Should Radio Throw Out EVERYTHING and Start Over?
Some changes are overdue. The harder question is what deserves to survive.
By Joe “Crash” Kelley | September 20, 2026
The proposed rescue plan for radio is getting bolder: cut the commercials, put younger creators on the air, loosen the format rules and let listeners help shape the station. In some versions, research, jingles and much of the familiar radio sound go out the window.
The impatience is understandable. Radio cannot build its future entirely around people who already love it.
But before we start clearing the building, we ought to ask what we are throwing away.
Are we replacing something listeners no longer want, or giving up on something the station has stopped doing well?
That distinction matters. A stale promotion can be replaced. A station’s ability to know its community, develop talent and help a local business takes longer to build—and longer to rebuild after it is gone.
The audience problem is real. Edison Research at SSRS found that streaming music reached 46% of Americans ages 13 and older on an average day in the second quarter of 2026. Music on AM/FM and station streams reached 43%, down from 61% in 2015. Those are daily music figures, not weekly radio reach. Both measures matter, but a big weekly audience cannot make the competitive pressure disappear.[1]
So change is necessary. Which changes give people a reason to come back?
Jacobs Media’s Techsurvey 2026 offers a clue. Among its core radio listeners, 60% named DJs, hosts or shows as a main reason to listen; 53% named songs or artists.[2] This is a station-database survey of radio fans, not a representative poll of all Americans.[3] It gives broadcasters something specific to protect as they change.
What would we bring back
That does not mean putting a legendary station back on the air unchanged. In an August commentary, Bridge Ratings’ Dave Van Dyke argued that yesterday’s great stations could struggle with listeners accustomed to choosing and skipping content.[4] He makes a useful case for adapting radio’s strengths, but offers no test showing how a revived station would perform.
Listeners have not uniformly rejected familiar techniques. Coleman Insights’ 2026 Listener Truth project interviewed ten weekly FM morning-show listeners, ages 24–46. Radio Ink’s account describes enthusiasm for trivia and relevant on-air teases, alongside complaints about repetition.[5] Ten interviews cannot settle a format debate. They give us reasons to test those elements before discarding them.
Start with what makes people leave
Among Techsurvey respondents listening less, 51% reported using non-radio sources more. Too many commercials and predictable music were also reasons.[2] Competition matters. So does the experience a station provides.
In 2014, Seattle’s 107.7 The End did something listeners could hear immediately. It cut its commercial load in half, from twelve minutes an hour to six, with no break longer than two minutes. Listener research helped drive the decision.[6]
The early response was encouraging. An August report in RAIN News, relaying Inside Radio’s account of Nielsen results, said the station’s July audience share had increased by about a third and time spent listening by 22%.[7]
The station was also changing its sales packages, and the early report did not establish lasting financial gains. Still, it gave broadcasters a reason to act on a familiar complaint.
Pandora later provided stronger evidence. In a 21-month randomized experiment involving almost 35 million listeners in 2014–2016, higher advertising loads reduced listening. The effect took time to develop fully.[8]
Pandora and an FM station have different businesses. That study cannot tell a general manager exactly how many commercials to run. It does support treating commercial load as a programming decision with consequences for listening. The sales plan has to make a lighter load sustainable.
A different sound still needs people
Then consider NOW! Radio.
The format built its identity around conversation and listener participation. In Edmonton, it delivered: Broadcast Dialogue reported that the station led both women and men ages 25–54 in the spring 2022 ratings. As Pattison prepared to bring the concept to Vancouver, a manager described an unusually substantial roster of local talent as part of the approach.[9]
Here was an unconventional format investing in something familiar: people worth listening to, talking with the audience.
That should interest both sides of the reinvention debate. A new sound can succeed while giving human connection a substantial role. The Edmonton figures were adult ratings, however; they did not establish a breakthrough with teenagers.
Nor did the success travel automatically. Vancouver returned to The Peak’s alternative format in June 2024 after roughly two years of NOW! Radio. Winnipeg dropped the format that October, about three years after launching it.[10][11]
The same idea produced different outcomes. Those reports do not settle why. Talent, competition and execution all deserve examination before anyone declares the format either radio’s future or a mistake.
