Guest blogger: Robbin Steif from Lunametrics!

[ I'm really happy to have my first "guest post" from blogger Robbin Steif from Lunametrics. Robbin really liked my "gradual building of context" post from awhile back and she and I have been discussing a related metric that she thinks builds nicely on my visitor engagement metric. Without further ado, Robbin Steif ... ]

On the one hand, I thought that Eric’s recent post, The Gradual Building of Context, was just awesome. Although every site has to define visitor engagement for itself, every site is still capable of pulling together similar numbers (which is why I loved it.)

On the other hand, I disagreed with Eric’s final conclusion, “I need to reach out folks like Matt, Marshall, and Clint and see if there is some way I can get them to more passionately advocate for my books in their weblogs. Given that their visitors are more highly engaged than the “average visitor’, I have to believe their is an opportunity to sell more books.”

I took one look at the numbers in the last chart and thought, well, that doesn’t make sense. Sure, Matt’s visitors (or Clint’s or Marshall’s) are somewhat more engaged than the average visitor, but their “start the purchase cycle” numbers are pitiful. If Eric were to put effort into this, the place to put effort in is where both those metrics are strong.

Eric was good enough to send me the spreadsheet, and I pushed the numbers. (Well ok, technically he and I pushed the numbers at the same time over the phone …) On the phone, he called it “Robbin’s Metric” and I left it that way. It is the product of his Visitor Engagement and Percent Buy Path Sessions:

[ Ugh! Yes, I know that image is hard to read! I will correct ASAP!!! ]

By multiplying the two metrics and then ranking all the referring blogs by that metric, you see where Eric should put in extra effort. I agree witrh the first conclusion that Eric already came to in his blog, i.e. that he needs to work out some kind of deal with Anil. However, the two blogs where he should put time/effort would be Justin Cutroni’s and ROI Revolution’s. Interestingly, they are both Google Analytic blogs, so there is a decent chance that the reader is newer at analytics and probably could really benefit from Eric’s books. I didn’t highlight Steve Jackson’s blog, Xavier’s or Aurelie’s because they are already converting well (if there is such thing as converting well.)

Finaly thoughts: An engaged visitor to a site that is also content rich, like Eric’s, doesn’t necessarily make a good customer. In fact, Clint did a survey on his blog and saw that many of his visitors already own many, if not all, of Eric’s books. When visitors go to the beginning of the checkout, we can actually see interest in the purchase, as well as interest in the content — and as direct marketers know, you should always pursue the customer who already has a propensity to buy.

[ Thanks to Robbin for taking the time to take visitor engagement to the next level! What do you think? Is Robbin on the right track? Did I miss the mark? As always, your comments are greatly appreciated! ]

Posted Monday, March 5th, 2007 | 10 responses | Add a Comment | Share, Save or Email

Clint

Robbin/Eric,
Interesting post. I was wondering though if you could explain why multiplying pct Engagement and Pct Buy Path Sessions is a valid transformation? There’s a little buzz in the back of my head that says that something’s not quite right. I ran a correlation test and the two metrics don’t seem to be horribly related to each other and that makes me wonder if the results aren’t just artifact. In other words, Robbin’s metric looks significant but actually isn’t.

Anyhow, just looking for the explanation/proof of the metric so that the buzzing will go away.

I can play with math too and it looks even more interesting when I add visitor conversion to the mix by subtracting Robbin’s metric from visitor conversion and multiplying the result by 100 – in that scenario, sites with a large negative score are the places to spend time (and it would add Robbin’s site to the her list of sites to focus on)

eric

Clint: My head was buzzing too, and I explained as much to Robbin. But in terms of generating metrics that help better identify opportunity it seemed, much like the Earth, “mostly harmless” to combine the metrics per Robbin’s suggestion.

Oh wait, are you just complaining because you shot to the bottom of the list? Because remember, we don’t expect your people to drop into the buy path ;-)

Seriously, Robbin made an excellent point that perhaps my best opportunity vis-a-vis visitor engagement and desire to purchase was with Steve, Anil, and Justin. Now all I need to do is find enough time to reach out to Steve and Justin (Anil and I have been talking) to see if I can make something happen … unless Steve and Justin are reading this, and if so, call me!

Thanks for your feedback Clint.

Clint

Eric,
I wasn’t worried about me (personally I think it’s amply proved that my readers already have your books). Just that funny feeling.

I also think that Robbin’s visitors represent an excellent opportunity and wonder why she wasn’t on the list. ;)

Ian Houston

Hi Eric,

There are a couple things bothering me about this too.

How does 21% Buy Path Sessions to 0% conversion for Anil’s referrals compare to the site as a whole. I don’t know that this tells me the opportunity to improve sales. Is it about average, worse than average, better than average?

Since I don’t know from the above the relative volume coming from each site I would be hard pressed to pick focus points. Relatively speaking, if Anil’s traffic only represents 1 user out of 100 from say Robin’s blog I would actually look into how I could improve conversions from Robin’s blog first.

