Best Pracitce Advice: Incoming from Online Ads

[Anonymous]

Hi everyone. We're about to launch our first online ad campaign with real-time bidding (woooo!) and are driving our ad clickers to Marketo LPs. We have three sets of ads running, each with different products targeting different online profiles.

My question is this: in terms of analytics, would it be smarter to have unique campaigns for each ad set or to have one set of campaigns listening for clickers/conversions on all three ad sets to move all clickers along the flow? I know that on the program level we'll be able to see everyone together either way, but in terms of comparing how each set of ad conversions compare against each other would it be better practice to seperate them?

Any advice would be greatly appreciated! Thanks in advance.

[Anonymous]

We use a single program for our ads, but they are for 1 product and generally 1 persona.  If you are able to track the cost separately for the ads for each product, you may want to set up separate programs to facilitate the ROI analysis.

[Michelle_Tizian]

Just like Elliott, I use one program per product type.  I do have a couple of landing pages (a/b test) per adgroup that way I can track which adgroup is performing better (But I obsess over cost per click and cost per conversion because of my limited budget).  I have a few status campaigns that track clicks, and trial requested.  

[Anonymous]

Thanks guys! We pay one fee for all our ads together so in terms of ROI calculations it will be most useful for me to think of them all as one big group, at least for this first test run. I'll definitely run different campaigns for each segement, though.

Michelle - can you give me a bit more info on how you construct your status campaigns that track clicks? I had originally thought that I would use the "Visits Web Page" trigger to get clickers into the program flow but realized that wouldn't work as they won't yet have converted on the form...

[Michelle_Tizian]

Hi Andrew,

It is Visits Web Page that triggers the click on the ad. I label it clicked instead of "visited" because the visitors had to have clicked the ad to visit the landing page (whether it's from the keywords search or from the display network.  Then the other is "Trial Requested" which is the same as Converted when they fill out the form. Same thing we do with our LinkedIn ads.  Since this is your first run with PPC, pay attention to your keyword quality score which looks at the following: Expected click through rate, ad relevance, and landing page experience.  I pause anything below 5/10. Have a keyword in your ad.  And know where to find the keywords that are triggering your ad.  If it's not relevant add some as negative keywords. 

[Anonymous]

Yes, using the "Visits Web Page" trigger is precisely what I wanted to do, but my concern is that it will only work (ie. get them into the program flow) if they're a known lead. Anonymous leads won't make it into the program if I'm understanding correctly...?

[Michelle_Tizian]

Andrew, that's correct, you won't see any anonymous leads in the program.  But in the status campaign for "Visits Web Page" You'll see the anonymous leads there.  

[Anonymous]

Thank you Michelle! I understand now...I'll just have to count my anonymous leads by looking at the campaign.

One final question though for the leads that DO convert. My campaign to move leads to the "Converted" status progression uses the "Fills Out Form" trigger on my form. When an anonymous lead fills out this form they will then become a known lead...but will they then show up as converted in my program summary? Or will they not because they were not known leads at the time they converted?

[Michelle_Tizian]

Hi Andrew,

Yes once they fill out the form and submit, they then become known and will show up as converted.  So yes they will show up on the program summary.  When you check your visits web page status campaign, you can see the unanonymous leads who visited the page, and you'll also see the leads who progressed because their contact information will appear.