Measurement and Metrics
I’m not sure it was a brilliant idea to put an analytics-based session at the end of Day 1, but here we are. At an analytics-based session on the end of Day 1. Ah, I crack me up. Good times, good times.
So, Rick Bruner (DoubleClick) is acting as moderator amongst panelists Young-Bean Song (Atlas), Chad Parizman (Scripps Networks Interactive), Darren Stoll (Macys.com) and John Squire (CoreMetrics). Am I the only one who just wants to scoop DoubleClick’s Rick up, lock him in a room, and grill him about Google? [Yes.–Susan] Okay, moving on then.
Rick starts off saying that the Internet is much more accountable than any other medium, however, it’s still very complicated. He asks the panel to talk about a client that from a measurement point of view has really made a breakthrough. Who’s getting it right?
Young doesn’t out a company, more a way of doing things. He thinks companies that are doing things right are those that combine their online presence with display advertising. He says it makes sense that there would be some kind of combined effect and obviously there is. He states that advertisers who run campaigns simultaneously with other online efforts saw a considerable lift in conversation rate. To those of us in search, this makes total sense.
Several of the panelists talked about the need for analysts to looking beyond the last click or the last ad seen in order to find a richer story. Darren talked about this in depth, highlighting some of the stuff Macy’s has been working on.
For example, he noted that Macy’s tracks their marketing programs through CoreMetrics, as well as a separate email tracking system. Often while comparing data he’ll find a significant discrepancy from what CoreMetrics reports to what their email tracking program reports. Digging deeper he can see that there is so much activity tied to multiple mediums that to have that cut and dry "this is what got them here" is insufficient. Often people will click through the email, then leave, and come back later in the day through Google and make a conversion. In this case, Google would get credit for the conversion even through it was really sparked by the email.
For advertisers, looking beyond the last click or last ad seen creates a richer story. Sometimes an ad that didn’t look good in reporting is actually found to be delivering really well. It just wasn’t the last ad seen so it doesn’t get the credit. More thorough analysis is used to see that credit and reallocate budgeting channels accordingly.
Darren notes that Macy’s is working hard to identify the different value points throughout the site. Historically, they’ve given the last click credit and used that to determine what’s working but now they’re looking at the long chain of events.
Chad chimes in from a publisher’s perspective saying that their differentiator is video. He says he’s had some really great success with custom Web-based content sponsorship. The hot buttons in measuring video are things like completion of content, watching another video, just starting a video, etc. He wants to see users watching multiple videos in the same session and coming back for more.
You can also using metric information to see the relationship between online researchers who buy things offline. John’s company does a lot of work with tracking customer behavior offline. He uses things like membership IDs, email addresses, and tracking cookies to see what ads are leading users into the physical store to make a purchase.
Young commented that you don’t even need a big CRN infrastructure to figure out the offline/online scenario. He says, imagine you have a simple promotion in the offline store where every time someone buys something you give them a ticket to go to a Web site and enter their email address to see if they "won" a free gift. You don’t need everyone to go, you just need a sample. Once they give you their email or some other identifier, you can go back and see if they’ve ever spent time on your site before. You can use that information to better target products to them in the future.
Rick polls the panelists and asks them to identify one aspect of analytics that is still a really big challenge for their company.
Darren says his company struggles with the cross channel analytics. Figuring out what their online activities are doing to drive offline activity. They’re striving to be smarter on an individual level in order to engage consumers more intelligently.
Somehow the other panelists are let off the hook and Rick moves on to his new question: Where do you think advertisers are not doing metrics well?
Chad says that marketers can come up with a compelling story but they have to get into the equation early on. The good thing about Web analytics is that you can track everything. The downside is that you can track everything. [*forced laughter*] The skill in analytics comes in knowing when to track, how to track and how to report on the date you’re receiving. Chad says this is his company’s stumbling block.
John says advertisers are getting caught up in the hottest new thing instead of spending that dollar on things that are already known to convert well. The thing people struggle with is that there are so many ways to go out and address customers that advertisers get distracted.
The always opinionated Young says advertisers need to get looking at view through conversions. VTCs are when a consumer sees an ad, doesn’t click on it, but then converts. Even though there was no click, the conversation is still attributed to the last ad seen. Young says he’s seen no evidence that VTCs are a significant measure of direct response. Actually, it’s not a measure of direct response at all. It is a measure of targeted reach. If you get more VTC from one site than another, you’re getting a more targeted reach. Basically, Young things VTCs are named wrong.
Darren agrees with Young and says he doesn’t put a lot of weight on VTCs. All they are doing is capturing a sense of the quality of the traffic that’s seeing that ad. When he’s optimizing his marketing he’s looking at what sales are being driven by this and clicks.
What is missing in the field of metrics?
Chad says talent. Heh. I think that’s a great place to end!
See you tomorrow. Blogger’s gotta eat.
One Reply to “Measurement and Metrics”
Commenting on the paragraphs regarding Macy’s:
At my company, we call the multiple clicks that contribute to a conversion/sale the “Purchase Path”. We track every click that is related to a conversion. Our technology has collected many examples of web consumers performing multiple searches through many different channels before purchasing (email, affiliate, search engine, CSE, organic, etc).
We provide this data to our clients so they can see how the various advertising channels are contributing to conversions/sales.
It is entirely possible (before this data is considered) that advertisers may be reducing the effect or eliminating channels that do not appear to be directly related to the sale, but that do provide an “assist”.
Now that this type of data is available, it provides the advertisers with additional knowledge that a particular channel is contributing to positive results. I think it also demonstrates how important it is to be “present” in multiple channels.
The point that is still up for debate is how to allocate credit to each of the clicks along the way. Currently, we consult with each of our clients to customize our reporting to best fit their company.
Lisa, thanks for the always entertaining and informative posts! Keep up the good work!
PS – am I allowed to say “huzzah”?