Why Steinberg Sold Cheddar

And how removing ads improved ad revenue

Before we get started, I wanted to thank everyone for sharing my last issue about The Athletic’s valuation. So many new people have subscribed, so welcome to all the new people! If you have any thoughts about anything I’ve ever written, don’t hesitate to hit reply. I’ll respond to everyone.

New here?


I had the pleasure of attending The Information’s Media Bootcamp last Friday. For those that don’t know, The Information’s founder, Jessica Lessin, is passionate about media businesses overall, but more specifically, subscription media companies. I’ve been a subscriber to The Information for about a year now and one of the things Jessica said really struck me.

“You can’t put a paywall on a pig.”

I imagine every single media company subscribed to this newsletter has either considered a paid subscription or is actively rolling one out as we speak. But introducing a paywall isn’t just something you can do if your content is only okay. It has to be exceptional.

In a recent interview, she summed it up:

As journalists, we think our story is so much better than this other story, but in the eyes of a reader who is maybe not paying a ton of attention, it may seem similar. I always say, you really need to go for the 10x and focus on the things that are 10 times better or different than what other people are doing.

If you’re going to start charging for people, whatever you create has to be exceptional. Her advice for any media companies was to create a franchise. For The Information, that’s their org charts. What could your media company create that is exclusive to only paying subscribers that no one else can—or will—provide. That is your 10x better.

If you’re seeing signs of difficulty with your subscription, really break down what the user is getting for signing up. Are the stories the same things that every other media brand is writing? Or is editorial providing that special something that differentiates your reporting from someone else’s? We’re seeing media companies succeeding with their subscription products, so analyze what is working for them.

But in this issue, I wanted to spend a little time breaking down why Steinberg decided to sell.

When Facebook was preparing to go public, everyone was so excited because Mark Zuckerberg had had the foresight to turn down so many acquisition attempts. He believed that the company was worth more, so he didn’t sell. This has turned into the cult of the founder, which has naturally gotten rather toxic, but it’s still interesting.

Steinberg had none of that. “Cheddar had to be sold,” he said at the event on Friday, referencing the $200 million sale to Altice that closed back in June. And his rationale was really interesting. I’ll summarize below.

Imagine an early stage pharmaceutical company. You’re constantly looking to find a drug that’s going to work. And suddenly you happen upon something. Clinical trials show things are going really well. You’ve got a very small group of sales people that are selling it and things are looking good.

This is where Cheddar was when it sold. They knew they had a product that people liked and they knew there was a business there. But, they had a choice.

That pharmaceutical company now has a choice. Are they going to build the factories required to make the drugs at scale? Are they going to hire the necessary sales team and supporting marketing? Do they have the resources to reach the scale required to make this product a success?

Steinberg summed it up: “We needed an ever-increasing sales force, we needed increasing production and hours and people to do deals… We needed scale.” And so, rather than try to build that scale, which would have cost a lot of money, he decided to sell and continue building the company with the added resources of Altice.

He’s not the only one to have made this decision. Last Tuesday, I talked about a recent media acquisition where the majority of Industry Dive was purchased, but the founding team was sticking around. They could have ventured forward, using their balance sheet to grow, or they could partner with a firm that could provide resources to help them achieve their goals.

But I would add a caveat to Steinberg’s point. General media companies absolutely need huge scale to survive. Vice, Buzzfeed, Vox, Refinery29, Group Nine Media, and all the others are going to need to continue growing considerably if they are going to be able to operationalize a profitable media brand. And even if they do get to profitability, they raised at such extreme valuations that I question whether they’ll ever reach those lofty goals.

One thing I overheard Lessin say at the event was that media companies don’t tend to like buying money-losing projects. It’s just not in their DNA. So, all of these digital media companies that are losing money are not going to be able to command anywhere near the revenue-multiples that they will need to reach those valuations.

Equally, it’s unlikely that they’re going to be able to buy each other unless they all accept that the valuations are too great. I referenced this piece by The Information last week, but for the new subscribers, it’s worth looking at.

The next 12-24 months are going to be very interesting for these general media businesses. They’ve raised so much money at such lofty valuations, but are now running out of steam. I overheard one attendee ask another if they were looking at doing any acquisitions and the response was telling: “We’re always looking. But the price will be better next year.”

While one or two of these might survive—my money is on Vox—the rest are going to ultimately get acquired for a lot less than they anticipated. If they buy each other, perhaps then they’ll have the required scale to really compete. Otherwise, there will be no new independent media companies.

You’ll notice I keep referencing “general media businesses.” By now, you’ll know I’m rather passionate about niche media. I think there’s a real opportunity here to build a lasting, forever business whereby founders grow their media brands methodically. Perhaps they raise a little money, just to get going, and then invest in themselves with their cash flow. Over time, this reinvested cash does compound and I’m sure many of the niche media brands I’ve referenced here are worth tens of millions of dollars.

