Great post, and your thoughts on monetizing are clearly defined and interesting. I’ll be using your definitions in future discussions. It occurs to me that what you need for the actual access control component is some kind of ‘content agnostic’ authorization service… I wonder where you might find one of them? We should talk.

Cheers,

Peter

Say I’m subscribed to a feed about vacuuming. If a link to “Vacumaster 3000″ pops up in the stream, I’m okay with it. In the consumer – app / service provider – data provider triangle no one blames the other. I think it’s very viable and will revolutionize marketing.

This post was mentioned on Twitter by ldodds: New blog post (prompted by blog post from @sbrinker): Thoughts on Linked Data Business Models: http://bit.ly/8XiSWp…

Thanks for the kind comments on my post and taking the discussion further. These are great ideas.

I completely overlooked “sponsorship” model, but I agree that it’s an important one. In my visualization, I’d picture it at the very top, as the most direct form of revenue: you’ve got the revenue (or mandate) even before the data is delivered to end-users. Government services, regulatory requirements, NGO missions, etc. all fall under that heading — and, at least today, represent a significant portion of the big linked data sets out there.

I think you’re absolutely right that freemium models are particularly powerful for subscription and value-add models. You’ve got a great list of ways in which the premium version of the data can be differentiated, without hindering the wide distribution of the free version. One other variation may be tied in with the authority model: validated/certified components of the data that are added for paying subscribers.

However, I want to better explain the idea of advertising inside data.

What I’m thinking of there is not advertisements the way they exist in the visible web, but sponsored extensions to a data set that would have actual utility to people consuming these feeds. Enough utility that people would be more inclined to use them rather than strip them out.

Here’s a hypothetical example. Let’s say Google starts making linked data available about what it knows of local businesses: location, business hours, categories of products and services, etc. Maybe they include ratings data (as part of an “authority” model). Now, imagine that they have an optional set of “coupon” fields for special offers made by those businesses. But in order to include coupon information in the data set, the local business has to pay a fee.

Such coupon data could likely be valuable to a set of folks consuming that data. The local businesses are then subsidizing the data set by their paid inclusion of certain extra data within it it. (Kind of like a supply-side freemium model: extra premiums for being in the data set.)

Granted, this is hypothetical. But does that seem like it might be more useful?

1. Reuse of data and search engines

There is a danger that Google will penalize those who reproduce data from other sites. Although their algo might be something of a blunt instrument at the moment, everything seems shrouded in mystery about what they deem “copied content” – expect this to become more sophisticated.

e.g. Your site about mytown legally reuses data from dbpedia. Which is the most believable? Which should be ranked highest?

2. Subscriptions

Turning this idea around there will also be a need for data to be corrected and sent back to the data supplier.

I can imagine this working in much the same vein as me selling electricity back to the grid.

Is that another model? Might you be at once a supplier and collector of data, or some kind of middle man?