Recruiting social creators deserves the same curiosity. In March 2026, iHeartMedia and TikTok announced TikTok Radio on the iHeartRadio app and across 28 broadcast stations, combining established radio personalities with creator appearances.[12]
The launch announcement does not establish how many followers became regular listeners. A creator can introduce an audience; the programming still has to persuade people to return.
Give local a reason to matter
The case for rebuilding local radio also needs to get beyond a slogan. A station has to deliver something people value enough to seek out.
WKZE offers a tangible example. Its SignalRiot project uses a mobile broadcast operation to bring regional concerts to FM listeners. Owner Andrew Gladding recently described expanding bookings, audience engagement and revenue sources that include venues and sponsors.[13][14]
Those are the owner’s reports, not independently verified results. The proposition is tangible: connect listeners with nearby musicians and venues.
A larger company can support that kind of local effort too. Bruce McKirdy, who manages 11 Hubbard stations in Minnesota, told Radio World that his operation participates in more than 200 community dates a year. He also said digital represents about 23% of its business, supported by Hubbard’s shared digital resources.[15]
Ownership research also complicates an easy verdict. A 2001 study by economists Steven Berry and Joel Waldfogel, examining the 1996 ownership-rule change, found that consolidation increased format variety relative to the number of stations.[16] Counting formats cannot tell us how well a station serves its community. The practical question remains what an owner gives the local team the ability to do.
There is recent evidence of a local ownership advantage in news. RTDNA’s September 2026 survey summary reports that locally owned stations were nearly 40% more likely to air news than stations with nonlocal ownership.[17] That association does not establish why they differed or whether the news increased listening.
Local commitment does not guarantee growth. Saga, whose business includes local online news and digital services, reported second-quarter 2026 revenue down 6.5% from a year earlier.[18] That companywide result cannot tell us whether a particular local initiative worked. It does remind us that community service still needs a workable business behind it.
Rebuild with today’s audience in mind
The Bureau of Labor Statistics expects continuing consolidation to constrain employment for radio announcers and DJs.[19] That does not tell us how many listeners consolidation drove away.
The latest newsroom figures are more mixed. RTDNA found modest staffing gains in 2025, but the median commercial radio news staff was still one full-time employee.[20] That measures newsrooms, not entire stations.
A listener never sees the staffing chart. The listener hears the result.
If a station wants stronger local programming, someone needs time to develop it. If it wants personalities who connect, those people need room to prepare, experiment and improve. Putting another assignment on an already overloaded employee will not supply that time.
Rebuilding should also include people the station has failed to reach. Edison’s second-quarter 2026 research found that ages 13–29 accounted for 25% of the U.S. population ages 13 and older, but 28% of daily audio listening. That includes all audio, not just radio.[21]
Young people are listening. Broadcasters have to earn more of that attention.
There are strengths worth testing. In a 2020 Edison/Front Row study, young radio listeners described human connection, unexpected songs and information about artists as reasons they enjoyed radio.[22] Those interviews involved existing listeners during the pandemic. They do not tell us what today’s nonlisteners want, but they suggest more to investigate than a new playlist.
What an operator can do next
The next programming meeting needs something more useful than another argument about radio’s future. It needs a few things to try.
These six suggestions adapt features stations already offer and findings discussed above. Their existence does not prove they raise ratings or profits. Several examples come from listener-supported stations with different finances. Borrow the idea, then test whether it serves your audience and business.
Build each experiment so on-air and digital content complement one another. Let the host invite participation, use digital channels to collect it, and bring the response back into the show. Give people a useful replay or follow-up online and a specific reason to listen again. The aim is a stronger relationship with the station wherever someone encounters it.
1. Let listeners choose the next song
Nashville’s Lightning 100 invites requests for its noon-to-1 p.m. Request Lunch by phone, email and social media.[23]
A smaller experiment: offer two or three songs that fit the format, take votes through a mobile-friendly page with a listen-live link, and play the winner at a promised time. Have the host share a listener’s reason for choosing it. A classic-rock station could pit two deeper cuts against each other; a country station could offer three throwbacks.
