Oracle has been discussing how the company intends to further develop its customer relationship management (CRM) products based on its own experience of what it terms "contextual social CRM".
Anthony Lye, Oracle's head of CRM, explained during his speech that the firm's concept of CRM is based on the hypothesis that relationships with customers are changing.
"CRM systems in their early designs were transactional, and built to serve a company's relationship with the customer, who was at this time either a company or a household," he said.
"Now customers want to talk to other customers and, because most CRM systems do not facilitate that, they go to other places to talk to each other."
This change makes it vital for companies to extend their CRM systems to listen to the conversations taking place outside their visibility, and to take the appropriate action.
While Lye acknowledged that many companies already monitor social networks and communities, he maintained that their methods of doing so are often " primitive".
Lye suggested that businesses should start by ensuring that their CRM systems listen to the "hierarchies in relationships".
"This includes the explicit relationships, like the 'I know you, or you know me' type, as well as implicit relationships, which are basically derived by sophisticated analytics and clustering," he said.
Expanding on the implicit relationships companies might look for, Lye pointed to Amazon, which he said employs an "implicit relationship structure. It tells customers that people who bought this also bought that."
Implicit relationships are important to understand, according to Lye, because people in social communities tend to listen to each other if they believe themselves to be similar.
After understanding social relationships, Lye suggested that organisations should understand their context and determine the relevancy of the comments to their reputation.
"Businesses search social sites or their communities for mentions of their brand name, and can panic when it comes back with all this feedback. But what they need to realise is that a lot of the feedback may not be relevant," he said.
"What you have to do is determine relevancy and influence. Who is actually listening to the comments? If someone says something today and no-one comments on it tomorrow, it becomes irrelevant.
"A good way to find out the context is to capture the conversations occurring in online communities, stimulate them and then mine and track them over time. It's all about analytics."
'Whatever the causes of political polarisation today, it is not social media or the internet,' claims Dr Grant Blank
Tesla founder leaves OpenAI group - while Valve Software's Gabe Newell joins
'If someone were to venture into one of these relatively benign black holes, they could survive, but their past would be obliterated,' he claims
Captured by accident by an amateur astronomer in Argentina