Tuesday, October 20, 2009
Amy Chang, Group Product Manager for Google Analytics, notes that when GA was bought and launched by Google, it was "all about democratizing web analytics," bringing more powerful web analytics functionality to a wider audience of businesses than ever before. Mission accomplished. Books will soon be filled with tales of Google's awesome data management capacity (they'll be dry books... but books, nonetheless). Providing such a massive exercise in computing to so many millions of customers -- with multilayered segmentation capabilities -- is a task nearly no competitor could accomplish, least of all for free.
Today Google continues its aggressive pace of development for GA, releasing a long list of enhancements. They'll delight power users and laypersons alike. Chang notes that the more customized (as opposed to customizable) features continue to be part of the trend of analytics moving out of its old role as the sole preserve of IT departments, and putting intelligence directly in the hands of marketers and managers.
On the power user front, we get more custom variable capability. This was available in the past, but it's becoming easier to use. To take an example, GA is now making it easier to track segments like session attributes (logged-in vs. not-logged-in users) or visitor characteristics (member vs. non-member, or even demographics like gender). Don't expect setting this up to be easy, but it's getting a lot easier.
For the rest of us, Analytics Intelligence is evidently the culmination of extensive product development. Rather than having to customize a "what's changed" type of report yourself, or having the canned report be limited to vanilla categories, this alert functionality is tuned so that it's better at predicting which change alerts are the most likely to cause a marketing department, specifically, go "eeek!" (or in other cases, trade high-fives around the M&M's dispenser). Note the slider in the screen shot below: you can set the "alert sensitivity" to "high" or "low". As Google aptly notes: "Now, you can spend your time actually taking action, instead of trying to figure out what needs to be done."
Other new toys include the ability to set 20 goals instead of 4, and custom engagement metrics considered as potential goals, as opposed to clunky custom segments you had to cook up from scratch.
And that good old IT person who formerly had the only set of keys to your marketing statistics, and sent them to you in whatever format she felt appropriate? Expect a lifelike facsmile to be staring vacantly into the camera for the Official GA Photographer, looking every bit like the lonely Maytag repairman.
Labels: google analytics
Wednesday, March 18, 2009
It is difficult to overstate how much of a data-driven culture Google maintains. And right on the heels of that aspect of their philosophy is their commitment to providing more of everything for free.
Recently I heard a bit of a different spin than I was expecting, casually chatting with a client about our recommendation that they finally switch over to Google Analytics. The reservations weren't coming from the expected places - the IT team not liking the idea, or the fear that Google would have access to private data. Rather, he said he fully expected Google to get people hooked on the free offering and then slap a hefty monthly price tag on it. Hmm! Hadn't heard that one. (And Google does offer a fuller-featured product, with pro versions of Urchin.) I opined that Google wouldn't be charging for GA: they might stop adding massive numbers of features, but there will be quite a bit left over in the free version whichever way the wind blows. And if you're an advertiser... you definitely don't have to worry.
If you gave me $500 and told me the winning number on the roulette wheel in advance, I'd probably fail to put a chip down. I'd shove the $500 in my pocket and make a dash for the nearest door at the casino. In other words, I'm not generally a betting man.
But I'll put a chip down on this one. I think it's unlikely that Google will raise prices on free Google Analytics for as long as you're running your business. (As long as you promise to sell it in ten years or less.)
Today we're seeing word that Google is offering a whack of new stats on YouTube. Instead of just seeing how many views your video got, you can look at where people viewed from, user ratings by geography, and more.
So is this functionality only available through some special program for advertisers? Nope! It's open to anyone with a YouTube account. Anyone who uploads video.
Google's data-driven culture and their compulsion to provide more services for free: the trend continues with this release.
Liz Gannes comments from the entertainment industry point of view, but in the end, this announcement is for every YouTube user that uploads video and wants to understand their audience better.
[Edit 10:51 EDT: link to item changed from outdated link to today's announcement on the YouTube blog.]
Labels: google analytics, youtube
Friday, December 12, 2008
For this client, using the "custom segment the most engaged users" method, I isolated the "engaged decile" of site visitors who view many pages and remain on the site for a long time. In the past month, that "decile" is 11.9% of visitors, but anyway... what is remarkable about this group is that they make up well over half the pageviews on the site! In fact, this 11.9% accounts for 56.6% of pageviews, leaving the remaining 88.1% of visitors to generate a pathetic 43.4% of pageviews. Overall, engagement on this site is poor. Many visitors are distinctly unimpressed. The "average" visitor views three pages. The "average visitor" of that engaged sliver views an impressive fifteen pages! (Average time on site: 1:58 vs. 13:22, a difference of more than 6X.) So that sliver of interested folks diverges sharply from the crowd. More proof: your tribe matters a great deal. The rest of the world doesn't give a hoot.
Labels: google analytics
Wednesday, November 26, 2008
I'm most of the way through Seth Godin's great new book, Tribes.
It's a book about leadership, and the need for leadership at every level to achieve meaningful change in any setting.
It's also about distinguishing between a passionate group that can "go places" and achieve something together, as opposed to a loose aggregation of people who don't really care enough to matter.
