The most obvious case is where APIs generate direct revenue by being pay-to-play (great examples include Xignite, Hoovers, Twilio or potentially Amazon AWS) – in these cases the API effectively is the product.

Beyond these more obvious direct cases though (and this seems to be where you’re headed judging by your examples!) there are lot of cases where APIs generate business value without being directly charged for and hence grow revenues overall but indirectly. These indirect impacts of an API come in different flavours.

One is where the API essentially projects the core product (whatever it is) to new places for users to use it – that might be mobile devices, integration with partner sites, to company intranets or new variants of the core product’s interfaces. All of these increase the value of the core product for it’s users (“Wow – I can get my salesforce reports on my iPhone”) and also allows third parties to add value. Examples of this include (which has 50k+ developers adding value to it’s platform), eBay, FedEx and a lot of others. The revenue impact is really two fold: 1) potentially increasing the spend of existing customers – since they’ll want to activate new options and features, 2) increasing the stickiness of the service by pushing up switching costs once people have come to rely on the service being present in all these environments. Dropbox and Evernote I’d say also benefit from this type of value.

A second area is where the API promotes some core product. Examples here are many media APIs such as Netflix, NPR and The Guardian – these APIs allow content to be extracted and reused by others. The content can be core content (in which case normally the organisation providing the API will require branding information to be shown when it is reused, or a share of advertising revenues) or be information such as listings which drive traffic directly back to the original product (see NPRs great APIs for this form example). Ecommerce sites that provide APIs also use this model – the API is coming to accompany affiliate programs by allowing partners to build increasingly smart “shop windows” away from the core site. The business value here again isn’t in charging for API use (in fact you may effectively be paying people to use it!) – it’s in the traffic or transactions you drive back to your main product.

A third area which may seem a little odd first is value generated by APIs which actually grow (or create the product) itself – I.e. they “feed” the product. The best example of this is Twitter (but there are plenty of others – YouTube, Foursquare, Facebook) – here the value is in what’s coming in via the API – content, information, check-ins etc. which the core product needs to make sense. Although Twitter is only just beginning to roll out it’s business models, they all revolve around the fact that twitter is a river of real time content – 75% of which comes in via the API. Hence it wouldn’t make sense to charge for (normal) use of the API at all – but later on either via adverts in the stream or charging for the Firehose (another API!) the company can generate revenue.

In each of these cases, there is no $ value you can really associate to each API call – but you can calculate the impact on overall revenue even if it’s indirect. Further, this calculation should include the soft benefits of putting out an API such as engaging your users more deeply and increasing their potential commitment to your solution.

So I suspect that Evernote and Dropbox’s API programs, whilst not generating direct revenue for them (they are free) provide value in at least one (possibly two) of these other dimensions.

– Edit: I should add that there are some useful presentations out there on the value of APIs, a great one to check out is Sam Ramji’s Darwin’s Finches presentation

Lastly it’s also worth noting that APIs can also reduce costs – by making integration quicker and easier with large partners and streamlining products + they can create opportunities which weren’t there before – you might be able to say “yes” to a big partner which you wouldn’t be able to do otherwise.

Steven Willmott
Co-Founder & CEO of 3scale.

NOTE: You can also follow the whole discussion thread on Quora