When data misleads

Following the data isn't always the answer

I’m on my way to Boston tomorrow for Hubspot’s Inbound conference. The last time I spoke at it was the day after Trump’s election. Talk about a weird vibe. If you’re there, come see my session on “the future of media.” It’s at 11am on Friday and thankfully we’re not up against Derek Jeter or Reese Witherspoon. Somewhat surprised an Obama isn’t speaking.

Today, I wanted to continue the conversation on editorial scorecards and how blindly following data can create misaligned incentives. First up, a message from Omeda.

6 actionable ways to manage your audience growth costs

Your audience and subscription numbers are flat and you’re feeling the heat. Time for a rebrand, a new podcast and a new Head of AI just for good measure.

All good, except now your budget’s out of control, you have less money to fund all those new ideas in your head… oh, and your numbers are still flat.

Sound familiar? Read on for seven ways to keep your audience costs under control — without sacrificing your growth goals.

When follow the data leads you astray

Last week, I wrote about editorial scorecards. My argument was that, as we enter a more-with-less era, many publishers will turn to counterproductive ways to wring efficiencies from their editorial operations through misguided efforts to measure editorial output and assign it a value.

Everyone is reflexively in favor of measurement and, even better, accountability, particularly when it’s happening to others. But I believe most of these efforts at editorial scorecards end up just creating misaligned incentives.

The influx of algorithms has flattened most digital content. Business models rewarded “eyeballs” and fleeting impressions versus depth and loyalty. In my podcast with Puck’s Jon Kelly, a veteran of Vanity Fair, he noted how many legacy publications have struggled to create credible subscription programs because they’ve been too concerned any revenue created by asking people to pay will not offset lost opportunities in advertising.

One veteran of a magazine brand wrote me: “The challenge we ran into was how to build those deeper connections. We also got distracted for a while by the Facebook games of the 20-teens, including a brief detour into on-platform videos.”

It puts many publishers in an awkward position when more focus is being placed on so-called made-for-advertising websites that clog up search results pages. A recent study by the Association of National Advertisers found that these SEO chop shops, who are master optimizers, can suck up 23 percent of open programmatic ad budgets. Much of this stems from the endless quest to optimize to a spreadsheet rather than the more slippery concept of quality.

“Advertisers prioritize cost over value, sometimes to their own detriment. They chase cheap CPMs. The primary incentive driving programmatic media buying behavior is cost—getting the most impressions for every dollar. On its face, this might seem like a sensible incentive. However, in the media world not all inventory and not all impressions are equal. Common sense should tell buyers that not all ‘cheap’ inventory is ‘quality’ inventory.”

Common sense doesn’t fit in a cell on the spreadsheet.

On My First Million, Colin and Samir’s Samir Choudtry spoke of how creating for YouTube means having a scoreboard on each piece of content you create. Inevitably and inexorably, it pulls you to optimize to big numbers.

“They’d rather have 5 million views than 4.5m views. When I look at it, I’d rather have 100,000 views or 100,000 dedicated fans who will walk away knowing something about me than rip all that out to make sure they get to the end of the video. The business of YouTube is predicated on did they click and did they watch to the end. That can give you the wrong signals.”

AI is poised to make all of this worse, but also presents an opportunity. Gaming of algorithms and growth hacks will become automated to such a degree that I believe humans will reward humans who go off piste. Maybe the internet will get weird again. AI is going to create this homogenous nature to content,” Samir said. “Everything is going to feel the same. That’s going to lead us to create something unique and different.”

Distinctive brands will be more valuable, and there are so many things to building brands that do not show up on spreadsheets. The risk is always that following the data is the safest course. My fellow podcaster Alex Schleifer often returns to the endless A/B testing that many tech companies use to torture design and product, effectively abdicating their responsibility to have a point of view.