But this approach requires patience. I think many of these new general media businesses lost track of the fact that media is a slow, methodical business. There was so much success around software companies reaching “unicorn” status that media brands wanted a taste of that too. But when running drastically different businesses, it’s best not to compare.

When removing ads grows revenue

Digiday did a good story on GiveMeSport, a sports publisher. In it, they talked about how they completely redid the website with a belief that an improved user experience would result in more revenue.

Before August, digital sports publisher GiveMeSport had struggled with a page-load speed problem. Its 5-year-old site had become cluttered with ads and heavy code causing its user experience to suffer and individual page loads to take up to 20 seconds, according to the publisher.

To tackle this, the publisher scrapped its former site and started afresh with leaner progressive web-app features like push notifications and the ability to work offline. Meanwhile, on the ad side, GMS flew in the face of typical viewability-target pressures and culled its ad count from an average 11 per page to four.

The outcome?

Page speed as drastically improved and direct traffic to the site has increased by 63% year over year. Obviously, this makes sense. Here’s an image that Google put out explaining how important page speed is to your audience.

Getting rid of ads to improve page speed is a tough pill to swallow, though. If you have four ads and you add a fifth, shouldn’t the eRPM go up? With programmatic ads, I have seen this occur. Each unit’s CPM might drop a little, but that fifth unit improves the overall revenue.

But, the reality is, if your experience is garbage, no one is going to hang around on your site to see those ads.

So, what else can be done? These are a few recommendations to improve experience and revenue…

  1. Audit every demand partner that you have in the stack. It’s not uncommon for publishers to have 15-20 of them all bidding for each impression on the site. What you’re looking for here are three things:

    • How much revenue has that demand partner generated over the past 3 months?

    • What percentage of total bids have they won?

    • How long is their call taking to load?

We do this because we’re looking to cut the number of calls on a page. Many demand partners exist without ever winning or generating money, yet they get equal opportunity and take up page load time. By removing them, you increase speed.

This also shouldn’t be a one-time process. If you have a demand partner that is winning a large percentage of the bids, but it takes a ridiculously long time for their call to load, try removing them anyway. See how much revenue the other partners can make up. At least with programmatic, there’s always another partner who has an impression waiting to fill.

  1. Lazy load everything on your site, especially the advertisements.

I say everything because images take some time to load. There’s no reason to have the 10th article load immediately. As the user scrolls, you can then start loading features that are below the screen. This will drastically improve the load time of your site.

On the ad front, it’s the same way. Especially on mobile, you should try to limit the number of ads on screen to one. As the user begins to scroll, you can start filling in other ads. Obviously we don’t want the content to be jumping around, so test your lazy loading so that it minimizes any damage to the user experience.

  1. Be smart with your ad refresh

Everyone does ad refresh with their programmatic ads. And if they say they don’t, there’s a good chance they’re not telling the complete truth. But that doesn’t mean that every ad should be refreshing all the time. Limit the refresh to only ads that are in view. This will reduce the number of active calls and also improve your viewability metrics.

If you can’t limit your ad refresh to just the units on screen, then do a test. Turn ad refresh off entirely, see the effect it has on your revenue, and then turn it back on. You may find a reduction in ad impressions when you turn it off, but still see an increase in revenue because ads out of view are not loading.

I keep mentioning viewability, so let’s spend a minute on that. It’s hard to imagine, but advertisers actually like their ads to be seen. And they’ll pay more to publishers that make their ads visible.

Google has an in-depth guide on Active View reporting, so it’s worth giving it a read. But, here is how you’d break it down:

Two important factors are (a) what percentage of an ad appears in a viewable space on screen and (b) how long that portion of the ad remains visible.

By and large, viewability appears to be measured at the site level rather than the ad-unit level. One of the moderator’s in the AdOps reddit channel summed up the benefits of improving viewability rather nicely:

honestly you need to be paying attention to your overall viewability for all auctionable sizes. using the adx active view rate as the proxy. we found a pretty big uptick when you move past 50% and then another big one right at 75%. we found paying attention to specific ad units or platforms didn't move the needle as much as fixing the overall.

People are inclined to gravitate to clean looking numbers, hence why this person found increases to revenue at the 50% and 75% mark. By improving viewability, you’re going to improve the amount of money per impression.

By reducing the number of partners in the stack, lazy loading everything, and only refreshing ads that are in frame, things will load faster and your viewability will improve, both of which should result in growth to your business.


And that wraps up this Tuesday’s issue of A Media Operator. Thanks to everyone for sharing last Friday’s. If you found this helpful, please consider sharing it with your colleagues. If you have thoughts, please reply. And if you’d like me to include those thoughts in the next issue, please let me know that also. Thanks for reading and see you on Friday!