Whenever a potential customer converts into a lead, I always look to see, how did they find out about us? (Well okay. Sometimes I am better than others. Especially when a lot of customers convert and then I just want to see trends.)

But it was a weekend. Quiet. So when one of the conversions came through, I went to my Google Analytics to learn who he was and how he found out about us. I started by looking at Marketing> visitor segment performance> geo location and saw that the conversion was in the US (go figure.) I kept drilling down by hitting the plus sign until I learned that the conversion came from Delray Beach. (Notice the 100% under the G1, Goal 1, next to Delray Beach.)

Then I segmented by Source to see where he came from, and it was a link from my blog! So I went into my blogalytics and did the same analysis, looking for the Delray Beach visitor, but this time, when I segmented by referring source, it was a search from Google.

Above you see how I do the segmentation, and below you see the result:

So I did the same keystrokes, looking at all the visitors from Florida, but this time, segmented the Delray Beach visitor by keyword. He had typed in, “website conversion funnel.”

Hmm, I thought, I didn’t even know that LunaMetrics ranked for that term. So I tried it in Google, and after I signed out, I saw that my blog (but not my site) came up on page one.

At this point, I felt like I was on a mission. I went over to ClickTale, which I installed on my blog but almost never use. (For those of you not familiar with it — it tapes the user session.) I rarely use it because I don’t learn much from it and it takes a lot of time. Mostly, I watch people scroll and it gets very tiresome, very quickly. But this particular one was fascinating, not because he did so much (he mostly scrolled), but I was able to see where he scrolled quickly (not interested) and where he took his time.

So now, I know that the customer is from Delray Beach, Florida, he is concerned with website conversion funnels, he read my six most recent posts on my blog as well as my post on “What Do Web analysts Do All Day Long?,” he completely skipped all the fun parts in my “Dear Sexy Web Analyst” post and went right to the Google Analytics. What he read was interesting enough to check out my site, and what he saw on my site was interesting enough to submit a form. And then I met him on the phone today. Sort of like Phil Kemelor’s 2×2 matrix — here was someone that I know both online and offline.

Robbin Steif

Eric Peterson’s free webcast for Web Analytic beginners is coming up, this Tuesday, March 6. So this also seemed like a good time to (finally) write about Phil Kemelor’s almost free ($25) e-book, Intro to Web Analytics. Phil, remember that most PR is good PR. And then when you get to the end of this post, remind yourself again.

I couldn’t get ten pages into the .pdf without strongly disagreeing with some of Phil’s basic premises:

1) I disagree when you write: “For your organization’s Web site, there is only one metric that matters: How much money is it making?”

Yes, Phil, you are often right — most websites are ultimately about making money or saving money, directly or indirectly. But more and more, people create websites just to change the world. Look at Wikipedia. Sure, you can argue that they accept donations now and need to measure that. But their donations are ancillary to their real goals. I’ll bet that their website’s most important metric is, “Did we help people today?” or “Did more Wikipedians feel like they are a member of our community?” OK, maybe you don’t like that example. So, how about Mozilla? Sure, they have a few revenue streams too, but I’ll bet that they use them only to stay afloat and pay for their server and executive directory. Like Wikipedia, the vast majority of their “employees” are volunteers, and their goals are not financial. So, Phil, when you write, “The purpose of your web site is to make money or save money,” I just have to disagree.

2) I disagree when you give a list of ten lousy reasons for people to avoid measuring and testing their site.

Yes, you are right — there are no good reasons to avoid measuring your site. At a very basic level — with the server side analytics package provided by your web hosting company — it is simple and free. But testing is different. I created this post on Increasing your Conversion rate for $1524, and the largest cost went to testing. Sure, you can do user testing for an hour or so (one day I user tested a landing page through three airports and two plane rides). You get awesome qualitative data and it helps you locate major problems. On the other hand, there are lots of things people just won’t tell you. Plus, how significant can it be when you are doing user testing with five people?

So when someone says (and now I am quoting from your e-book), ” “It will cost us too much to measure and analyze site use,” they are really saying, “We don’t get enough value from our website to invest a thousand (or ten thousand) dollars in it.” This is exactly how I feel about my company’s brochure. Resources are always constrained, and companies have to invest where they get the highest return. Websites make your business happen. Websites make my business happen. But one size doesn’t fit all.

On the other hand, I read all the way to page 70 without disagreeing again, and by that time, it was just a technicality. So, for twenty-five bucks, pick up a copy of Phil’s ebook and see if I am wrong.

Robbin Steif
LunaMetrics

So you did everything Google told you to do, but you can’t get your AdWords to talk to your Google Analytics.

This question came our way this week, and it was interesting, because I just went through the same problem myself. So here is some advice.

Problem #1) You want to link AdWords and Analytics, but when you follow their directions, you don’t get an “I already have Analytics” option.

I learned this one the hard way. Because you can only have one AdWords account associated with your GA account (even though you might have multiple Analytics profiles and even be able to look at multiple Analytics accounts), you have to unlink the old AdWords account so you can get the one you really want. However, unlinking is fraught with problems. (Thanks Justin, for teaching this one to me.) In fact, Google even says, if you want to unlink, contact us (use their contact form, they are pretty responsive.)

Problem #2) You successfully linked AdWords and Analytics, but you don’t see your campaigns.

This can be caused by autotagging turned off. You might have your autotagging turned off so that you can tag your AdWords just the way you want, but you have to tag them all. If you want to turn it back on, it is in AdWords, under My Account > Account preferences.

Problem #2a) You successfully linked AdWords and Analytics, you have autotagging turned on, but you still don’t see your campaigns. Or maybe, you see some of them, but not all of them.

When you see some but not all (and you have autotagging turned on), that’s a red flag — you don’t have Google Analytics installed on the landing page. “But wait,” I can hear you say, “I have my Analytics in a file that automatically gets copied to every page on my website.” Yes, I work with a customer like that, and two (count ‘em, one-two) of her campaigns landed on very specific landing page, stripped of all navigation and include files. Hence, stripped of Google Analytics.

Problem #3) No, none of those are the problem. Your account was never linked to another AdWords Account, you have auto-tagging turned on, you have analytics on your landing pages. This was the problem that came to me this week. Finding the problem was real gumshoe work and it was exciting to discover it.

Here’s what happen. I was getting ready to send it to someone on high, but decided to have one last look. I wanted to be sure he had GA on his landing pages. And then I noticed that he did have code installed — but he was running two different Google Analytic accounts. The account that was used from his landing pages was nothing like the account for the rest of his site. (I mean, the numbers, like this: uacct=”UA-10005-8″) So yes, there is a Google Analytics account somewhere, someplace, that is reading the AdWords, but the other account, the main one, can’t see that the clicks are coming from AdWords. And isn’t the whole idea to be able to see it in one place and then make decisions based on your data?

Robbin Steif