October 24, 2020
Members Only

Should More of Us Be Thinking About Data as a Product?

Many of the sites I look at on the b2b side are primarily content-driven businesses. The team structure is straight forward. On one side, you’ve got the editorial team. On the other, the business side.

This is a straightforward media business.

When I get excited about the next decade of niche media companies, this is what I think about. Hundreds and thousands of new media companies will launch that are focused on very specific topics on both the consumer and business side.

However, I’ve been thinking a lot about data as a product. In this case, I’m not talking about user data, but instead, I’m talking about data specific to the particular industry in question. Readers of a news organization could also use data to help make decisions. Rather than them going off to find it in other places, why not present a single destination for that?

There are, of course, a few reasons not to venture down this path, which I’ll start by covering. Nevertheless, it’s an interesting opportunity that is worth diving a little deeper into.

But first…

Reasons to avoid a data play

There are two major reasons why it would make sense to avoid a data play and they are related.

The first has to do with the company’s DNA. In that linked essay, I wrote:

Who you are as a business matters. It explains a lot about the approaches taken to accomplish things and what the priorities are.

Each business has a DNA. When a business is small, that DNA can change (a term in genetics known as a mutation). As a business grows larger, though, it becomes increasingly difficult for that DNA to change. This is why we find large organizations often get disrupted by little ones.

In media, there’s a specific DNA associated with the business. It’s very much a people business. Day in and day out, your journalists are talking to sources, reporting on what’s happening and filing stories. Each day is a new day and you have to build the product all over again. It also doesn’t really scale without additional inputs.

Compare that to a technology business. For Twitter to add another million accounts, it doesn’t need to hire a linear number of employees. That means the business thinks about things in a different way.

The differences are vast between a media company and a software company. How can a media company that has historically focused on creating content then shift and create broader data products? Consider this a little more…

With sustainable media, it feels very much hand to mouth. The flywheel is straight forward. We create content, build an audience, sell against that audience so we can then invest into creating more content.

For many software companies—at least the ones that get the most press—the goal is to generate a return in the future. There’s plenty of money available to invest in growing the product with a hope that revenue will follow.

The other major reason to avoid data plays is that it’s expensive. You need to hire software engineers, data scientists, product managers, etc. That sort of fixed cost really adds up.

If you’re not careful, you could have a large technical team on staff building out data products when the DNA of the company really isn’t suited for that. Before you know it, you find yourself hopping onto the treadmill of fundraising.

This is why I have looked at media as a very circular business: content to audience to revenue to content all over again. The formula is simple and I find it works if operators stick to it.

Why look at data

And yet… There are a variety of reasons that a smart data strategy can be incredibly compelling, powerful for the business and financially rewarding.

Data is a great tool for explanation

For some people in your audience, raw data is the name of the game. They want as much of it as humanly possible, want to download it into excel spreadsheets and will pay obscene amounts of money for that.

Those people tend to work for hedge funds and are gobbling up data with the goal of earning just a few more basis points in returns. Fine business to be in, but it doesn’t feed into what makes media companies unique.

Our business, by and large, is explaining what’s going on, what it means and why it matters.

This creates its own interesting flywheel that can be incredibly powerful. When you have data coming out of your own operation, it becomes fodder for potential stories on the website. Let’s use an example…

Imagine you’re running an insurance site. And you have the average car insurance premiums in New York running through the dashboard. If there is, for some reason, a sudden spike in premiums, your editorial team could write a story about that. Rather than crediting some other data source, your editorial team is able to reference your own data.

This is powerful for two reasons…

First, you become a source rather than a writer of other sources. As I wrote in my piece about SEO tent poles, providing data to journalists is a great way to get other people writing about you, which creates a new audience development tactic.

Second, and more importantly, you’re showing your prospective customers how to think about the data. Like I said up top, what makes media unique is that we explain things. Therefore, using your data as a reference and then explaining its context becomes an added benefit to the prospective buyer.

Reduced CACs

A content/data mixture also reduces your customer acquisition costs (CACs) considerably… Users are coming to your site to read content. They see one of the articles your team has written about data found in your platform. That’s marketing for your offering.

Therefore, the CACs are effectively the cost to create that content divided by the number of customers you get to sign up for the data, right?

