Give Me Back My Wrecking Ball

Reading Eric Peterson’s devastating remarks on Brandt Dainow’s Google Analytics post reminded me a little of watching the explosive demolition of an old, rickety building. The spectacle is compelling, and while you know that the thing has to come down, it’s almost a little sad watching it implode.

I felt a little guilty. As if I’d loaned Eric my wrecking ball for the day!

I toyed with the idea of commenting on that post as well, but I really couldn’t think of much to add. Still, it got me thinking about Google, what they are really all about, what their impact on analytics is likely to be and, ultimately, why it’s even kind of understandable that some people see GA as a Killer App.

Google’s an amazing company, obviously. But what is that makes them amazing? The key to the Google enterprise certainly looks like Search. Search not only drives the vast majority of Google revenue, it’s also the primary brand identity. It’s funny though, because in some ways Search is the least impressive part of what Google has accomplished. I think most serious net users would agree that while Google did once enjoy a substantive advantage in Search technology that advantage has pretty much evaporated. Competing engines are largely equivalent in quality and often substantially better in UI design. But instead of chipping away at Google’s search dominance, most rivals have actually lost ground since achieving technical parity.

On top of the Google Search System resides AdWords – and this really is the goose that lays the golden eggs. Of course, it wouldn’t be worth anything at all except for the market in the Search system underneath it. But AdWords did break some pretty significant ground. The bid system was superior to Overture’s – both in terms of creating an efficient market and in maximizing Google revenues. In addition, AdWords was very, very easy to get started in; much easier than Overture. This is a consistent theme in significant Google products – they were usually important because they extended the reach of a product more than the technology.

Next up for Google was the AdSense Content Network. Once again, Google didn’t really break any technology ground here. Their content matching is awful. And the system is rife with fraud which they appear to have done nothing to prevent. But Google made it easy for publishers to capture revenue from the vast pool of Google AdWords buyers. So even though this system is, in some respects, very poor - it’s still a massive revenue generator and a considerable success in other ways – mostly having to do with reach.

Google really proved that they could broaden out beyond their basic Search/Advertising niche with Google Mail and Google Maps. Gmail is a nice but hardly revolutionary technological achievement. Its biggest impact was in the fact that it was free (reach) and that Google’s ability to scale left it without crippling restrictions around storage that plagued similar efforts. Google Maps is a very clean implementation of mapping with some gee whiz components. But a good part of what really makes it compelling is the API. This, once again, was a case of Google expanding the potential reach of something by opening it up to a wider audience.

Google Analytics is very much in this corporate tradition (all of which is made possible by Google’s profound understanding of how to scale technology – the knowledge that is, ultimately, the secret sauce behind most of what Google has accomplished). I suppose just about everyone would concur that the initial rev of GA was a flawed, me-too, low-end product (especially since it just was a flawed, me-too, low-end product they acquired). But it had, nevertheless, a profound impact on the marketplace for two very good reasons: it was free and it was Google.

Unlike the creation of an advertising marketplace, however, it’s less clear that advantages of scale necessarily translate into any customer advantages. Does it make much difference if 100 people or 10000 other people are using your analytics package?

I’d like to say a little yes and mostly no. Unlike an advertising marketplace, where penetration and scale are critical advantages, in web analytics they aren’t direct advantages at all. But there are secondary advantages to using a widespread tool – the most important of which is the potential availability of analysts.

At the moment, this advantage is more apparent than real. The vast majority of Google Analytics implementers and users aren’t doing serious analytics, and are only casual users of the tool. In addition, while a very healthy consulting marketplace has sprung up around GA it isn’t as large as might be expected because the companies most likely to hire consultants are much less likely to be using GA than "enterprise" analytics solutions.

Nevertheless, there’s certainly an advantage to ubiquity and the GA advantage in analysts might eventually turn out to be pretty significant. For now, though, it’s hard to see that it provides any measurable advantage at all if you’re a significant enterprise - probably the reverse.

Still, if Google does end up a power-house in the enterprise market-space, I think it’s more likely to be because of this than anything else. The "free" aspect just isn’t that compelling. Enterprise buyers understand the concept of "total-cost-of-ownership." They know that people are almost always the biggest expense in anything they do – and that giving people the right tools is essential to getting good returns on that expense - so semi-plausible canards like the 90/10 rule aren’t going to cut much ice.

Regarding the product itself, I think Eric’s post is both compelling and on-the-money. The bottom line in the analytic space is that most enterprise tools (Omniture, VS, WebTrends and Unica) now support full OLAP multi-dimensional reporting and unlimited visitor segmentation integrated with that capability. Advanced features like scoring are becoming increasingly commonplace. In addition, most enterprise vendors support a range of data integration options – both internal and external. And these integration options are not only vital to most enterprise clients but are pretty much the battleground on which the tool wars are now being fought.

So it’s pretty much as Eric says – the claims Dainow makes for Google being a killer app vis-à-vis the enterprise class competitors are absurd. But take this down one level – to the class of products that exist beneath the Enterprise tier, and the claim looks a lot more plausible. When GA 2 came out, I wrote that it would put enormous pressure on the low and mid-range WA vendors. It has.