Great changes can be achieved by leaders with faith in their vision, followed by a relatively small number of people who help make it happen. One of many examples Godin gives in the book is the animal shelter activist who was able to make San Francisco into a No Kill zone (typically, 70% or more of healthy animals in shelters are eventually destroyed). He went on to replicate the feat in upstate New York. In this case, many people in established agencies actively resisted the initiative, and many others in these locales simply didn't care. The ones who did care made the difference.
Thinking about this I can see many parallels in business and online community, in things I see or do every day. Take Avinash Kaushik's common-sense exhortation that "aggregate" is never the name of your website visitor. If you get bogged down in aggregate statistics, you might be overwhelmed with just how many loosely-engaged, "valueless" clickthroughs come to your website every week. Yes, but why not make an exercise out of ignoring the 80% of people who aren't connecting with you and zero in on just the 20% of those who do? Study their characteristics. Build and grow with them. (And it's easier than ever to study them. This week, using Google Analytics' custom segment features, I hand-built a segment called "engaged quintile," for the 20% of website visitors on a client's site that stayed a long time and viewed many pages. By definition, guess what the "bounce rate" was for that segment? Yes, it was 0%! It's heartening and inspiring when you look at life through that lens.)
I've also been noticing how some online communities take off, and others have much more trouble doing so. I think that's because we often wildly overestimate the importance of numbers, and underestimate the importance of engagement, as Godin says. Even in communities that are supposedly in a niche, we wind up with a mass, semi-engaged group that never takes anything and runs with it. One online community of 2,000 people (sounds like a niche, right?) spins its wheels and goes nowhere, while another of the same exact size takes off. The second one doesn't take off because 2,000 people are doing stuff. It takes off because the core leadership is making this place their main mission in life, and an outer core of maybe 20-30 enthusiasts make it their mission, too. Most everyone else joins in over time because of that passion displayed by 20-30 people. Momentum starts with leadership, and a tightness and clarity of mission that breeds passion.
The first group is doing what Godin calls sheepwalking. The second is on a real mission.
This is also why broad online content plays like Yelp take so much money to get off the ground (and why InsiderPages, Judy's Book, OurFaves, and maybe MojoPages are fighting an uphill battle, or have already run aground). They're very wide, so nothing is happening in a lot of parts of the network. Yelp managed to roll ahead despite their breadth because of the passions of Yelpers, and the relentless effort to build community offline as well as online.
But other, similar communities just have a much tougher time of it. It goes back to another Godin recommendation, from Unleashing the Ideavirus or perhaps somewhere else, to pick a hive you can dominate (read: a genuine, small, focused niche). Weak passion spread across a wide field provides little sense of mission.
TripAdvisor got lucky, and maybe fooled quite a few copycats into thinking they succeeded because they were in a "niche". Sure, travel was a "niche" in 2001. But it's actually far too wide for most aspiring community builders to attempt today.
If you talk to a venture capitalist about a user-generated content "idea", they might ask you about what major cities you plan to roll out in. I hope you expect to raise rounds B and C to the tune of $20 million or more in total, because on a shoestring ($1-2 million) budget, you can dominate maybe one or two major markets, if you throw all of your passion and resources at them. Angie's List started off in Columbus, OH, and expanded slowly. They kept in business for 17 years before some Web 2.0 guys decided to pony up that big $20 million+ to help that business "go big." But the majority of the action is in the hives Angie's List was able to dominate: perhaps six or seven midsized markets. Few large ones. Many markets remain weak, with few homeowner reviews, memberships, etc. And Angie's List is considered to be wildly successful, relative to competitors.
What if you were actually starting on a shoestring as Angie Hicks did, with less than $1 million in capital? Probably you would be best off finding passionate groups of homeowners in midsized markets, and trying to get passionate leadership growing those markets. Try doing the same in a half dozen even smaller markets, even, in metro areas of half a million people. That's sort of how Angie did it, and that's how she grew passionate followings (tribes) of homeowners in the markets she does dominate.
Eventually, building a brand in that way, national dominance is indeed possible. But for the time being, forget New York... unless you really know what you're doing.
Amazingly, Angie's recent round of investors expected some of the proceeds to be used for European expansion. Insane. If you're the size of eBay or Amazon, that makes sense. It makes no sense for a company of mid-sized American tribes like Angie's List.
Meanwhile, I'm working with a niche (very small) ecommerce site that also focuses on community. It is in only one segment of travel, evaluating a particular type of vacation. On this site, in spite of the fact that they are open about the fact that they're trying to sell you a trip, the visitors are extremely engaged. Passionate people are writing reviews. And the "engaged quintile" is spending 10-15 minutes on site, on average looking at 14-15 pages on the site (regardless of their geographic locale or other characteristics). (And did I mention that 0% bounce rate in that quintile?) I believe the reason for the surprising level of engagement with this startup is that they truly understand what a passionate niche is today, and they mean to provide true leadership to that niche.
A niche like "travel," or "bars and restaurants" sounds like a "niche" to the uninitiated, but unless you've got bottomless pots of money, remarkable and unflinching leadership, a core of 20-30 passionate co-leaders, and a really different hook, those "niches" today are just far too broad to maintain and grow a passionate tribe.