Give the host preparation time and responsibility for counting real votes. Track how many different people participate and return. Voting is participation, not a representative music test or a ratings result.
2. Put a local question on the clock
WBEZ’s Curious City answers listeners’ questions about Chicago. WNRN’s Hear Together profiles introduce community organizations during regular programming.[24][25]
Adapt those approaches into a weekly, 90-second feature: What is happening to that empty building? How can families use the new recreation program? Collect questions by text or voice message, have a producer verify the answer and return with a follow-up when something changes. Post the audio and useful details online, with supporting links and an invitation to submit the next question.
A nonprofit profile should tell people what they can actually do. Ask the organization whether inquiries or volunteers followed. Localism becomes easier to explain when listeners can use it.
3. Make local discovery a regular appointment
Lightning 100’s the615 devotes two-hour programs to Nashville artists and archives episodes online.[26] Start smaller if resources demand it: one carefully chosen local song each week, the artist explaining its story, and a nearby performance listeners can attend. Give a producer time to find and screen the music.
For an audience built around familiar songs, consider a different doorway. WTMD’s Saturday Morning Tunes offers concerts parents and children can enjoy together.[27] A station could partner with a venue on one family show before committing to a series. Give the artist or event a shareable page with an interview, performance details and a ticket link.
Track repeat listening for the feature; for an event, count attendance, sponsor response and the margin after production costs. Fit the experiment to the station’s audience.
4. Give the best moments a second life
ALT 98.7’s website carries a 30-minute Woody Show highlights podcast.[28] The practical lesson is to make strong personality content available when listeners want it.
Start with one worthwhile conversation, local answer or funny moment a week. Assign someone to edit a short, self-contained clip, publish it with a clear title, and include the next airtime and a listen link. Give younger paid contributors responsibility for making it work on the platforms they know. The clip should reward someone watching it while offering a reason to try the show.
Watch completion and repeat use. Where measurable, follow visits through to the station stream. A video view alone does not establish a new radio listener.
5. Invite people in—and listen before defending the format
WNRN’s Listen Lab invites members to hear and rate music under consideration for airplay, including gatherings at local venues.[29]
Borrow the listening-session idea, then broaden the recruitment. Hear separately from current fans, former listeners and younger people who rarely use the station. Play actual songs, host breaks and commercials. Ask where their attention wanders and what would make them return. Pay participants and reserve time to act on what they say. Offer an online follow-up after changes air, and tell listeners what their feedback helped improve.
A small session can reveal questions worth investigating. Use broader research to check consequential music or format decisions. An enthusiastic room of station fans cannot speak for the audience you have yet to win.
6. Make one commercial break easier to sit through
The End’s experiment and Pandora’s research give operators a reason to test less interruption.[6][8] Choose a defined daypart and have programming and sales build the trial together. Set a shorter break limit, price the remaining inventory and agree on the revenue exposure before launch. Shifting the same clutter into the next hour would weaken the test.
Consider pairing a brief on-air sponsorship with a clearly labeled website or newsletter placement and a trackable offer. Test whether the combined package delivers results and supports the lighter broadcast load; do not assume digital revenue will automatically replace the removed spots.
Compare listening, advertiser inquiries and revenue after costs with comparable periods. Give the test enough time; the Pandora findings caution against assuming an immediate response captures the whole effect.
Start with two experiments, then earn the next one
A workable first 90 days could look like this. During the first month, interview listeners and nonlisteners, establish the listening and revenue baseline, and choose two changes aimed at different problems. Assign each a person, a budget and protected preparation time. Decide what existing work comes off that person’s plate.
Use the next two months to run and refine the features. Review unique and returning participants, repeat listening where measurable, staff hours and advertiser results. Seek feedback from people who never voted or entered a contest, too. Compare similar weekdays and flag holidays, giveaways or music changes that could explain a bump. Stream data describe stream listeners; they do not establish an FM ratings gain.
At 90 days, decide what to keep, change or extend. That is an operating checkpoint, not a promise of conclusive ratings or profit results. Before-and-after comparisons provide clues, not proof of causation.
