Gannett papers using AI for high school football

[Slacker]

"Starting at center, No. 78436."

Ricky Vaughn, California Penal.

 

[LanceyHoward]

We decided to put it behind the hard paywall (no sports coverage generates subscriptions as effectively as preps), and it worked well. Each post would generate 2-7 new subscriptions — three times a week. There ended up being a lot of value for readers in having it, even if that value was just the score. We were able to fill a gap in the market with the lack of MaxPreps coverage.

How did your paper determine what stories brought in subscribers? And if covering high school sports brings in lots of subscriptions what sports topics are duds at bringing in subscriptions?

 

[Matt Stephens]

How did your paper determine what stories brought in subscribers? And if covering high school sports brings in lots of subscriptions what sports topics are duds at bringing in subscriptions?

Two main metrics were LTCs (Led To Conversion — stories that didn't necessarily convert a subscriber directly but was one of the 3-5 free articles a user read before they hit the paywall and paid) and direct conversions (article they hit the paywall and became a subscriber for).

Preps content was always high traffic and there wasn't competition in the Charlotte market. Decided one fall (either 2020 or 2021) to put all preps content behind the paywall and almost everything started driving multiple subscriptions. Charlotte is fortunate to have Langston Wertz Jr. dedicated to covering high schools — he's done it since the 80s and loves it — and that always helped with the page view numbers, but the thing about high school sports even in markets where page views aren't great for preps (which is a lot of them), that coverage can still drive subscriptions if you cover it well — most likely more subscriptions than any other best, regardless of desk. And there are always new parents and extended family members every fall who are willing to subscribe to read about their kid's team.

For Panthers game day coverage, we'd usually only hard paywall the in-depth analysis of a game that would go online a couple of hours after the game's end. (We'd convert the instant analysis/gamer URL that went live as soon as the game ended and drove the most page views because of Google to the in-depth paywall piece because it already held that domain authority.) That article would usually get 3-15 direct conversions each week. More subscriptions early and late in the season. When the team is bad, mid season traffic and conversions are harder. All other game day content would usually at least generate a couple of LTCs.

We made our NASCAR coverage model focused on page views rather than subscriptions. It drove a ton of traffic (like 250,000 - 1 million PVs in a day for Daytona 500 coverage alone, with the 2020 race and wreck driving the most). We traveled to Daytona, Talladega and any regular season race within a 6-hour drive, plus all playoff races, and it was worth it for the traffic we got from articles, but we learned pretty quickly our online audience for NASCAR was national and not local, so getting paid subscribers was much more difficult. We still did a lot of in-depth features and found quality exclusives and analyses, and they'd again get good page view numbers but not conversions.

Raleigh was a more unique situation with colleges coverage and how saturated the market was for UNC and NC State. We had solid page view numbers across the board, and steady subscriber numbers, but no beat stood out as the one that could always get the most conversions. (I should add that when Natalie Pierre arrived to focus on leading our colleges coverage, everything got better across the board.) There were also outliers, of course. When we were able to break Hubert Davis becoming the next UNC coach, traffic went nuts. Plenty of subscribers came, even if it wasn't behind a hard paywall. We did a FOIA to understand what exactly happened with NC State's ejection from the MCWS due to COVID and did an exclusive on that, and that drove A LOT of direct conversions. (Plenty of Andrew Carter stuff we put behind the paywall drove subscriptions, because Andrew's articles are just that good.)

The one high risk/high reward strategy I tried in Raleigh with subscriptions was with columnist Luke DeCock. His columns were great every time, but the page view traffic was rarely as good as it should have been. Despite the traffic numbers (it wasn't bad, it just less than we felt it should be), we decided to put every column Luke wrote behind the hard paywall, and the strategy instantly worked. He quickly became one of the 2-3 top subscription drivers in Raleigh, which was amazing, especially with the quality of news reporting that newsroom did that drove plenty of subscriptions.

There's a long-winded answer, full of lazy writing and typos from my phone, I'm sure, but it's helped me kill time waiting for my oil change.

 

[LanceyHoward]

Two main metrics were LTCs (Led To Conversion — stories that didn't necessarily convert a subscriber directly but was one of the 3-5 free articles a user read before they hit the paywall and paid) and direct conversions (article they hit the paywall and became a subscriber for).