It's interesting that we were able to reach all of these conclusions without a single mention of technology, or "feature sets," isn't it? Makes ya think. At Yelp, Angie's List, Epicurious, Craigslist, and Flickr, the technology ranges from clunky to cutting-edge. But it's always in service of the tribe and its passions.
Labels: google analytics, leadership, marketing, seth godin, web analytics
Wednesday, October 22, 2008
Enterprise class web analytics for everyone? Who could argue with that? And who could have predicted this ten years ago?
For this camper anyway, this is the most significant announcement to come out at e-metrics summit here in Washington DC. Advanced reporting that gets closer to the enterprise class feature sets available elsewhere. Talk of a new API is welcome, too. If using third party apps to create multi-source reports for clients, it will be easier to integrate the rich data from GA.
It's all good.
Edit: In the Big-Brains-Take-on-the-News-for-Dummies Dept: On the elevator, I just talked with the co-founder of a leading analytics company and asked what he thought was the most important thing about this announcement. "It's the segmentation," he replied. "Doing it well is so incredibly hard." Peering into interfaces, it's all too easy to think stats are stats and they're easy for software vendors to serve up to you. In reality, with less powerful analytics products, "you can't get that info" is the answer to tough questions a marketer might have. Particularly intriguing is the segment "only visits that converted." It should be dead easy to check out whether most of your sales conversions came from the East Coast, or whether buyers were more or less likely to be using Firefox as their browser. Knowing more about your buyers is vital, but outside the capabilities of garden variety analytics. That, in a bit more than a nutshell, explains the pithy reaction "It's the segmentation."
Or in the words of Avinash Kaushik: "Analyzing data in aggregate is a crime against humanity." For his reaction to the announcement (some of the feature set, he no doubt had significant influence over!), see this post at Occam's Razor.
Labels: google analytics
Monday, March 24, 2008
If you delve into Google Analytics reports by geography in my neck of the woods, you get some curious results: places that are really neighborhoods are classified as municipalities. Places that are only distant memories, officially speaking, are still ontologically in your face in GA reporting.
For example, one site I work with gets a lot of traffic from Etobicoke, Malton, Weston, and Islington. If we're laying out the information architecture of our website, should we perhaps use Google Analytics as a guide in planning? Definitely not.
Problem: Etobicoke, once a town near Toronto, then a borough of Metro Toronto, and finally, a mere ward district and place with a name that is meaningful from a real estate standpoint, is definitely not a city or town today. Malton is a mere "neighborhood" within Mississauga, although it might once have been a town, city, postal unit, etc. etc. in the distant past. "Islington" is a nice name for a certain intersection and surrounding areas, in loving memory of many decades ago when Islington was a town and postal unit. Rexdale and Weston are much the same as all of the above. The problem is compounded by the fact that many users are still associated with national ISP's like Rogers and Sympatico that have IP addresses assigned to these locales.
Not only isn't GA hip to the subtleties, it's using designations that don't exist and phantom names-of-things that recall fond memories of malt shops and filling stations from the 1940's.
We can all agree that an information architecture on a site like Toronto Life should be as flexible as possible, and include neighborhood names and informal names. (I personally get a great kick out of people from West Queen West who write letters to the editor decrying Toronto Life's version, Queen West West. Ha ha! I'll say it again - Queen West West! Queen West West! Hope that guy's reading.)
If it's neighborhoods you're after, there are certainly databases of neighborhoods for every city and town out there through major data providers. But if you want to know the difference between a neighborhood, a ghost town, a real town, a city, or a metropolitan area, don't look to GA. It hasn't a clue.
Rather disappointing. Beyond that, for people that don't re-examine their geographic assumptions, they're likely to get tripped up when using the information for real-world purposes.
In general, geography is a lot harder than it looks at first glance, especially when you're building a website. Looks like that could be an interesting panel for the Local Search tracks.
Labels: google analytics, local search
Tuesday, May 08, 2007
Apparently, ceiling height alters how you think.
Not surprising. On a related note, I'm pretty sure that the formats offered by your web analytics reporting affect not only how you think, but how you discuss among yourselves in your company. In many companies -- admit it -- discussion revolves around the easiest reports to generate. Shouldn't it revolve around the information that relates most closely to profitability and to your key user behavior goals?
For most analytics packages in the past, the effort required to implement segmentation often meant that you made a conscious decision not to think about certain things, until someone in your company sets it up so you can get the information you need.
Speaking with Brett Crosby of Google yesterday, I was delighted to hear about all the new features that are being rolled out in the revamped Google Analytics. One big one is that they'll offer by default more reports that online marketers typically want, rather than forcing you to go in and customize in order to generate those reports. As Brett said, the old version used to be a bit like a 747, and the new version, well I forget the car he used for his example, so let's say a 2008 Subaru Legacy GT. Pretty easy to drive, but if you want to impress, there's a sporty leather-wrapped steering wheel just so you feel tough and quantitative like you did before.
This will take some time to digest. I'll comment more extensively as I get comfortable with the platform myself.
Labels: google analytics
View Posts by Category