“I get all my news from Twitter” has become a common boast. Quite often you hear, “I found my new job via LinkedIn”. But you rarely get professional people talking about their Facebook presence. Yet the market cap per user of these companies is the complete reverse - $54.4 for a Twitter user, $64.2 for LinkedIn and $167.9 for Facebook, according to Forbes in October 2015.
That is not just down to sheer number of users - it also reflects levels of engagement, such as posting, commenting and clicking through links. When it comes to revenue-drivers, that’s what makes the difference because advertisers are looking for audiences who will notice and interact with their content. Facebook gets a big thumbs up from investors on this front, whereas Twitter is looking for a buyer and is said to be running out of cash.
Monetisation of data has therefore become a crucial component of the business plan for these - and other - social networks, such as Twitter selling its data via Gnip. Good news (for now) for media planners and buyers who need to understand exactly where and how these audiences engage with content.
But there is a problem. At this week’s Demographic User Group conference, Seth Spielman, associate professor of Geography at the University of Colorado, made the point that 95% of the data being created in social networks sits in walled gardens. That means it is only available at a cost or subject to tight usage rules, rather than being accessible via the open internet.
For him, that creates problems when trying to find data sources to improve the accuracy of the American Community Survey - a rolling research sample that replaced the decennial national Census in the US in 2010. On some variables in the survey, the margin for error is greater than the estimate. As Spielman pointed out, the confidence level in some data on which major federal funding decisions are being made can be between 0 and 100% - or useless in any meaningful statistical sense.
One of his tasks is to identify additional data sources that might help to improve the survey. Social networks are an obvious place to look as they contain a lot of self-generated information about lifestyle, household composition, affluence and so on. But the costs can be prohibitive, even assuming the data is available to third-parties.
This is not just a problem for academics. Investors are heavily backing the “platform economy” and technology start-ups, many of which have data sharing at their heart. Monitoring devices, from personal health to home security, only really have long-term value if that data can be shared, benchmarked and used to optimise services for each owner by leveraging the network effect.
When a new technology is launched that plans to keep data within its own walled garden, it may look like a smart play because it gives the business control over how to monetise that information. But it also constrains the ability for an eco-system of third-party services to grow up around that data.
It is a conflict of interests which is a long way from being resolved, even where it is being thought about. Investors know they want to put their money where the data is. Right now, walled gardens look like the place to be. But that may only be true in the short-term if open internet eco-systems prove to be where users look for better service.
Amit Pau, managing director, Ariadne Capital and Ros Singleton, COO, UK Broadband and angel investor with Angel Academe, will be discussing investing in data and technology at DataIQ Future on 20th October. For more information and to book your ticket, go here.