Scorched Earth in San Diego

It seems like all of Southern California is ablaze. And I don’t just mean the hellish fire consuming homes and causing epic dislocations. That’s a far more serious business than any we pursue in the normal course of our business lives. But the acquisition of Visual Sciences by Omniture completes the work of self-immolation pursued by the VS management team. They can put their gasoline cans away – their work is done.

If ever there was an example of one company being out-fought and out-thought at every turn by a rival, this has to be it. And the end result – this acquisition – only accelerates greatly the verdict that the market was increasingly making anyway.

But the wider implications for the analytic world will need much time to really see and understand.

Let’s consider first the customers of Visual Sciences. With HBX essentially in sunset mode, they will have to transition to Site Catalyst. This is far from a tragedy. Many of them were already considering such a move given the struggles at VS.

What’s involved in that move? Changing from an HBX tag to an Omniture tag is neither a complete slam-dunk nor a gigantic hurdle. The tags are structurally similar. And many decisions made with respect to HBX tags will carry over into the Omniture world. But this will take work. Don’t believe that you can use an automated translator and all will be as it should in the new world!

If you intend to actually get value out of the switchover, some portions of the tag need to be carefully re-thought. Some areas that have the biggest and most fundamental differences are Campaign Coding (especially if you were tagless), eCommerce coding and custom variables. Site Catalyst is much richer when it comes to custom variables – and you should take advantage of this when you transition or you’ll be crippling your implementation.

Beyond tagging, the two biggest concerns will probably be around reporting (all your ReportBuilder stuff will have to be moved and that WILL be a hassle – and you’ll need to reinvest in some training) and active segmentation. Site Catalyst ASIs are not the same as Active Segments. Well, they are. But you can’t use them the same way and if you try you will be sadly disappointed. Instead, if you are significant consumer of Active Segments you will have to transition to Discover. If you can afford it, you’ll want to go to Discover 2.x (especially if you use Active Segments for reporting more than research), and if you can’t, make sure you do get the Discover 1.5 version.

The VS clients that are going to be hardest hit are the ones with many sites and a heavy investment in reporting infrastructure. If you are a single site and aren't heavily invested in ReportBuilder, the transition probably shouldn't be too hard.

What about Visual Site and Visual Workstation? Well, if you didn’t make this transition, you should probably be counting your blessings. You’ll probably be better off in the Discover 2.x world anyway. Apparently, Omniture says they will continue this code line – with Discover supporting software-as-service implementations and VS technology supporting the potentially growing market for in-house processing.

I have a hard time believing that. Why Omniture would want to keep two completely different – wildly different – implementations and GUI’s around just to bridge these two worlds is, to me, obscure. Perhaps Omniture does need a non software-as-service offering. But, if so, it ought to be in Discover clothes – not some completely different outfit.

I’m betting at Omniture they already know that, and the VS product-line will be allowed to die a quick death. Maybe they needed to commit to the two-product direction to keep the VS management team on board, but those sorts of decisions don’t usually survive the first few months. We’ve all noticed that merger plans don’t always go the way the management teams expect!

But in this particular case it’s hard to believe the implications weren’t clear to everyone. If we were in ancient Rome, the VS management team would be falling on their swords.

So what exactly does Omniture get from this acquisition? It sure isn’t technology. Essentially, they get a boat load of customers. They get relieved of the necessity of aggressive pricing in bringing those large accounts into the fold. They get the opportunity to upsell all those accounts on a range of new products.

All of that’s important. They also get some relief across the board from competitive pricing pressures. Probably best of all, they make it much harder for a new strong number two to emerge. The longer VS was around, the more chance a WebTrends or Unica had to become a strong #2 challenger. Now, Omniture enjoys a massively dominant (instead of just clear leadership) position in the enterprise space – and that will make it much tougher for any other challenger to emerge.

And, of course, they really didn’t pay all that much. All in all, it looks like another pretty darn good move by the Orem crew.

Omniture has long preached their vision of the analytics space as being essentially bifurcated. Google Analytics at the low-end and Omniture at the enterprise level. To achieve that goal, they’ve blended excellent product, great sales, a cohesive team and – perhaps most important – the relentless pursuit of a strategic vision.

You don’t always need “that vision thing.” Many a war has been fought (and even won) without a great general. But the striking disparity in fortune between Omniture and Visual Sciences in the last 18-24 months – during a time of unparalleled growth and prosperity in our industry – shows just how important it can be. Starting at almost the same point two years ago, one has become a giant. The other burns upon a funeral pyre.

Comments

[paul legutko]

Well, Gary, I think that the future really depends on what Omniture wants to adopt in the VS toolkit. I’ve already been asked by a few people what this move means in web analytics. The answer for us, I think, is: not much right now. Per Phil’s post, it’s going to take a while for Omniture to integrate VS clients into any new hybrid version of SiteCatalyst/HBX. And a hybrid version could be a great thing – imagine Omniture’s flexibility combined with HBX’s Report Builder – or a SiteCatalyst with 100 active segments instead of 5 – or a SiteCatalyst with tagless campaigns. Then again, they could do the opposite: a hybrid version without Report Builder, only relying on Discover 2.x for segmentation, with Omniture subrelations replacing Custom Metric Dimensions or Content Hierarchies. That would be tragic. But I don’t see a hybrid version coming out anytime before the end of 2008.

The other thing Omniture might do is keep HBX intact, treating it as a simplified version of SiteCatalyst and giving it a separate tab in the Omniture interface (like Discover, TouchClarity, and SearchCenter are currently). This would be a shame for current HBX implementations, but might be the path of least resistance for our friends in Orem. Or they may do nothing by way of integration, and current HBX clients will simply go on as usual and end up transitioning to Omniture as their HBX contracts expire.

[Rocco Albano]

I do not think Omniture will keep HBX alive as a standalone platform or even as a tab in the current Omniture portfolio of products beyond the middle of 2009. In terms of the migration path for HBX customers to SiteCatalyst, the strategy employed here by Omniture will be critical to ensuring ROI on the acquisition.

Feature wise I do think Omniture would be wise to integrate several key items within HBX including:

HBX's in-system charting and graphing layering capabilities

HBX's browser overlay tool which I have found to be slightly better then Click Map

HBX's Population Group feature, although I suspect you could do this with S.Props now in Omniture

I have also heard as mentioned in Paul's comment that HBX Report Builder is one of the best Excel analytics modules out there.

In regards to VS/Platform 5...well that is a whole different story. I think that with some Omniture enhancements to the Visual Site/ Visual Workstation interface the tool could be a replacement for Discover or at least a significant extension of Discover.

I think that Omniture should leverage the VS DPU, FSU, and Visual Load architecture as the foundation for scaling the Genesis network. To truly map upwards of a million impressions from a display campaign and actively correlate/segment that data to site click stream data will require serious processing power and the VS underlying architecture surely has it.