Preps content was always high traffic and there wasn't competition in the Charlotte market. Decided one fall (either 2020 or 2021) to put all preps content behind the paywall and almost everything started driving multiple subscriptions. Charlotte is fortunate to have Langston Wertz Jr. dedicated to covering high schools — he's done it since the 80s and loves it — and that always helped with the page view numbers, but the thing about high school sports even in markets where page views aren't great for preps (which is a lot of them), that coverage can still drive subscriptions if you cover it well — most likely more subscriptions than any other best, regardless of desk. And there are always new parents and extended family members every fall who are willing to subscribe to read about their kid's team.

For Panthers game day coverage, we'd usually only hard paywall the in-depth analysis of a game that would go online a couple of hours after the game's end. (We'd convert the instant analysis/gamer URL that went live as soon as the game ended and drove the most page views because of Google to the in-depth paywall piece because it already held that domain authority.) That article would usually get 3-15 direct conversions each week. More subscriptions early and late in the season. When the team is bad, mid season traffic and conversions are harder. All other game day content would usually at least generate a couple of LTCs.

We made our NASCAR coverage model focused on page views rather than subscriptions. It drove a ton of traffic (like 250,000 - 1 million PVs in a day for Daytona 500 coverage alone, with the 2020 race and wreck driving the most). We traveled to Daytona, Talladega and any regular season race within a 6-hour drive, plus all playoff races, and it was worth it for the traffic we got from articles, but we learned pretty quickly our online audience for NASCAR was national and not local, so getting paid subscribers was much more difficult. We still did a lot of in-depth features and found quality exclusives and analyses, and they'd again get good page view numbers but not conversions.

Raleigh was a more unique situation with colleges coverage and how saturated the market was for UNC and NC State. We had solid page view numbers across the board, and steady subscriber numbers, but no beat stood out as the one that could always get the most conversions. (I should add that when Natalie Pierre arrived to focus on leading our colleges coverage, everything got better across the board.) There were also outliers, of course. When we were able to break Hubert Davis becoming the next UNC coach, traffic went nuts. Plenty of subscribers came, even if it wasn't behind a hard paywall. We did a FOIA to understand what exactly happened with NC State's ejection from the MCWS due to COVID and did an exclusive on that, and that drove A LOT of direct conversions. (Plenty of Andrew Carter stuff we put behind the paywall drove subscriptions, because Andrew's articles are just that good.)

The one high risk/high reward strategy I tried in Raleigh with subscriptions was with columnist Luke DeCock. His columns were great every time, but the page view traffic was rarely as good as it should have been. Despite the traffic numbers (it wasn't bad, it just less than we felt it should be), we decided to put every column Luke wrote behind the hard paywall, and the strategy instantly worked. He quickly became one of the 2-3 top subscription drivers in Raleigh, which was amazing, especially with the quality of news reporting that newsroom did that drove plenty of subscriptions.

There's a long-winded answer, full of lazy writing and typos from my phone, I'm sure, but it's helped me kill time waiting for my oil change.

Thank you for such a detailed response.

 

[Inky_Wretch]

Could AI cover city hall?

I ask because our Gannett paper no longer has a single employee here. The last one left in mid-August. The last local copy came out on Aug. 17. The University of Arkansas beat writer is now getting front page bylines though - just a day late thanks to early deadlines.

 

[JM22720]

Two main metrics were LTCs (Led To Conversion — stories that didn't necessarily convert a subscriber directly but was one of the 3-5 free articles a user read before they hit the paywall and paid) and direct conversions (article they hit the paywall and became a subscriber for).

Preps content was always high traffic and there wasn't competition in the Charlotte market. Decided one fall (either 2020 or 2021) to put all preps content behind the paywall and almost everything started driving multiple subscriptions. Charlotte is fortunate to have Langston Wertz Jr. dedicated to covering high schools — he's done it since the 80s and loves it — and that always helped with the page view numbers, but the thing about high school sports even in markets where page views aren't great for preps (which is a lot of them), that coverage can still drive subscriptions if you cover it well — most likely more subscriptions than any other best, regardless of desk. And there are always new parents and extended family members every fall who are willing to subscribe to read about their kid's team.

For Panthers game day coverage, we'd usually only hard paywall the in-depth analysis of a game that would go online a couple of hours after the game's end. (We'd convert the instant analysis/gamer URL that went live as soon as the game ended and drove the most page views because of Google to the in-depth paywall piece because it already held that domain authority.) That article would usually get 3-15 direct conversions each week. More subscriptions early and late in the season. When the team is bad, mid season traffic and conversions are harder. All other game day content would usually at least generate a couple of LTCs.

