The classic problem goes something like this:
1. Marketing: We need advanced customizations, deep integrations, increased scalability, better performance, and more control overall over Web analytics.In a nutshell, this is one of the reasons why hosted solutions exist (SaaS, ASP, on-demand, whatever). While it's hard to do Web analytics, it's even harder to do it internally using actual software that you run. Wouldn't we prefer it to go something like this:
2. IT: We don't have resources, time, or budget to help you right now. Fill out these forms and in the future maybe we can help.
1. Marketing: We need advanced customizations, deep integrations, increased scalability, better performance, and more control over Web analytics.
2. IT: Yes. Can do. Will do. What do you need and when do you need them by?
My belief is that to "do Web analytics" the right way, you need an allocation of IT resources to support your implementation and extend it to fulfill strategy and improve business performance. After all, I firmly believe Web analytics is for optimizing business performance, guiding strategy, and supporting tactical decisions. And to do all that, you need resources when you need them. The larger your site or portfolio of sites, the more resources you need.
It's all pretty logical. Getting back to IT, if you're using a hosted solution, you need fewer IT resources. The vendor takes care of a lot of IT stuff. If you are running your analytics in-house, you need a team of IT resources because you will be doing it all yourself.
I would prefer those technical resources report into Web analytics, but I'm not sure if the general business world (as in non-Internet companies) sees the ROI of Web analytics clearly enough to immediately delegate a full-time "mini IT" team to support analytics at phase zero (i.e. when you first get hired and plan the rollout). And that's why you need to be very wary of what vendors tell you about IT requirements and Web analytics.
If management expects that you just need to tag the pages and you, the analyst, can do that yourself, your company will be in for surprise. It's never that simple. Smaller companies with one or a few sites that use the same technology may be able to pull off the solo cowboy analyst including tags and doing all the tech work. But larger companies that have many sites and many different technologies serving those sites are a much different animal.
My advice is that you can't be fooled by vendor messaging that claims "you don't need IT." That's just not true. Marketers can't do Web analytics alone. You will need IT to help you extend your Web analytics solution. And as I've already stated, the level at which you need IT will vary on how you "do" Web analytics. It differs greatly if you are running an in-house proprietary solution, an internal vendor solution, or a hosted solution.
If you are doing Web analytics using a proprietary solution, you created internally, you may probably then already understand what I mean when I say "Web analytics needs IT." Chances are you are using an OLAP-based solution that has huge BI infrastructure behind it and the cubes contain latent information. Your data model may be limited compared to the major vendors. Your tool may be overly complex, hard for business users to use, and limited in terms of features, or it may be the coolest thing since sliced bread, and the people who created it may know more than the vendors. Still, unless resources are adequately delegated to support analytics and extending the implementation, your tool users and report consumers will make thousands of requests to IT, and they will go unfulfilled leading to user frustration.
If you are running an in-house software solution, such as that provided by Unica, WebTrends or Omniture, you will rely on IT for all sorts of things, like hardware and software maintenance, database administration, network support, and will need to leverage help desk and ticketing systems. In addition, Web analytics projects become part of the IT project planning cycle with budget requests and consideration.
If a business wants to succeed with Web analytics, it must determine how to effectively resource the implementation and ongoing extension of an analytics platform. Here are some tips for ensuring you get the resources you need:
Factor Web analytics resource needs into the capital budgeting and yearly planning process. Business stakeholders must identify the IT resources they need in advance, and then align the IT team according to business goals. Resources must be allocated according to financial guidelines that maintain corporate profitability.
Document your Web analytics projects and business requirements and share the documentation with IT. Whether your Web analytics projects are related to implementation, campaign optimization, data description, or integration, you need to share that information with IT so they can determine how to support analytics.
Identify and document why you need IT resources. In other words, identify and document what IT will be doing for Web analytics and how their work is necessary for improving corporate performance. On the business side, explain that you won't be able to fulfill X business goal without IT resources.
Leverage a project manager. Project managers are critical and important to cross-functional team success. They focus work, align people, determine tasks, monitor completion, and allow a multifaceted team of business marketers and IT to do what they do without worrying about managing the project.
Share your analytics success with IT and let stakeholders know how IT has helped you. Often corporations forget that these very talented IT folks are working really hard behind the scenes, often without getting much (or any) credit for the complex work they do. When you have an analytics success, share it with the folks that helped you tag the pages or configure your servers. When people are singing your praises in the cafeteria because they now have the data they need to do their jobs and/or you've improved their business performance, let them know IT backed you up and helped you deliver. There's enough glory to go around.
If you do what I'm saying in this column, the problem of marketing versus IT will minimize. IT will be able to keep up with all of your constantly evolving business requirements and the dynamic, high-maintenance nature of your Web analytics program. And your marketing department, business stakeholders, and executive team will be very happy with the results.