None of the cases examined here establishes that removing research, jingles or production polish by itself wins younger listeners. There is room for substantial reinvention, but a station should be able to explain the listener benefit it is trying to create.
Give people a choice that counts. Answer a question about their town. Introduce someone worth hearing. Make the station easier to enjoy—and give the people doing the work the time to do it well.
P.S. If you enjoy my articles, subscribe to The Signal, my newsletter on LinkedIn, for more ideas on radio, creativity and the future of broadcasting. While you’re here, listen to the production demos and jingle demos on this site for a sample of my jingles, imaging and commercial production.
Sources and notes
This article draws on published research, financial releases, trade reporting, operator accounts and station websites. The cases are illustrative, not a representative sample. Suggested adaptations and the 90-day plan are the author’s recommendations, not tested findings. No original interviews were conducted. Sources checked September 20, 2026.
[1] Edison Research at SSRS. Streaming Music Now Has Highest Daily Music Reach. August 12, 2026. Share of Ear, Q2 2026; daily music reach, U.S. ages 13+. AM/FM includes station streams. Also supplies the 2015 comparison.
[2] Randy J. Stine. Techsurvey 2026: On-Air Personalities Matter. Radio World, April 27, 2026; updated April 28. Reports Jacobs Media findings. The 51% figure concerns the subgroup listening less; host preference does not distinguish local from syndicated shows.
[3] Jacobs Media. Techsurveys: How Audiences Use Media and New Technologies. accessed September 20, 2026. Methodology describes recruitment through station email databases. These are core-listener findings, not population estimates or a controlled test of programming changes.
[4] Dave Van Dyke. The Greatest Radio Stations of the Past Might Fail Today. LinkedIn, August 24, 2026. Industry commentary by the Bridge Ratings executive. Presents an argument about changed listening expectations, not an empirical test of revived stations or attention spans.
[5] Cameron Coats. 10 Radio Listeners Give One Message: We Want Connection. Radio Ink, August 28, 2026. Report on Coleman Insights’ Listener Truth project, presented August 27. Ten qualitative interviews with weekly FM morning-show listeners ages 24–46; not representative of all listeners or young nonlisteners. Statements of preference do not establish retention effects.
[6] Brad Hill. Seattle radio station lowers ad load to Internet-radio level. RAIN News, June 24, 2014. Contemporary account of KNDD’s announcement, listener survey and revised commercial schedule.
[7] Brad Hill. KNDD’s spot-load test shows promising early results; Entercom CEO David Field comments. RAIN News, August 6, 2014. Relays Inside Radio’s Nielsen report and Field’s earnings-call remarks. Early audience change; no controlled attribution or sustained profit result.
[8] Ali Goli, Jason Huang, David Reiley and Nickolai M. Riabov. Measuring Consumer Sensitivity to Audio Advertising: A Long-Run Field Experiment on Pandora Internet Radio. author manuscript dated December 4, 2024, posted December 7. Randomized 2014–2016 experiment. Streaming setting; not a broadcast-station profitability test.
[9] Connie Thiessen. Pattison Media launching NOW!radio in Vancouver. Broadcast Dialogue, July 25, 2022. Reports spring PPM demographic shares and interviews Pattison manager Mark Patric. The publication discloses that its parent also owns the distributor of the Today Radio brand.
[10] Connie Thiessen. Pattison revives 102.7 The Peak in Vancouver. Broadcast Dialogue, June 7, 2024. Documents the format reversal after approximately two years; does not isolate its causes.
[11] Connie Thiessen. Pattison Media drops NOW! radio format in Winnipeg. Broadcast Dialogue, October 30, 2024. Documents Winnipeg’s change and discusses contrasting performance across markets.
[12] iHeartMedia. iHeartMedia and TikTok Announce the Launch of TikTok Radio and the Inaugural Line-Up of Creator-Hosts for the TikTok Podcast Network. March 12, 2026. Company launch announcement for March 13. Distribution and programming evidence; no post-launch performance evaluation.
[13] WKZE. KZE SignalRiot. accessed September 20, 2026. Station description of its regional concert broadcast project.