That’s where media is unique again…

When I was talking to Craig Fuller (who unintentionally inspired this essay) on the podcast a while back, he talked about the negative CACs. This is exactly how it sounds. Rather than spending money to acquire a new customer, you are actually getting paid to acquire a new customer.

This works because most smart media companies are diversified. So, on that content using your data as a source, there are likely advertisements of some sort. That advertising revenue is earning you cash while you are theoretically acquiring new customers to the business.

Compare this to a traditional software company… Its primary strategy for getting people to sign up for its data offerings is more cost prohibitive. Some data-first companies—CB Insights comes to mind—do an amazing job using content to help drive business, but for most of these companies, they just don’t know how to do it. Therefore, they’re stuck hiring expensive biz dev people and paying large advertising budgets.

The multiples are greater

When the time comes for a liquidation event—partially or outright—the multiples on a data company are simply greater than on a media company. I’ll use that same podcast episode for reference:

More than half of the revenue at FreightWaves comes from traditional media sources: advertising on the site and video and subscriptions. However, the company has been able to raise tens of millions of dollars because FreightWaves is also in the data business.

When investors look at the business, they see the entire community across media, data and events and were willing to value the business much higher than if the company had just been a traditional media company. Data multiples are higher than media.

For media companies, I’ve heard of acquisitions in the range of 2-4x revenue. For software companies, multiples can sometimes push the 10x revenue. Obviously a lot goes into this and I’m simplifying for the sake of this article, but the point stands. There is a drastic difference between a 2-4x multiple and a 10x multiple.

How I might do it…

All of this is fine and good… But when we still think about the underlying cost structure and the DNA of the business, it can be tricky to make this work. I’ve started to formulate a thesis on how I think media companies could start to do this in the right way. By no means is this a concrete thesis, but I think it’s worth discussing.

First, simply wait. When you’re first getting started, it can be difficult to invest in everything you want to do. Trying to embark on data too early can eat into your available cash that would otherwise be allocated to hiring the right editorial and business teams.

Remember, the real value of a media/data blend is the audience is already coming to the site. You can’t achieve negative CACs when you’re still fighting to bring audience in.

Second, experiment with rented data before you embark on a massive investment in owned data. Depending on the industry, there might be a variety of small data companies out there that might be interested in some cross promotion. You pull the data onto your site in a limited fashion and, in return, they get some credit.

This obviously doesn’t help you from a primary source perspective, but it does start to get the company DNA used to this new type of reporting. Additionally, you could use this data as some sort of a subscription product where you do a revenue share with the other data providers out there.

I like this approach because it treats you like the industry’s tent. Your audience knows that it should come to your site to get all of its informational needs rather than finding disparate sources for information.

This is still tricky and I would mostly use this strategy to gain an understanding on how your audience engages with this and use it for research on what sort of data you could invest in.

Next, the data team should be entirely separate from the rest of the organization. Remember, the DNA of a media business runs counter to the DNA of a software/data business. Therefore, if it tries to fit within the normal organizational structure of the business, it’ll fundamentally suffer. Priorities will be confused and things will slow down.

As the one running the company, clearly you’ve got a vision for where this should go. And this data department should report up to you. However, this is where you want to be more hands off and build a team you can trust. A media operator’s instincts are more closely tied to the original media flywheel of content, audience, revenue. This team needs to have the time to build out something remarkable.

Finally, be very methodical about it. You’ll likely never stop adding new data points to the platform. But in this case, consider Pareto’s law. 80% of all outcomes are derived from 20% of effort. Yes, you could try to get that last 20% of possible outcomes, but it’s going to cost you so much more.

Therefore, understand exactly what your audience needs and be methodical about adding it. If it’s not being used, remove it from the platform. Data processing costs money. If you see a data type being used a lot, identify if there are ways to add even more of that. Your users will let you know what they need.

To sum up…

For some markets, I think there is a really intriguing opportunity for media companies with more resources to invest in data as a product worth selling. It’s a tricky business and can screw with the DNA of the company; however, if you get it right, it can create a nice flywheel of data informing editorial informing data.

Additionally, the negative CACs that come from earning revenue on stories that promote data found in the platform is incredibly compelling.

It’s not a one size fits all strategy, but it’s worth discussing at the very least.

Join A Media Operator

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