Image via istockphoto
Written by Eric Gockel
ClickTracks‘ CEO John Marshall took time out of his busy day for this interview with Eric Gockel via Instant Messenger about cookies, analytics, and Tufte.
Originally Posted May 28th, 2005
EG: OK, let’s hit the ground running: Are you concerned about the increase of users deleting their cookies regularly (or software doing it for them) — doesn’t that impact the effectiveness of your product?
JM: The question should be divided into two, relating to first-party and third-party cookies.
Third-party cookies suffer from the problem you describe; first party do not — so our logfile-based products are unaffected. They are implicitly first party. Also, our high-end product JDC is effectively first party.
For example, Intuit is switching to our technology for that reason. [With] our hosted products, the user has a choice – either first or third.
So I am worried for the small-time vendors that cannot switch off third-party technology, but not us.
EG: You’ve stated in the past that you’re specifically not trying to be a HitBox with all its charts, reports, and “path” data. Why is ClickTracks better than their product, or would you recommend it (HBX) for a certain type of user?
JM: Hitbox is great for executive dashboards. It’s clearly designed to impress the CEO and CIO. Our product is designed for the analyst – the person that actually has to crunch the numbers.
EG: And the interface designer and marketing crew.
JM: Right. Basically, our product is used by people that need to use data to drive decisions. For example, a web designer uses our stuff if they have some basic understanding of the numbers in a small company. That might also be the marketing guy, of course,
and in a bigger company, it’s a function of a department
EG: You mention Intuit uses your product. Does it satisfy the CEO and CIO (and CFO) — or do they even access it?
JM: They don’t access it. Intuit delegates that down to each individual business unit. That’s a philosophical difference. Traditional analytics products assume a top-down command-and-control structure:
“The dashboard tells me X, so tell the troops to do Y”.
Our product assumes a bottom-up structure:
“Hey, boss, I discovered this page doesn’t work, so I tweaked it with red instead of blue, and I improved conversions from 2.23 to 2.37%”
“Great job, junior. Have a bonus”
The top-down companies are too sclerotic actually to do anything with the data anyway, so we bypass them and appeal to the nimble companies who can integrate analysis into the process of understanding customers at a tactical level.
EG: The other analytic software packages have copied your “overlay” approach. Are any changes coming up on your end — or does ClickTrack still have an advantage over theirs?
JM: We have currently several huge advantages:
It actually works. We’ve tried Webtrends and Urchin on demand. We couldn’t make it work.
In Urchin, it’s also worth noting it’s buried inside the product. It should be right up front, but they know it doesn’t work, I expect. This is frustrating for me because it devalues the feature across the industry. It’s another example of “looks great in the demo, fails in real use.”
Another advantage is that with ours, you can view it with segmentation applied; of course, the segmentation is ad hoc. The report is useless without comparing different segments, and I’m not sure *anyone* does this except us. Again, very frustrating. People are buying a product that is fatally flawed if you can’t do this.
Finally, we can cope with Javascript pages and other complexities that others can’t.
EG: Have you seen Google’s new free conversion tracking component for their Adwords? For someone who is purely interested in conversions and not behavior, is this good enough for them?
JM: Yes, it’s pretty good if you are willing to accept the following limitations:
1. No organic search data and no ability to understand if paid vs. organic on the same keyword is better
2. Third-party cookie (see above)
3. You’re OK exposing this data about your business to Google
4. You’re willing to have another separate measurement system for your other campaigns
EG: Aren’t I exposing my business data to your hosted solution?
JM: Yes, and that’s important for customers to understand, but we can’t make you pay more for each paid click because of good conversion. Google can.
I strongly believe that your analytics should be independent. It’s like hiring an auditor. Without independence, it’s a charade.
EG: For site developers, do you have any tips on SEO that might prove helpful beyond the basic ones we already know about? Any practices that we should stop doing?
JM: Hmmm….very interesting. Pay attention to not just the keywords but the keywords for entry pages.
For example, look at your top entry pages, and in ClickTracks, label those visitor segments. Then examine those keywords on those entry pages. Designers don’t pay enough attention to those pages. Remember: that page is the user’s first experience of your site/product/service. NOT the home page!
EG: Can you tell us about your latest baby, Optimizer?
JM: We previously had two products. Analyzer at $495 and Pro at $2,995. Analyzer has the advantage of very simple installation, simple configuration, etc. It’s perfect for small businesses, time-starved consultants, etc.
Pro has more features, but perhaps most importantly, it takes a few hours to get everything set up and configured.
People want the more advanced features of Pro, but they don’t want to install the server code (Pro is client-server architecture, and it’s setting up the server that is the issue), so we took some of the more advanced features of Pro, like the what’s changed report and the robot report and put them into a mid-range product that is as easy to set up as Analyzer, i.e., 10 minutes.
Pro, of course, includes ROI on PPC campaigns, conversion, etc. That’s a key reason why it’s more expensive; Optimizer doesn’t do ROI.
EG: For someone new to web analysis, how do you recommend getting started on their own? Say the VP of Marketing for a small company, etc.?
EG: I think most people have seen some kind of primitive web stats report, analog or similar, by now.
The biggest barrier to getting started is realizing how much web analytics has moved on from this type of thing, the exploratory kind of analysis you can now perform.
It’s so much farther ahead; it’s hard to conceive if you’re stuck using whatever freeware junk your ISP gives you, so getting started first requires that you ignore that stuff. Next, I’d recommend they download ClickTracks Analyzer and throw a log file at it.
I think people also need to realize that the key to good data is to segment users into different groups and compare them.
Often the old-style stats reports are ignored by marketing guys because there’s a huge “so what” factor. I have “n” page views. So what?
On the other hand, segmenting into paid vs. organic search, I have “x” page views for one and “y” page views for the other. Bingo! Organic search results drive more page views, indicating greater interest in my site–that is compelling.
Sorry to pitch my own product here, but honestly, there’s nothing else that’s so easy to get started with, cheap to buy, and that you can simply download and try.
EG: Hey, you’re entitled.
JM: (if there is another product, please mention it, I honestly don’t know one)
EG: Finally, has [Edward] Tufte returned the favor and started recommending ClickTracks? Do you know him personally?
JM: No way! Actually, I don’t think Tufte is a big fan of anything on the computer screen. I always sense that his bias is toward what you can draw by hand in Adobe Illustrator and then print at 1200DPI
EG: Well, thank you for your time today, Mr. Marshall.
John Marshall, CEO of ClickTracks, is a Netscape alumnus responsible for two other successful software companies.
Eric Gockel is the Principal and Founder of cre8, a consultancy providing internet strategy, design, and development services since 1997.