5 Replies Latest reply on Feb 13, 2014 5:52 PM by Josh Hill

    To Nest or Not to Nest . . .

      Hi Everyone,

      I have chatted with a lot of folks lately and it seems there is some confusion and inconsistency in how the industry is leveraging the Engagement Module together with Programs. 

      A few notes that I would like your opinion on.

      1.  What is the fundamental deciding rule you use to decide whether an email should be nested with a program or put directly into an engagement stream?

           Maneeza's thoughts:  A program should be nested in an engagement stream ONLY if part of an overall nurture      stream. For Example:  Email 1 --> Email 2 --> Webinar Program (with 4 emails in it) --> Email 3 --> Email 4
      If ALL you have are programs in an engagement stream, the only benefits you are getting are:

           1.)  Send cadence controls like former traffic cop functionality.
           2.)  Organization.
           3.)  Your egagement score is simply a reflection of program success in this case.
           4.)  If you only have 1 asset per program in this stream, and you are leveraging RCA, this has implications for multi-touch attribution in RCA (we won't open this can of worms just yet) 

      Would love to hear your thoughts!

        • Re: To Nest or Not to Nest . . .
          Jeff Shearer
          I've generally avoided nested emails or programs in an engagement program, and prefer to link to outside programs instead. Most of the stuff I'm sending via nurture is content like whitepapers, webinars, etc, and I have dedicated programs that centralize tracking for each piece of content. Thus, I want to ensure if someone recieved a piece of content outside of my nurture that I'm able to track that activity, and skip sending that piece of content again if they end up in my nurture stream. 

          I also prefer to keep creative centralized where possible, so if I need to make a change to a standard email or landing page, I can do it in one place and have the changes reflected everywhere. I also a/b test everything, so I'm forced to use programs with smart campaigns and the random sample approach.

          One downside I've encountered is that engagement programs that primarily use outside assets appear to lose some of the tracking and dashboard benefits. For example, an email sent via an outside program doesn't appear to register the unsubscribe rate on the engagement program level. The "engagement score" appears to still calculate though. So there's definitely a tradeoff, but in my situation, there's far more benefit to using outside programs than local assets.
          • Re: To Nest or Not to Nest . . .
            I have seen some customers nesting programs when they have a limited number of engagement streams, and are trying to "aritificially" create more. For example, they have one engagement stream that has member leads from multiple industries. They add a program to the engagement stream with a smart campaign that sends one email if industry=retail, a different email if industry=financial, etc.

            This setup makes it difficult to add or subtract content, or to see who has received what email. (And my understanding is that those are the primary benefits of the engagement stream methodology.)

            I think a better solution to that problem would be a single email with segmentations. If the content or audience varies so drastically that segmentations are not enough, and a limited number of engagement streams are available, an engagement stream may not be the correct vehicle for the campaign.

            Engagement streams are meant to simplify, not complicate--so I would also vote for only nesting programs if they are part of an overall nurture. That's my 2 cents worth!
            • Re: To Nest or Not to Nest . . .
              Hi Guys,

              Thanks, I tend to agree with both of you.  Jeff the reason you are missing the engagement score for assets outside of the engagement engine is that the asset needs to be in the Engagement Module to be part of the algorithm.  

              Also, Jeff, not sure if you have RCA, but if you do, how are you able to measure channel level attribution if you are measuring the success of your content centralized from all channels?


              • Re: To Nest or Not to Nest . . .
                Jeff Shearer

                I'm aware of the limitations of tracking with non-local assets, but it's actually not the engagement score that's the problem- that actually seems to be tracking just fine, which was a pleasant surprise. It's the "unsubscribe" metric that I'm specifically missing in the dashboard. It's not a huge issue- I can still get to that data other ways, but it's an annoying limitation all the same.

                Regarding the RCA question, I've essentially decoupled content from channel in my program design. If I have a new whitepaper, all the assets that relate to it are located in one central content program. If I want to use that piece of content in a particular channel, such as online advertising, I have a separate program for the channel that's used for tracking, and references the content program's design assets. Then, a lead is essentially a member of both programs:I can see that the lead converted on X channel, and meanwhile, that they also downloaded Y content.

                So I can still do channel level attribution, but now I can also do content-level attribution. With this structure, there admittedly isn't a great way to look at content performance WITH channel in Marketo's RCA, but again, the structure I've pursued has benefits for me that outweigh this reporting limitation. Ultimately, I can can get at mutual program membership data via SFDC reporting if needed.
                • Re: To Nest or Not to Nest . . .
                  Josh Hill
                  This slide show covers some of these issues