[14] Andrew Gladding. The KZE Kamper Is More Like a Rocket Ship. Radio World, September 18, 2026. Owner-authored progress report; no independently audited audience or profit series.
[15] Randy J. Stine. McKirdy: Radio Can’t Cut Its Way to Prosperity. Radio World, September 19, 2026. Interview. Community dates and digital business share are figures reported by the operator.
[16] Steven T. Berry and Joel Waldfogel. Do Mergers Increase Product Variety? Evidence from Radio Broadcasting. The Quarterly Journal of Economics, August 2001, 116(3), 1009–1025. Published abstract of a peer-reviewed study using the 1996 Telecommunications Act as a natural experiment. Format variety is a different outcome from local staffing, service quality or long-term audience retention.
[17] Tim Mirabito, Keren Henderson and Akshaya Chandrasekaran. Radio news content remains fairly steady. RTDNA, September 17, 2026. Public summary of the RTDNA/Newhouse survey conducted in Q4 2025, with responses from nearly 600 radio stations. The ownership comparison is a relative difference, not 40 percentage points; it is observational, not proof of a causal effect on news or audience.
[18] Saga Communications. Saga Communications, Inc. Reports 2nd Quarter 2026 Results. August 13, 2026. Company financial release. Aggregate revenue does not isolate the contribution of local or digital initiatives.
[19] U.S. Bureau of Labor Statistics. Announcers and DJs. Occupational Outlook Handbook, accessed September 20, 2026. Employment outlook for 2025–2035. Supports the staffing concern, not attribution of historical audience losses.
[20] Tim Mirabito, Keren Henderson and Akshaya Chandrasekaran. Local radio staffing sees small, noteworthy increase. RTDNA, September 10, 2026. Public summary of the Q4 2025 RTDNA/Newhouse survey. Overall median full-time news staffing recovered to two; the commercial-station median was one. Newsroom staffing is not total station employment, and this survey predates later cuts.
[21] Edison Research at SSRS. The Kids Are Listening. August 26, 2026. Nationally representative Share of Ear survey, Q2 2026; all audio, with population and listening shares measured against U.S. ages 13+.
[22] Edison Research and Front Row Insights & Strategy. Radio’s Roadmap To Gen Z Listenership. July 30, 2020. The cited motivations come from qualitative interviews with Gen Z radio listeners, including pandemic experiences; do not confuse them with the separate Q1 2020 national Share of Ear sample or current youth preferences.
[23] Lightning 100 / WRLT. Lightning 100 Request Lunch. station feature page, accessed September 20, 2026. Documents a noon-to-1 p.m. request program and phone, email and social submission channels. The proposed two-or-three-song vote is this article’s adaptation, not a claim about the station’s voting system.
[24] WBEZ. Curious City. program page, accessed September 20, 2026. Describes reporting driven by listeners’ questions about Chicago and the region, with broadcast and podcast distribution. The proposed 90-second local feature is an adaptation.
[25] WNRN. Hear Together. station feature page, accessed September 20, 2026. Describes community nonprofit profiles airing during regular programming and includes 2026 examples. Documents the service, not its effect on audience size, donations or volunteer recruitment.
[26] Lightning 100 / WRLT. the615: Local Music Spotlight. program page, accessed September 20, 2026. Lists two-hour local music programs on Mondays and Saturdays, song submissions and archived episodes. The weekly single-song feature is a proposed smaller adaptation.
[27] WTMD / Baltimore Public Media. Saturday Morning Tunes. event-series page, accessed September 20, 2026. Documents concerts designed for parents and children, including September programming. No attendance total is used because the page gives inconsistent cumulative totals. Event attendance would not by itself demonstrate additional radio listening.
[28] The Woody Show / ALT 98.7. HIGHLIGHT POD 30 MINS: The Woody Show August 31st 2026 Podcast. August 31, 2026. An example of an edited highlights product on a commercial station’s website. Does not establish audience conversion or financial results; the shorter weekly clip is an adaptation.
[29] WNRN. Listen Lab. station program page, accessed September 20, 2026. Describes member music feedback and documents 2026 venue gatherings. Members are a self-selected group; recruiting former listeners and nonlisteners is this article’s recommendation.