For software competing in the low and mid-market, I don’t think there is much doubt that GA is in fact a killer app. It is functionally very much on the same turf as these other products. In the 2.0 rev, it is better designed and easier to use than most. It’s free. And it’s from one of the strongest brands in the world. If I was selling a low-end WA tool, I think I’d pack it in too.

So I think some of what Dainow wrote may be a matter of perspective. If you’re a WAD or a Semphonic and your clients have traditionally used enterprise class solutions, GA hasn’t made much more than a good ripple on the surface of your world. But if you’ve been trying to sell a relatively inexpensive WA tool to small and mid-size businesses, you got hit with a tsunami!

Will GA ever be a tsunami in our world as well? I don’t have a cut-and-dried answer to that. I’ve heard skeptics argue that most enterprise clients wouldn’t trust Google with the analytics because it’s a conflict of interest with their ad spend. I don’t buy that argument. It may cut a little water – but not much. On the other hand, there do seem to be some real barriers here: both in terms of support, integration and GUI that may not be easily transcended.

I think the potential issues around support and integration are obvious. But the GUI issues might be less apparent. Enterprise-class web analytics tools are increasingly for analysts not for marketers. And the type of GUI appropriate to an analyst may just be fundamentally different than what is going to work for a "mass-market" product like GA.

For example, to get the kind of power in statistical analysis that an analyst wants, you need tools like SPSS and SAS. I doubt whether such tools could ever be "fit" into a mass-market GUI. Of course, web analytics tools in their current enterprise state are well short of that level of complexity. So, to me, it’s an open question whether a mass-market product could also own the enterprise analyst’s desktop. I suspect not, but it doesn’t strike me as clearly impossible.

Right now, if you are an enterprise class customer, there is precious little reason to have GA on your short list of tools. It isn’t, at the moment, a feature competitive tool. And even if you don’t have lots of experience with web analytics, the idea that you should try it because it’s free is deeply flawed. The effort invested in implementing, training and acculturating your users to a web analytics tool is not less (and probably more) than with most other software products. And it is in that process that the bulk of your costs with almost any piece of software always lie. Nor is it clear (at least to me) where this tool is headed. I know where Omniture and VS and WebTrends are taking their products. But what’s the Google vision when it comes to large enterprise analytics?

With issues around documentation, support, integration, product capability and product roadmap, GA hardly seems like a compelling bet at the enterprise level. That doesn’t mean those issues won’t get resolved. Google is a company that has surprised us before and clearly has the intellectual and financial capital to compete – and even dominate – in the enterprise web analytics market. But just as "might" doesn’t make right, "might" also doesn’t make "will."

In the end, I don’t think I’d describe an Enterprise that went with an inferior analytics tool just because it’s from Google as "bold." Instead, I’d probably use words like "risky," "foolish" or "misguided," because in the world of enterprise software, the world of "might be" has almost always proven to be an illusion.

Comments

[Eric T. Peterson]

Gary,

I will gladly send back your wrecking ball since it's not really my style to do a tear-down job like that. But sometimes someone writes something so dumb and so mindless it just makes my innards go "pop!" and I am compelled to say something.

Google Analytics is a freaking great tool for a certain audience ... but little more. Compelling, yes. Visually appealing, yes. Best at its price point, yes. But appropriate for some of the companies I see deploying it out there? Nope.

Which makes me wonder who is giving them the "get Google Analytics, it's free and it's totally great!" advice? People with vested interests, for sure, which is ironic because one person wrote me after the Dainow post (and the free vs. for-fee report) asking if I held shares in OMTR and VSCN and was just trying to run them up (as if I have that kind of clout ...) I don't, for exactly this reason.

I learned at JupiterResearch that being beholden to a company or an idea is very dangerous. When I was at Visual Sciences, of course I loved the application. But I was ** paid ** to love the application. And when you're paid to believe something, or write something, or say something ... that's what you do. That's what a product manager does, and what an evangelist does. Easy!

Still doesn't explain Dainow, but I suspect that too will become clear in time.

Anyway, great post.

E.

[Avinash Kaushik]

Gary: I really enjoyed reading this post. I think it was becuase it was a thoughtful dissection of the current state and presentation of your cogent perspectives on possible outcomes and scenarios.

I might quibble on a point here or there (especially your dig on "semi-plausible canards"!! :), but that does not distract from a well written post and a depersonalized presentation of your perspective.

I greatly enjoyed reading it.

Thanks,

Avinash.

[Chris Grant]

Excellent viewpoints.

Regarding your statement "the idea that you should try it because it’s free is deeply flawed" ---

Because GA seems to be paying more attention to its GUI than to connectivity, it requires a lot of clicks to get to anything followed by copying data into summaries. Some of us, on the other hand, are working with post-processed data pulled automatically from many individual reports in enterprise tools. So our GA-using clients end up paying a lot more for mid-level weekly report summaries than they would pay for the one-time implementation of really interesting processed stats using a fee-based tool.