This content has been marked as final. Show 12 replies
Unless they have turned off their cookie settings, any website or email open would start the cookie path (even anonymously on the web property) based on the Munchkin being in place. A MKTO landing page would be able to identify your cookie details to help pre-populate a form on that landing page - but general web browsing can be done anonymously or known - but cookied.
Hope this helps.
Just a slight update to what Brigid above said - an email open will track the lead's email behavior but won't nessecairly coorelate that name with any web browsing history.
Every visitor to your site is tracked every time they visit (unless they block cookies). In order to associate the name to the unique cookie (tracking ID) however, the lead needs to click on a link in a marketo email back to your site or fill out a form on your site or a landing page, an email open alone is not sufficient. Once they do this you will see their browsing history on the lead history. You don't need a MKTO landing page, you can do this with an embedded form on your main site as well.
Also, its worth noting that if someone clears out their browser history (eg: clears cookies) they will need to click another link or fill out another form to re-associate their browsing with their cookie.
@Nate Taber Clicking a link in an email does not associate the lead. Leads remain anonymous until they are confirmed via form submission or via Munchkin, REST, or SOAP APIs.
@Vicky K You should ignore email Opens because a user who opens and thoroughly reads an email still won't get logged unless they have images enabled in their mail app. Opens can also be backfilled synthetically by Marketo which makes them ultimately ambiguous. Use Clicks as your reliable metric.
Thank you for your replies!
What if these emails are being send to our current database? Does an "open" email action allow tracking of their digital behavious in the future?
@Sanford, so are you saying "opens" is not a good metric to use in this case?
If we have an email that has no landing pages asspociated, how does it help to track then?
@Vicky K an Open doesn't allow ongoing tracking at all, even when it succeeds. Think about Microsoft Outlook with images enabled. Even if a user does download the tracking pixel (which tells Marketo they opened the email, at least for a couple of milliseconds) there is no way to track their later digital behavior because they are not in their browser at that time. You can't put cookies into Outlook.
Opens is never a good metric because, for security reasons, many/most (depending on your type of lead database, % of Gmail users vs. corporate users, etc.) mail apps simply don't allow Opens to be logged. That's why having clickable links is so essential. Links get the person into their browser, where you can be (reasonably) sure that only deliberate user actions are recorded and (usually) can set cookies for tracking over time.
I don't mean that seeing an Open in somebody's ActLog is completely devoid of meaning. It's that not seeing an Open is meaningless. Each lead could've paid just as much attention to the email, but only one of them was logged. Thus reporting on Open activities to measure success is grossly misleading.
@Vicky, there's a good overview of when munchkin tracks a lead and how a lead moves from known to anonymous here: https://community.marketo.com/MarketoArticle?id=kA050000000LCxMCAW
Essentially, a lead becomes known through a form fill, marketo email link click, or the SOAP Api (3rd party form) - an open is not sufficient.
@Sanford - thanks for the good explination on opens. I agree that the reliability of the opens varies based on your DB. Gmail, for example, downloads images for all emails, as do many mobile clients. So if you have a consumer DB than you will probably see more accurate open rates.
@Nate A lead cookie doesn't become known just by clicking any link in a Marketo email. We may wish that were so, but it doesn't work that way.
Yes, documentation is wrong, or at best misleading. It's pretty easy to test + prove....
Hi folks, so I haven't proven the docs are wrong here. My understanding was that:
- This unknown lead then fills out a form and is associated to an email, becoming Known. This is the basic function of marketo.
- This Known lead then receives an email from you. Their Open, Click, or Unsubscribe can now be directly tracked. (I agree Opens is a bad metric, but it will be TRACKED if possible).
- Known Email Address inside Marketo - but they were never cookied by marketo before. This could happen when you first turn on Marketo and migrate the database. Technically they are known, but uncookied. Marketo urges people to do a tagging campaign
- Send Email to lead. If they click or Open the email, then we know only that they did this behavior.
- If they click through to any page with munchkin code, they will be cookied and should be assigned a cookieID, associating their email and cookie automatically. So in theory, the lead was already known, they just assoc to the cookie.
- I know this works because I can put an email in the system, send an email, click on it, and get a visits web page log activity.
Or we could say that my browser is cookied anonymously. Later, I get an email and click on it. will Marketo NOT assoc the cookie and email address unless I explicitly ask it to?
We've found automatic lead association does not work across all browsers and versions of Munchkin. I'm pretty sure it hasn't been working as reliably as you think! The only way for us to test this functionality was to use Private/Incognito browsing and/or one-time virtual machines. Using the same browser session you use for daily work breeds confusion -- there was huge confusion at my workplace because people thought auto-association was working, when they had actually filled out a form. Once we ensured that everybody was testing with a fresh, uncookied browser session we were able to conclude the magic didn't always work. We had to use the Munchkin API to associate leads based on a special token stored in SFDC, which we in turn added to the query string of email links. This gives us the 100% assurance.
EDIT: The testing period I'm referring to was October of last year. @Josh - Given your confidence I have assigned our engineering team to reevaluate the most current version. It may be flawless now, though if anybody is hosting older Munchkin versions they would still be problematic. (Tightly run development teams host locally for change control, for better and worse.)