Karma: Electronic Feedback and Identity Management

Users deciding to make a purchase using popular online classifieds are faced with a series of seemingly intractable problems:

  • How sure can I be that the anonymous seller is trustworthy?
  • What is the seller’s history of reliability in the past? How many appointments has he failed to attend?
  • Has the seller described previously sold items accurately?

Sellers have their own set of questions regarding buyers:

  • How often has the buyer failed to show for an appointment?
  • Is a buyer who wants to issue a check creditworthy?
  • Is the buyer likely to post negative feedback about the seller?

Current online classified/selling sites offer few answers to these questions, yet for thousands of years buyers and sellers have used exactly the same currency to decide whether a transaction is wortwhile: reputation, or as Kobo calls it, Karma.

The Karma Patent

Reputation is a tricky business, and automatically scoring it requires multiple axes of measurement.  Karma is measured using a proprietary algorithm to score the following areas.

Identity: from email address to credit check

Who are you dealing with? Karma measures from weakest to strongest (accreted step by step; each level depends on the previous)

  1. Weakest: unconfirmed email address.
  2. Confirmed email address
  3. Unconfirmed physical address
  4. Physical address confirmed through automated means
  5. Credit check, results not disclosed publicly
  6. Credit check, results disclosed publicly using simple Weak, Average, Strong representations
  7. Strongest: Place of business visited by identity inspector

Buyer feedback

Optional feedback can be left by sellers for buyers. Scoring includes but is not limited to:

  • Promptness in communication
  • Promptness in payment
  • Made in-person appointments

Seller feedback

Optional feedback by buyers for sellers. Scoring includes but is not limited to:

  • Item described accurately
  • Original images supplied
  • Item packaged appropriately
  • Made in-person appointments

Community involvement

Deciding whether to do business with someone might hinge on their involvement in the online community. A seller who constantly posts unfocused, hateful messages about buyers in general may be considered a risk for buyers to deal with. A buyer who posts racist messages may not be the kind of person a seller wishes to negotiate with. Buyers who post helpful messages about how to buy a particular kind of good could be considered beneficial to both other buyers and sellers.

A member who catches obvious spam messages quickly and efficiently is an enormous asset to the group as a whole. Merchants to flag their competitors’ honest ads as spam are deleterious to the group as a whole and should be dealt with accordingly.

Don’t forget positive contributions. Members who frequently post useful messages can be thanked by other members.

Community involvement scores are derived from transactions including but not limited to:

  • Member flags spammers appropriately
  • Member flags as spam ads the community does not agree are spam
  • Member frequently posts constructive messages
  • Member frequently posts unproductive or hateful messages
  • Member frequently uses obscenities (can be determined per forum; some forums may be more lenient than others)
  • Member has been censured frequently
  • Member is in good standing for  short period
  • Member has been in good standing for a long period