10 Comments

• 

  Fred Tabsharani says:

  Rather than “stars” could you give each list segment an acutal score? In fact have a weighted rating for each important metric and design of a cumulative email score for each campaign. It can be a four digit integer. Basicially, each campaign will give birth to a Quality Email Score that includes a different weight rating for each metric used. Sends, Engagement, ROI, bounce, etc.

  This is much more involved and a big hill to climb. I have more information on this idea if you think it’s relevant to your clients.

  thanks,
  \fred

  Reply to this comment

• 

  Tim says:

  Does anyone actually see this as a spam fighting measure? This is a relationship analysis technique, one of the most powerful analytical tools in the intelligence community. Maybe I should explain another community term “flexible morality”…

  Gmail makes money by reading your emails and providing advertising relevent to your discussions. Now they propose to try to find out which advertisers (or people…) you are engaged with and you can bet your hat this will all going into their marketing database. It has nothing to do with fighting spam. This is the lipstick in the “lipstick on a pig” analogy.

  If anything, we should view this is an anti-competitive measure and you can watch as the rules for delivering to Gmail will become more opaque over time. Remember their ads are in competition with email marketers, the fewer of us their are, the more their ads stand out. Where FBLs represent a legitimate business concern of ISPs, this is merely a new business area for Gmail not a way to fix an old problem.

  I’ve said before that I wouldn’t use Gmail because I don’t want even their machines reading my email. I certainly don’t want them keeping track of who I’m “engaged” with. And as an email marketer, this is clearly a bad development.

  Tim

  Reply to this comment

  • 

    Ben says:

    Hi Tim, I’m a super paranoid guy, and generally appreciate your hearty skepticism of almighty google, but I think if Google wanted to track your relationship with marketers they could just do that anyway, without really changing the “images on” setting or deliverability. On a somewhat related note, I bet you’d really like this video series entitled, “google is our roommate” – http://www.youtube.com/watch?v=9RDe2Ia6YlM
    

    Reply to this comment

    • 

      Tim says:

      Ben,
      Nobody appreciates a good paranoid episode more than I, but that doesn’t mean they aren’t out to get us… And thanks for the link, I’ll check it out.

      But regarding this “spam measure”, Google is the big dog and subsequently is picked on from all sides from very high levels (I was recently interviewed by the FCC on a Google related matter…).

      So far, they have been reading your emails and applying ads. For an infrastructure provider to actually extract information from your email, store it, and organize it is a much greater step. You have to be able justify it in altruistic fashion (your ‘cover for status’) to protect yourself.

      Google is looking at Gmail as part of their multibillion dollar ad business, but it does nothing for them unless they can monetize it. Why would they hesitate to change functionality in their email client to accommodate a new monetization strategy?

      Seriously, I realize it hasn’t been completely unveiled yet, but do you see any benefit at all vs other methods against spammers? If so, please lay it out! If you can find a superior spam fighting method in there, I’ll eat my hat and apologize to all.
      Cheers,
      Tim

      Reply to this comment

• 

  George Bilbrey says:

  Although engagement is increasingly important, you are overhyping the demise of complaint rate data and missing some additional new metrics that are equally important.

  Over the last few years, ISPs *have* been investing in a new set of metrics on top of traditional reputation metrics such as complaint rates, unknown user rates and spam trap hits. In particular, they are doing this because complaint rates (please *don’t* call complaint messages a ‘feedback loop’) can be gamed by mailing to inactive subscriber bases.

  1. For all the large ISPs we’ve talked to (including the one that you are refering to in the blog post) complaint rates are still part of the reputation mix. Complaint rates aren’t going away.

  2. Engagement is far from the only “new” thing that they look at. Newish metrics include:

  * Trusted User Complaint Rates: ISPs only look at complaints from what appear to be active and real accounts.

  * Trusted User “This Is Not Spam” Votes: ISPs look at whether trusted users move messages from Junk/Bulk folders

  * Panel voting systems: ISPs such as Hotmail ask volunteer panels of user to vote messages intended for them as “spam” or “not spam”.

  Reply to this comment

  • 

    Ben says:

    Thanks for the clarification(s), George!

    Reply to this comment

  • 

    Tim says:

    George,
    Thanks for the additional information, adds some transparency to the argument.

    I see your point about gaming the complaint rate system by sending to inactive mail accounts. I would not be opposed to accounts with no recipient activity for 6 months or so being eliminated from the complaint rate calculations. It would also be nice if the ISP would feed those accounts back so we can remove them (with a news service you can’t always tell).

    The term “trusted user complaints” as complaints from people with active user accounts, I’m not sure I follow.. Is this as opposed to complaints from “inactive” accounts (in which case I’m really confused…)? Or does the ISP select a posse of activists whose complaints are weighed more heavily. If the latter is the case, then my poor conservative newsletter is going to get whacked.

    As far as user panels go, these get corrupted very quickly and turn into content censors. I’ve been the victim of this. I would be opposed to any system that puts a small group of people in charge of my delivery based on content.

    ISPs should not be in the business of judging content, nor should a small group determine what content is blocked from the rest. You can’t say that any particular content is “spam” You can, however, note the sending practices, the changing source IPs the lack of CANSPAM measures, the accreditation, authentication, etc.

    Let’s keep the playing field level, the criteria objective and the ISPs from reading our email.

    tk

    P.S. George, I notice you are with RP, I’m a fan of your SS product, its the closest to objective I’ve seen and was my first read each morning.

    Reply to this comment

• 

  Alec says:

  It seems that the new system is going to be a lot more subjective, and that doesn’t necessarily reflect the level of ‘engagement’. How are they going to weigh each action (open, click, etc.)? Would they register opens in preview pane or just webmail? These type of questions are important to senders to evaluate the success of their campaigns, but would they be publicised by ESPs?

  Reply to this comment

• 

  Bronwyn says:

  Wow, that is definitely something to be aware of! In fact, I think maybe this is ISPs going a little too far and being a little too intrusive. No one likes span, but no one likes interference either!

  Reply to this comment

Leave a Reply