|
|
Yesterday’s posts by Justin Cutroni and Jason Van Orden dealt with Analytics as a Business Process and the Measurement of Podcasting metrics. Today we have Marshall Sponder and Gary Angel.
Marshall Sponder is the author of the Blog: WebMetricsGuru.com which is part of the “Know More Media” Team.
Marshall’s Thoughts…
I have different types of issues depending on the complexity of my clients’ environment.
For a few of my SEO/SEM clients it’s getting them on board for analytics and tracking properly. One client of mine is publishing marketing material pretty soon and I want to have tracking code placed on the published works so we can track response and cross channel conversions.
Taking it to the next level, another client is a large News Video site wanting to increase their SEO presence but their template design and redirects interfere with getting their online videos indexed properly and before the right audience. I instituted changes that are helping their videos be indexed – but we need to know more – which videos are being clicked on and how often they’re being downloaded. Much of their pages is Flash based – and difficult to measure with Web Analytics – and getting the right tracking solution in place is a struggle and has been put off till next year.
And at IBM, I currently work with the US Homepage promotions to measure traffic, promotions, conversions, engagement and branding effectiveness. On a Enterprise level – template standardization and tagging are vital – and I’m often talking to stakeholders on getting their pages tagged for Coremetrics or IBM Surfaid (we’re in the process of transition from Surfaid to Coremetrics) – again – it’s a struggle.
At the Enterprise level – tagging standards are really, really important. I have found events that are only partly tracked – mostly it’s due to analytics limitations, not tagging URLs properly or not setting the analytics to track the right variables or all of them.
Finally, on a level beyond small and large companies – there is the effectiveness of a site / page and the promotions sending traffic to it. I’ve often been frustrated at not being kept in the loop about promotions running that drive traffic to a site – the site/page appears effective – but the targeted traffic is really being driven and you need to know that in order to determine how effective your site/page is.
Gary Angel is the president and chief technology officer of SEMPhonic
Gary’s Thoughts…
To be honest, when Manoj asked me to contribute a few thoughts on common pain points that our clients share and how they might be tackled, I figured it would a slam dunk. But the more I thought about it, the more I felt like doctor diagnosing Evel Knievel – it would be easier to list what doesn’t hurt than what does!
But as painful as implementation, rollout and reporting tend to be, I think the biggest pain point arrives when an organization has finished an implementation, is satisfied that what they are measuring is something like reality and all the hullabaloo around training and reporting has settled down. And then comes the moment when the implementation team has to sit back and say to themselves – what now?
That moment transforms what should be a satisfying sense of accomplishment into a scary look into the abyss. We find that for a lot of clients, getting started on real analysis is by far the hardest step – so here’s how we tend to guide clients about where and how to dive into serious analytics.
Start with where the money is on your site – for most people that means analyzing the sales process pages and the conversion process. A formal analysis here will almost always yield some dividend. And never, ever finish an analysis that doesn’t have actual suggested changes for testing. If your analysis doesn’t have data-driven changes for testing, then you didn’t do it right.
For many sites, an analysis of Internal Search as a router/convincer is a good next step. Search is a vital site component and will nearly always provide many optimization opportunities. Follow that up with an analysis of routing and navigation. Then look at cross-sell opportunities. While you’re doing all that, setup some longitudinal studies of 1st Visitors, 1st Purchasers and regular customers – after you’ve cycled through five or six months of other studies you’ll have enough data to tackle these.
By the time you’re done with all this, measurement will probably seem like a normal part of your business – and all those tests will be there for the measuring. That’s the way things should be – and you won’t have to worry so much about that scary abyss!
