May 1, 2020
Members Only

Is There a Way to Build a Revenue Server?

Whether the media company is hundreds of people or just a few, once there are additional revenue streams, silos start to form—especially if you have two separate teams managing them.

In their simplest form, media companies tend to have two revenue streams: advertising and subscriptions. Which is more important? At many older media companies, this is certainly a debate.

In the Digiday Plus Talks a couple weeks ago, Tribune Publishing’s Mark Campbell said:

It will reshape the age-old tension between volume and revenue. I’ve been in circulation for like 25 years and forever we’ve been talking about, you know, is circulation revenue driver or is it a volume driver that drives ad revenue. That balance is something we [circulators] have always been trying to strike, so as we’re thinking about future proofing our businesses, I imagine that tension will be front and center because we’re going to know that, remember that, subscription revenue is an annuity that isn’t as ephemeral or as oscillating as ad revenue.

It’s a basic debate, but it’s one that can have a serious impact on the business. The ad team is going to, naturally, fight against the subscriptions team because they, in theory, have conflicting objectives. The ad team wants as many pageviews as possible whereas the subscriptions team wants the paywall to appear as soon as possible.

Often times, the strategy would be decided by whoever was in charge. Did the “boss” come up on the ad sales side or the subscription side? That would dictate the answer to that question.

However, what’s best for the business is what drives the most revenue. Sometimes that’s going to be advertising and other times that’s going to be subscriptions. But how do you decide? And, more importantly, how do you unify your marketing efforts to ensure that you’re chasing the most dollars versus specific silo dollars?

Enter the Revenue Server

After last week’s essay about experimenting, one reader shared some thoughts on maximizing revenue depending on what type of user they are.

Effectively, different types of users are likely to engage with your content in different ways and, just as importantly, generate revenue in various amounts. Some users might be prone to subscribe while others might never.

I summed up the discussion as such:

You’re basically talking about being able to holistically understand what a user is worth to you and unifying your various department’s marketing exercises against a single user, right? Rather than the sub team thinking about subs whereas ad ops is thinking about how to increase ad impressions versus the third team focused on getting people to register for an event? A single view of a user so you’re maximizing how you monetize to them?

What if there was a system that would allow us to get that clean picture? It’s not an original phrase, but we’re going to call this the “revenue server.” Unlike an ad server, which is trying to maximize ad revenue, or your registration software, which is trying to get more conversions, this is a central system that gives a single view on the user so you know whether you’re truly monetizing them most efficiently.

In other words, it’s a user monetization system versus a tactical monetization system. It’s trying to identify what types of users you have and the right ways to monetize them.

I should make one giant caveat… I am not an engineer. To build out the system I’m about to describe would require considerable investment. If I had to guess, only the largest media companies will be able to do this. I could also see Vox and The Washington Post do this as part of their ever expanding suite of products. Right now, they offer CMS and ad technology. But as time goes on, will they start to offer holistic revenue technology that is better able to monetize users across a suite of diversified revenue streams?

I had a conversation with a friend of mine who has a ton of experience in the ad world. We discussed how this could be built and the problem is that there’s no unifying data point about these people that can make it easy to track. True, we can drop cookies, but that’s messy and doesn’t take into consideration that people are on their phones and computers at different times. Not to mention, with 3rd party cookies going away, I’m not even 100% sure analytics providers are going to be able to accurately track people. Say hello to a big boost in unique users.

But the one thing that every one of these users has is an email address. What’s unique about the email address is that, whether I’m on my phone or my computer, the email address stays constant. By using the email address as the unifying data point, it becomes easier to track that user and understand their behavior.

I haven’t hidden my belief that all publishers need to be working hard to start collecting 1st party data on their visitors. Email address is fundamental. That’s why The New York Times is gating its content after one article, asking users to create a free account. Honestly, everyone should be doing this right now.

Once you have that email address, you’re better able to keep track of what they’re doing even if you never want to actually launch a subscription product of any sort. A basic, free sign up can give you the ability to really boost your understanding of how to monetize.

What then?

The revenue server is effectively a CRM. But rather than simply tracking whether someone opened an email, clicked a link, consumed an article, or what have you, it’s tracking how the user is monetized. Here are examples of the types of data you’d be collecting:

  • Ads: Did they see ads? How many ads did they see? Were they direct deals or programmatic? What’s the eCPM of those deals? For my B2B readers, did they download a sponsored white paper?
  • Commerce: Have they engaged with a commerce ad? Did they ever visit your store? Did they add something to the cart? Did they buy something? How much did they buy?
  • Subscriptions: Have they signed up? Have they churned? Have they ever even engaged with the subscription page? Have they churned and signed up again?

