Should Analytics Be Bundled or Cobbled?

The question may be age-old, but it bears stating afresh: is it better to have your tools conform to the paradigm of compatibility/workability; or unique/targeted (assuming the latter is also “better at what it does” than the former)?

With the web analytics vendor space at once atomizing and consolidating, many marketers may be concerned that technology choices made now may play out poorly later on, leaving them either behind the curve on features or around the bend on complexity.

On the consolidation side we have older companies like Omniture, now part of Adobe; Coremetrics, now part of IBM (as well as the Unica offerings); Webtrends, continuing to expand its product line as a remaining independent; and free offerings like Google Analytics and Yahoo Web Analytics adding features and insight capabilities that many find attractive.

Yet there is also a move toward specialized, standalone products that do one thing (for the sake of argument, we will presume that they do that one thing well) or a handful of things; and which range on the scale of compatibility from near zero to near 100 percent. In social media and reach measurement especially, there seem to be new names entering the fray every day.

One Big Box – Less Assembly Required

I asked Rand Schulman, an industry founder and managing partner at Schulman Thorogood Group for his thoughts on this subject and he said, “With a packaged tool like Webtrends or the Adobe suite [“Omniture”] you can have access to measurement capabilities along a broad spectrum without having to worry about integration or the relationships between the way different tools collect data. This tends to make comprehensive tools the rule rather than the exception, even if certain standalone, targeted technology may offer some single advantage all on its own.”

Ease of integration, one-stop shopping, and some form of platform stability is clearly attractive enough to some marketers to overcome any concerns about “best-of-breed” or depth of feature in any particular area.

Smaller Boxes – Better Gifts?

Whatever dominance enjoyed by the thunderous vendors we all know, we know too there are smaller mammals roaming the countryside. A couple of companies I know of offer rather targeted web analytics. For instance, Adometry, based in Austin, TX, provides (among other things) “actionable optimization recommendations for a significant return on overall ad spend within the online advertising industry.” One of the most interesting ways the company does this is by delivering a deep dive on campaign attribution – partly by algorithms we have no space to detail, but also by verifying that all purchased impressions were actually served, “viewed by a human,” and, moreover, were part of the campaign’s target audience. In essence, it’s a sophisticated campaign attribution scorecard.

Another company, Vitrue (note the spelling on this one), delivers a form of management within the social media space by measuring engagement, loyalty, “fans” and “most engaged fans,” as well as other valuable tracking options. This data can be integrated with other analytics tools like Webtrends and Google Analytics. The suggestion is that Vitrue will deliver a more targeted and a deeper dive on its particular specialty – social media – than is otherwise available, and then you can integrate that with your comprehensive tool and have the best of both worlds.

Integration claims are often made, and they are well-intentioned. Without commenting on the integration characteristics of any tool in particular, it’s an open secret in the deployment world that “integration” can be one of those “famous last words” that one hears before chaos reigns. Often there is a fair amount of hand-stitching involved in combining disparate sources to get a full picture of overall traffic: call it “cobbling” if you will.

Also, we made an assumption at the top of the column – that the one-off solutions are “better than” the possible equivalent inside the major offering. The buyer is, as always, cautioned to verify the quality of the product before placing an order. There is no necessary factor that could cause a standalone to be better – the features simply have to be compared.

What’s Your Maturity Level?

Any recommendation about the modeling of your approach to “one vs. many” will depend on the maturity level of the existing analytics effort, along with the type of organization needing the deployment.

For immature environments of any kind, it seems the packaged tools offer more comfort. Best to stick with what you know will work (“has a track record”) rather than grasping at the next shiny thing.

For mature environments inside large organizations, a hybrid approach seems to make sense. A bedrock tool is needed, along with service-level agreements, great support, and enough vendor-reliability to get through the toughest senior meeting. But clearly many large organizations are finding ways to utilize the standalones, knowing what they can deliver and with enough confidence in their own mature analytics to know where and how to use these tools and their data to best effect.

For analytics power-users – those enjoying both mature environments and possessed of a nimble enough outlook to make quick changes (and with the technical chops to deploy) – it may make sense to go and find a targeted tool for each purpose. A reasonably well-funded, do-it-yourself organization can benefit in this way by avoiding higher license costs while seeking best-of-breed across the spectrum. It’s not for the faint-of-heart, but it can be enormously successful with the right team.

Companies may change and be bought and be absorbed; but often enough platforms remain. So even with the rapid changes taking place in the vendor space; choose for the platform, and choose what’s right for your organization in whatever stage of maturity it finds itself.