That said, I am not ready to lead a pastoral life just yet, so I understand the utility of using data to understand how to better serve audiences and, if you’re looking at the right information, to build deeper relationships. One big media company editor noted that the use of data in their shop has moved on from the era of “blinded/single-minded data you refer to, where analysis is low quality and clicks at all cost is the goal.” but

“I’m happy to say that’s not the way we’re looking at it now. We’re using data to better understand our different audiences, and to look closely at what each of them wants, so as to better move them down the funnel, and better understand who our core users are. Is that not a noble use of data—to better understand and super-serve our super users? I wish I had had the kind of direction my editors get when I was in their shoes. It would have made my job much easier, as it would have made my decision making clearer. It doesn’t need to be a one-way relationship, editor to reader. Getting that feedback in terms of how people interact with the output feels like helpful feedback to me!”

I always appreciate feedback, so please continue to send it in, good and bad. I feel like newsletters are best when they are a form of a conversation. Hit reply or send me an email directly to bmorrissey@therebooting.com.

Journalism's AI balancing act

Safeguarding democracy from false narratives is hard work, sustainable quality journalism relies on robust revenue and people-focused experiences. While AI introduces opportunities for advertising and subscription growth, not all pathways are lasting and aligned with the greater good. House of Kaizen's people-first revenue optimization steadies the course for news publishers, helping journalism stay upright.

Let’s make news more sustainable, together.


The open web has been imperfect, and rife with its own misaligned incentives, but it’s on balance provided an amazing array of mostly freely available content and exploded the number of people who can find audiences. The shifts that AI will make to search – it’s just a matter of timing, in my view – is somewhat underappreciated. Big name publishers are lining up to block AI crawlers from sucking their information into their large-language learning models. Lawsuits are coming if big dollar bills aren’t struck. This leaves Google in a pickle. The leverage publishers have isn’t just their information – much of it isn’t as valuable as they think – but it’s the dent to Google’s entire premise of “organizing the world’s information.” (The Guardian)

Walter Isaacson’s new Elon Musk book will provide plenty of fodder for his fan boys and detractors. In a new episode of People vs Algorithms we recorded yesterday, we discussed our need for heroes and villains, and the thorny issue of the “golden age for assholes.” The WSJ excerpt of the book shows that Musk is the prototypical asshole boss, yet his record of success doesn’t neatly fit into our narratives that the “good guy” wins in the end. More tomorrow. (WSJ)

I’m fascinated by how Barstool will fare as a newly independent company. It caught lightening in a bottle with the legalization of sports gambling. Many sports media have had to awkwardly retrofit their programming to push gambling. (Do not tell me that business models don’t dictate content direction. I cannot find an NFL podcast that doesn’t talk about parlays, teases and other nonsense.) Barstool had a nice few years as a trust fund kid of Penn Entertainment, but now it’s faced with figuring out a sustainable business model without the loss-making subsidies. Pre-Penn it had some success selling its own merch and putting on live events. Its ad business will always be challenged by skittish brands. That pivot to reality starts with a 25% cut to its staff. (New York Post)

My suspicion is many publishers got over their skis with commerce and got too far from their core strengths. Adweek’s Mark Stenberg has a piece on Hearst’s Cosmopolitan killing its CosmoTrips booking service after just a year. Part of this is new leadership not being invested in something they didn’t do. But more publishers will look at focus on what they’re best at, and that’s often branded content. Content and commerce is often better on whiteboards than in reality. (Adweek)

The unbundling of Vice has put I-D Magazine up for sale, and one of the prospective buyers is Karlie Kloss. Puck’s Lauren Sherman broke the news and adds a tantalizing tidbit that a further unbundling could result in Refinery29 back in the hands of founders, Philippe von Bories and Justin Stefano. I’m wholly in favor of founders of these companies – Shane Smith at Vice, Rich Antoniello at Complex, hell even Pete Cashmore at Mashable – returning for another shot. (Sponsored – Puck))*

* This is something new I’m trying with Puck, a publication I like quite a bit. We’re seeing if TRB can drive subscriptions. I’m one. Check it out.


Thanks for reading. Please send feedback by hitting reply or sending me a note at bmorrissey@therebooting.com.

Sign up for invites to The Rebooting Dinner Series. We have upcoming private executive dinners focused on direct-sold advertising, making money from video, subscriptions and a pair of Advertising Week dinners for both ad buyers and publishing executives.