False Positive problems

Yesterday, we talked about the grey-area between spam and legitimate direct marketing (DM). We mentioned that there are problems when legitimate DM is filtered as spam, becoming a false positive.

You may argue that DM mail isn’t critical, so this sort of collateral damage isn’t important. Indeed, some anti-spam vendors prefer to talk about "false critical" rates, instead of false positive rates, as if non-critical false positives don’t matter. There are three reasons why this isn’t the whole story:

  1. Any false positives can undermine confidence in the spam filter. Or undermine the organization’s confidence in the IT dept.
  2. Time wasted by checking the quarantine is far more expensive than time wasted deleting spam, on a per-message basis.
  3. Just try telling the direct marketers that their mail isn’t
    critical. The DM industry is in constant turmoil, creating new
    solutions to their "deliverability" crisis.

Steve’s right when he comments that people don’t want to have to train
a filter
, that they just want it to work. However, if there’s
disagreement between individuals in the organization about what is and isn’t spam,
training by exception will be necessary.

One Comment

  1. Posted February 8, 2005 at 8:00 AM | Permalink

    People and organizations need to deal with exceptions. In my view, white lists and black lists are usually going to be the best way to achieve this (rather than training mechanisms)

    Steve

Post a comment

You must be logged in to post a comment. To comment, first join our community.