As I talked about in my last post, “Friend Discovery” is a concept for a new branch of Social CRM.
Friend Discovery connects the Customer Relationship Management (CRM) database in an organization with the social networks of its customers and other stakeholders.
You’ll definitely want to read that first post before reading this one – otherwise, this won’t make much sense.
Three Principles:
I’m part idealist and part pragmatist, which makes me a big fan of starting with first principles. Being clear about what you want – and don’t want – in a project is the first step to making it succeed. So with that in mind, here are three principles for helping Friend Discovery achieve its full potential:
1) “Do Not Share” is the Default Setting:
If handled carelessly, Friend Discovery could be a privacy nightmare; serious enough, I think, to fundamentally question the idea in the first place. Imagine, for example, if my friends visited the website of my local pharmacy and discovered I regularly purchased some medication that would make me really embarrassed – take your pick on the drug. Imagine how disastrous a slip up like that would be. At a minimum, I would never buy anything from that pharmacy again, and maybe much worse.
There are exciting upsides to Friend Discovery, however, that I think still make it worth pursuing. The question is how do we realize those opportunities and mitigate against its real threats to privacy.
Organizations will inevitably make mistakes with something as powerful as Friend Discovery; there’s no way to completely prevent that. One principle to reduce the probability of mistakes is to make “do not share” the default setting for all information about customers’ interactions with a company. In other words, Apple only tells my friends I have an iPad if – and only if – I explicitly tell Apple that’s ok.
To simplify the user experience, companies may need to clump categories of goods and services. I might, for example, tell Apple that I’m ok with sharing my “music purchases” but not my “electronics purchases.” Whole categories of goods and services may require special confirmation settings, so that I might have to re-confirm if I chose to share my “medication prescription purchases” at that drug store.
Best practices on confirmed and unconfirmed email opt-in may be a good model here.
2) I Trust Some Friends More Than Others:
When I’m sharing updates on Facebook or tweets on Twitter, I don’t generally have to be all that selective about which friends get what information. But when I’m sharing information about what consumer electronics I have in my home, which stocks I own, or my plans to attend a particular out-of-town conference, I want to be a little more careful about who’s seeing that information.
This is the reciprocal of the first principle. In the first principle, the organization needs to carefully control which types of information about me it exposes from its CRM. In this second principle, customers need to be able to control which friends in their social graph get access to which types of information. Ultimately, this ability will depend upon the capabilities of the customer’s tools for managing their social graph.
Today, Facebook allows an approximation of this kind of ‘permissions setting’ with the lists you use to categorize your friends. With Friend Discovery, the stakes go up, and there would need to be more explicit tiering of friends based on your levels of trust in them. That’s a tricky issue, to be sure, but one that’s critical if customers are to trust sharing information with people they know via Friend Discovery.
3) My Relationship Data is My Relationship Data:
Friend Discovery needs to ensure that relationship data stays where it belongs. Information about the organization’s relationships with its customers and other stakeholders belongs in the organization’s CRM database. Information about customers’ relationships with their friends belongs in those customers’ social graphs.
Friend Discovery needs to be able to temporarily combine that data without compromising it. The organization should not be able to use Friend Discovery to build a copy of the customer’s social graph, and the customer should not be able to use Friend Discovery to generate a copy of the organization’s CRM records.
There may well be other important principles beyond these three. I welcome your thoughts on additional ground rules to ensure Friend Discovery is developed in such a way that it helps promote good behavior in the organizations that use it, and that, more generally, it’s built in such a way that it contributes to making the world a better place.
I’ll be exploring more about why I think Friend Discovery could be a real force for good in the world in a future post, but coming up next – some of the challenges of turning Friend Discovery into reality.