I think it was Jim McIntyre from Visual Sciences who re-ignited the cookie fire with his post titled “Head in the sand” on the Web analytics Yahoo group , prompting 12 or 14 responses from vendors, customers, analysts. I had the chance to discuss this matter with quite a few people over the last few weeks with industry peers, consistently getting the official party-line from their respective product marketing teams.
And that’s why it’s so difficult to sort this cookie issue out for the Web analytics user. Many, including myself, aren’t exactly clear about the motivations of the “experts” on this cookie matter. Until we are more data and clarity on the matter, I think the most productive use of time is to accept the possibility of a cookie deletion problem and quantify its impact. And that should not take more than 10-20 mins.
Don’t be evil but give me some more of your money - like $3-4B.
Forget building an e-Marketing suite. It’ll be g-Marketing. Is that good or bad? Not sure but I’m buying GOOG. 
Update: While on the topic, here’s an interesting ClickZ article about Google testing commercial listings embedded in the organic results.
I had an interesting discussion with a marketing agency on CPM advertising – what is Google set to do in the ad serving space, how much market-share they can hope to get, etc. Then we touched on Web analytics, concluding that one piece of the puzzle is missing: capturing banner views and allowing marketers to do very detailed analysis on the effectiveness of CPM advertising over long periods of time, with their favorite Web analytics tool.
Clearly a complex problem but I think that part of the reason why Adwords is so successful is because the returns are easy to measure and therefore predictable. That’s not the case with banner ads yet.
Ken Novak from Metrist Partners uses the beer and diapers classic to make a point about the future of analytics, read it here.
“Beer and diapers” was a classic nugget from data mining. For all its hype, there wasn’t much there to benefit from.
In mining, you extract a hunk of something that is either durable and non-reactive, or burns for a short time. Plus, you have lots of tailings to throw out. It’s not self-sustaining or pretty.
Personally, I don’t want to mine. I want to do something valuable and sustainable.
Here’s an alternative paradigm: data farming.
[…]
And many of us are already doing data farming: extending successful promotional elements to new prospect lists and new vehicles, learning from last season in ths season’s re-activation programs. This is farming, sustainable and nurturing, not at all like mining. And smart farmers don’t just produce; they learn so that next season will be even better.
Right on!
Last week I spent some time with a new marketing manager at retailer XYZ.com, who just got hired. This is the 3rd marketing manager in less than 12 months - in our discussions it became clear that I was sharing with her some of the lessons her company had learned with previous marketing managers (this won’t work, this has been tried before, why don’t you try this, etc).
I think we all understand the value of industry-wide best practices, but there are site-specific best practices as well that are worth saving somewhere. Here’s an idea: create an internal blog at your company to talk about the lessons learned. Or even better, create a set of documents describing your experiments, the tracking methodology, the conclusion, and the future work - save them somewhere everyone can access, have weekly or monthly cross-functionnal meetings to share your progress.
WebSideStory’s Active Marketing Suite is starting to look pretty good – Analytics, Bid Management, Search and Content Management. I like the idea of Web analytics evolving to be the core of an eMarketing suite. I think it makes sense for the customer – and it makes sense for the Web analytics vendors as they try to grow their business. A true win-win, isn’t it? Well, maybe.
One the customer side, if you had asked a marketer to pick a few modules to build an e-marketing suite, neither Site Search nor Content Management would have been top choices. I’m inclined to think analytics, Email, Bid management/SEM, and affiliate management are truly the modules every marketer would like to be integrated. And to be honest I’m not even sold on (Analytics + Bid management/SEM) because I don’t see any compelling new value created by combining the two - other than having a single vendor and having more consistent reports. That’s slim compared to the advantages of sticking to the experienced and feature-rich AtlasOnePoint for example.
On the vendor side, there obviously are some benefits – higher revenue per client, increased customer loyalty, etc. This is something many have tried, and few have done right because there are some big challenges. I think it’s absolutely essential that a/ the right modules are picked for the marketer, and b/ when combined, the modules get you the 1+1=3 kind of effect. And yes, doing all this while staying focused on the core business. Hmmmm… the DoubleClick ghost is playing tricks on me tonight.
… to know:
a/ that this site is losing tons of people on this page
b/ what to do to fix it
Which leads me to a more important point - as a Web analytics vendor, when I’m asked to “fix” a site, I now spend at least 2 or 3 hours browsing it with my browser overlay tool open (or not), taking lots of notes. For example:
- Home page seems heavy, check display time
- Checkout button not obvious, check add to cart rate
- Password selection clunky, check drop off
ETC
Then (and only then) I look at the detailed reports. In my experience, it’s the best to wear a consumer hat first then switch to being an analyst. Many readers of this blog know their sites inside out, so I guess my suggestion to them would be to get their hands dirty once in a while.