Also, to further one of Clint’s points, none of the three metrics really correlate strongly to the others. It doesn’t look like engagement is a real strong indicator of someone’s propensity to buy.

I’m just trying to wrap my head around how the above really helps me to decide where I should focus my efforts. Feel free to prove my concerns unfounded by convincing more of the traffic you get from Anil to convert. ;)

eric

Hey Ian! You make an excellent point, that I didn’t provide absolute volumes of traffic for any of the bloggers in question. While I’m happy to share some of my data with the world, I prefer to not share all of it. You know the drill.

I’m going to kick your comment over to Robbin to discuss correlation since I too didn’t see the correlation and prefer to treat each metric as separate and view them together “in context” instead of “in combination”. Still, it’s Robbin’s Metric (says so on the column header) so we’ll see if Robbin will comment on your comment about Clint’s comment.

Thanks as usual for the feedback!

Robbin Steif

Am sorry that I didn’t comment earlier. Crazy crazy, I think everyone in our industry is crazy, amazing that we have time to blog.

1) Heads that buzz. Right, Clint, that’s pretty much what Eric said when I brought this all up. But the point wasn’t to take two nicely formed metrics and multiply them and cross out the matching numerator – denominator stuff and come up with another beautiful metric. Instead, I just looked at this and said, there needs to be a way of pulling buy-path sessions and engagement into one metric, since the visitor from a blog/site that has high “scores” in both those areas should be the one that Eric would focus on. Maybe a discount to the readers who are clearly going to the path. This is in contrast to say, your readers who are highly engaged but don’t go to the buy path, probably because they already have the books.

2) How important is engagement? Good point Ian, if Judah Phillips were in this discussion, he would insist that I pull out my minitab. Am sorry you guys are speaking during the same time slot I am. Well, since engagement is made up of about seven (?) variables that Eric made up, probably some of them matter and some don’t. I don’t have to pull out minitab or even excel to see that the correlation isn’t awesome. BUT, that is because of external factors, like Clint’s readers who already own the book. Clearly, an engaged visitor who never even starts to buy is not as likely as one who does start to buy.

It’s when I look at the results of my “analysis” (if you can call multiplicatio of two numbers an analysis) that I really begin to believe the metric. It makes perfect sense that the readers of ROI, Justin (Epikone) and my blog would be the customers that don’t have the book and are just taking tentative steps toward buying it. As strange as that sounds for something that costs just a few bucks. Hope that helps. (But go ahead, say that I’m nuts. You can even say it at the Summit, and you won’t even have gotten first dibs on that right.) Robbin

Digital Media Analytics » Blog Archive » Pitiful “Start the Purchase Cycle” Numbers…

[...] More recently I spotted Robbin’s guest post on WebAnalyticsDemystified, and thought I’d weigh in. [...]

Judah

Robbin and Eric,

Cool stuff indeed. Like Ian says, knowing the relative traffic volume referred by these sites would be helpful. I would also like to see how the data trend over time.

The challenge with proving out the very interesting “Robbin Metric” is twofold: 1) the dataset is very small and 2) the arrays “% Buy Path Sessions” and “Visitor Engagement” have a small negative correlation (-0.2633), which means that as engagement increases, the % Buy Path Sessions” decrease (and vice versa). A reason for this negative correlation could be that a majority of “engaged” visitors referred by those sites have already read Eric’s fantastic books (and thus have less “% Buy Path Sessions”). That seems to make sense and, if true, would help validate the Engagement metric.

If we look at the correlation squared (.069) and multiply it by 100 to express as a percent, we find 6.9% proportion of variance in common between the two arrays. Interesting!

I think I would concentrate on the following data for my frame:

*Cutroni: 0, 14.6, 28.6
*Xavier Casanova: 2.4, 13.3, 28.6
*Aurelie Pols: 2, 13.3, 27.1
*ROI Revolution 0.6, 14.8, 21.1

At a very high level, and from a different angle (that of a person during a coffee break)… I tend to agree with Robbin’s hypothesis, based on her metric, that Eric may want to reach out to ROI Revolution and Cutroni to compare their audience persona/metrics with that of Pols and Casanova.

If the revenue opportunity were great enough, Eric’s time would be well spent determining why ROI Revolution and Cutroni sessions enter the funnel more often but convert less often with lesser engaged visitors than Pols or Casanova…

eric

Robbin: I think the only thing I would add is that where you say “made up of about seven (?) variables Eric made up” … I didn’t make any of the components of the visitor engagement metric up. They’re pretty standard key performance indicators, treated as indices, and combined in a way that I believe to be meaningful and useful.

Ticky-tack, I know, but worth correcting.

Judah: Excellent analysis! Now I know why the guys at Unica love you so much … I have a hunch about why ROI and Justin’s visitors enter the buy path but don’t complete, but I need to design a good controlled experiment to test the theory. Watch this space!

Thanks to both of you for your comments!

Increasing your Website’s Conversion Rate » Blog Archive » Read my guest post (on Eric Peterson’s blog)

[...] So, I hope you will read my guest post on Eric’s blog. (We even named a metric in my honor…) [...]

Add a Comment