Data Availability and Use

Productivity Commission Draft Report October 2016

I had the opportunity last Friday morning to attend a round table discussion with the Association for Data-driven Marketing & Advertising (ADMA), hosted by their CEO Jodie Sangster. The discussion focussed on the Federal Government’s draft proposal on the Productivity Commission’s Data Availability and Use draft report.

The purpose behind this meeting was for ADMA to hear from organisations, in the business world, about aspects of the draft report that would potentially impact their business.

There were a few areas of concern raised by the group in the meeting. Two key issues raised related to “Inferred” data and also the term “in the national interest”.

Inferred data — based on the analysis of other types of data. For example, retailers can observe changes in shopping patterns, and use these to draw inferences on a range of topics, potentially including people’s health concerns or other personal information (in the US, retail giant Target identified customers’ pregnancies from changes in their shopping habits (Hill2012)). Further, information about one person’s location or activities can be used to infer information about other people as well.

SOURCE: Page 190 Data Availability and Use, Productivity Commission, Draft Report, October 2016

Datasets that are of national interest should be recognised as strategic assets, with their curation funded accordingly to ensure their quality is maintained. As these datasets would, by definition, be capable of generating significant public benefits, their maintenance should be funded by taxpayers upon designation.

SOURCE: Page 357, Data Availability and Use, Productivity Commission, Draft Report, October 2016

What does this mean?

Giving the consumer the right to access an organisation’s “inferred data” became a sticking point and something that the group largley opposed.

As one of the banks put it. If through data the bank has access to or generates in the course of doing business with a person, they’ve determined an individual is 70% more likely to need a home loan in the next 12 months, that inference is based on the intellectual property developed by the bank.

I see this happen with Eloqua customers. Developing campaigns to deliver the most relevant message to a group of people means you need to take explicit data, blend it with implicit data and then make some determinations from there.

e.g. If a contact has looked as a range of web pages covering training services and then they’ve looked at pages about training delivery options and they’ve recently opened a number of emails that cover end-user education, it’s not a stretch to infer from this that they’re looking to arrange training.

That inference is not data explicitly sourced, submitted or provided by the contact, it’s inferred from their behaviour, or their Digital Body Language™ as we like to call it.

The “national interest” discussion was even more robust. In the absence of any definition, it was agreed the term is potentially all to encompassing. While the term “national interest” appears many times in the document, there is no specific definition.

The Internet of Me®

When I got back to the office, I talked with the team about the roundtable. We talked about the number of online locations that we provide our personal data to.

The diagram below, or “The Internet of Me” as we decided to call it, shows an example of one person and their many online personas, who they’re with and the relationship between each one.

For example, the person below uses their Google Apps account ID to login to various platforms. From WordPress they use a plugin to share blog posts to various social sites. It quickly becomes very fuzzy about who owns a person’s data and unclear to the average person, exactly where their data is stored and who has access to it.

In the diagram above this person manages multiple WordPress blogs, has multiple Google Apps accounts and administers various social properties for work, personal and a not-for-profit they volunteer with.

It’s easy to see how quickly core, personal information about yourself is disseminated across many companies. There are at least 18 different companies represented above with many having more than one profile of a single individual. e.g. the person has three Twitter accounts, two XERO accounts and multiple Facebook accounts.

Access the government reports

It’s a commitment to read with just over 600 pages, however your legal team may want to at least read the draft report and the proposals.

ADMA is preparing a formal response on behalf on the digital marketing industry. As soon as we can obtain that report, we’ll share it with you.