We made our NASCAR coverage model focused on page views rather than subscriptions. It drove a ton of traffic (like 250,000 - 1 million PVs in a day for Daytona 500 coverage alone, with the 2020 race and wreck driving the most). We traveled to Daytona, Talladega and any regular season race within a 6-hour drive, plus all playoff races, and it was worth it for the traffic we got from articles, but we learned pretty quickly our online audience for NASCAR was national and not local, so getting paid subscribers was much more difficult. We still did a lot of in-depth features and found quality exclusives and analyses, and they'd again get good page view numbers but not conversions.

Raleigh was a more unique situation with colleges coverage and how saturated the market was for UNC and NC State. We had solid page view numbers across the board, and steady subscriber numbers, but no beat stood out as the one that could always get the most conversions. (I should add that when Natalie Pierre arrived to focus on leading our colleges coverage, everything got better across the board.) There were also outliers, of course. When we were able to break Hubert Davis becoming the next UNC coach, traffic went nuts. Plenty of subscribers came, even if it wasn't behind a hard paywall. We did a FOIA to understand what exactly happened with NC State's ejection from the MCWS due to COVID and did an exclusive on that, and that drove A LOT of direct conversions. (Plenty of Andrew Carter stuff we put behind the paywall drove subscriptions, because Andrew's articles are just that good.)

The one high risk/high reward strategy I tried in Raleigh with subscriptions was with columnist Luke DeCock. His columns were great every time, but the page view traffic was rarely as good as it should have been. Despite the traffic numbers (it wasn't bad, it just less than we felt it should be), we decided to put every column Luke wrote behind the hard paywall, and the strategy instantly worked. He quickly became one of the 2-3 top subscription drivers in Raleigh, which was amazing, especially with the quality of news reporting that newsroom did that drove plenty of subscriptions.

There's a long-winded answer, full of lazy writing and typos from my phone, I'm sure, but it's helped me kill time waiting for my oil change.

darn, I left McClatchy around 2 years ago after covering HS and college sports (won't go into any more detail so not to out myself) and this post brought back soooo many memories.

 

[dixiehack]

Could AI cover city hall?

I'm convinced half the government here is run by AI. Call it ChatGOP

 

[Hermes]

fork AI.

He led the country in rebounding at Ohio. He was a father and business owner. AI called him "useless."

https://fox8.com/news/useless-at-42...OduiXkZmm8AMXYXsRQwtwqes49nfaBEwYo2rK5wxQcARQ

 

[Inky_Wretch]

Arkansas had noon kick. Of course, the Times Record also has a noon deadline so the out-of-state design hub and editors can ship it before larger papers.

Currently, the only in-state employee is their Arkansas beat writer. So there is zero local coverage of the school board or city board of directors. Maybe AI could handle that?

The last local copy filed at SWTimes.com was Aug. 17. They have not filled the editor/reporter/photographer/copy editor/one-person band position since the last one left in August. Aside from Razorbacks copy, the website and paper is filled with wire stories and content from other Gannett sources.

Our metro population is 110,000. Our MSA is 290,000.

 

[I Should Coco]

The last local copy filed at SWTimes.com was Aug. 17. They have not filled the editor/reporter/photographer/copy editor/one-person band position since the last one left in August. Aside from Razorbacks copy, the website and paper is filled with wire stories and content from other Gannett sources.

Our metro population is 110,000. Our MSA is 290,000.

That sucks, Inky ... but your area is far from alone.

In southeast Washington state, the Tri-City Herald (more than 300,000 population in Kennewick, Richland, Pasco) no longer has a sports editor. It's a McClatchy daily newspaper.

 

[Inky_Wretch]

That sucks, Inky ... but your area is far from alone.

In southeast Washington state, the Tri-City Herald (more than 300,000 population in Kennewick, Richland, Pasco) no longer has a sports editor. It's a McClatchy daily newspaper.

How many are left in that newsroom? Because the SWTR is at zero right now (not counting the Razorbacks beat writer who lives an hour north in Fayetteville).

 

[Slacker]

How many are left in that newsroom? Because the SWTR is at zero right now (not counting the Razorbacks beat writer who lives an hour north in Fayetteville).

Man, that really sucks, and especially for a market of that size.
Are the locals making any noise about that, or do they not care anymore, or what?