As we collect this data and start to better understand what type of user we have from a monetization perspective, we can start making informed decisions around the actions we take. Let’s go through a few scenarios as examples.

User that doesn’t buy

We have a user that comes to the site, consumes the content, but never finds their way to the commerce sections of the site. Whether that’s through advertising or a big CTA, they avoid it.

You can do two things here. Try harder to get them to go shopping or recognize that the user is not a good target for commerce CTAs. In place of these commerce CTAs, let’s put a higher impact advertisement. (As a quick aside, I don’t mean a popup.)

As that user comes back to the site, you can customize their experience to be less heavy on commerce and more on advertising. It’s more money for you and less wasted impressions promoting something they don’t want.

User that blocks ads

Perhaps the user bounces because of too many ads. Or, they come to your site and immediately have an ad blocker in place. Both are probable, especially if we look at some of these sites with an obscene number of ads.

Here you may decide to go incredibly light on the advertisements. One or two at most. But you really ramp up the subscription calls to actions.

One option could be that you make a trade with the user with the ad block, especially if they’ve come to your site multiple times. Offer an ad-light experience in exchange for them signing up for a newsletter or creating an account.

Why do this?

Rather than promoting everything to everyone, we start to get a bit smarter about promoting the right things to the right people. We try to force too much on our users. On a single page, there might be banner ads, commerce promotions, subscription CTAs and various other opportunities. Digital media sites are typically very messy.

Instead, we can create a more efficient page that is focused on what the individual is most likely to engage with, providing us the opportunity to maximize our revenue down to the user level.

Now, let’s say we have this system in place. Here is where things get a bit interesting.

When a user does log in, you’ll call the revenue server to see what the user classification is. The revenue server’s data will dictate what the monetization layout is. It’s similar to how an ad server works to deliver the right ads based on targeting, frequency caps, etc., but instead, it’s pulling in all your monetization methods.

The good thing about this system is that it wouldn’t be static. If a user is classified as someone likely to subscribe and then you get them to subscribe, they’d move into a new classification. That classification would have different rules on how you want to monetize them.

The other benefit of this is that it dictates which of your users you should be reengaging through ad buying.

Assume you have two teams. One is focused on commerce revenue and the other is focused on subscriptions. They both have marketing budgets. A user pops up that subscribes, but hasn’t read much recently (hinting at a possible churn event), and has also added things to their shopping cart on your site and not finished the purchase. Which wins?

Normally, both teams are going to market to them at the same time, which we don’t want. It’s spending double to get to one user with competing objectives. With clear visibility into the user, the right creative can be delivered—perhaps in this case, promoting the subscription.

One way or another, your teams are no longer competing for the user. They’re unified toward the goal of growing revenue.

Like I said up above… This is a complex system to build. Even if you can’t build the above system, you should be looking at this time as an opportunity to understand more how your users are being monetized. You might find that you are promoting things that are, relative to other opportunities, really under-performing.

I see this a lot with paid ads and house ads. Let’s say we have a single ad unit on the site with two opportunities. The house ad will promote an event. That event costs $1,000 to attend with the goal of 1,000 people. The paid ad is a campaign with a $10 CPM.

Which should win? I’ve heard a lot of people say the event. That’s the “bigger” revenue driver. However, there are two things we don’t know here.

First, did the user already buy a ticket to the event? If so, promoting it still is just a waste of space.

Two, what is the cost of that house ad? When dealing with a limited amount of inventory, we have to assign a cost to the house ad impression. Here’s what I mean…

Let’s say we run 100,000 impressions on the ad and get a single person to buy a ticket. That’s $1,000 in revenue for 100,000 impressions, which is a $10 CPM. Good, we’ve got the baseline now. Now, let’s say we run 500,000 impressions and get 4 registrants. That’s $4,000 in revenue for 500,000 impressions, which is an $8 CPM.

Which would you prefer? An $8 CPM or a $10 CPM?

When we start to think about our revenue as a total pie versus individual products, we start to better understand how we should be monetizing our users. And if we can start to build our feedback loops, we can start making even smarter decisions. A user has already bought a ticket? Then never show the house ad again. The user is already a subscriber to a newsletter? Stop wasting real estate on that promotion. The list goes on.

Join A Media Operator

Consider becoming a premium member so you can receive even more analysis and insights.