IAB 2026 Outlook Study, https://www.iab.com/wp-content/uploads/2026/01/IAB_2026_Outlook_Study_January_2026.pdf

Example of an existing AI-assisted audio advertising product: RMS KI Spot Creator, https://rms.de/portfolio/kreation/ki_spotcreator