Webtrends Review from long experience
Overall Satisfaction with Webtrends Analytics
Our use case is as an interactive agency. Some of our clients come to us as Webtrends customers, and we provide implementation and/or support as part of our overall web site development and maintenance practice. The Webtrends consumers at our clients' organizations tend to be in the marketing area, as the tool is used for one or more of the following: 1) assess the performance/value of the web site itself, 2) assess the performance of the marketing that brings people to the web site, or 3) for diagnostics, i.e. what parts of the content, navigation, etc seem to be causing site visitor successes or problems. The audience for the first is execs, for the second is marketers, and for the third is site managers and strategists.
Pros
- More than some of its competitors, most Webtrends' configurations (reports, dimensions, filters, content groups, measures) can be done in the admin UI, rather than in the tagging and site code. The tag itself is smart - it can sense offsite links, clicks on pdf downloads, form button clicks, and so on, which eliminates a lot of extra coding or tag modification that has to be done with other products.
- There are so many levers and buttons in the configuration that nearly anything can be turned into a report, or a report dimension, filter, or measure.
- It allows re-analysis of past data as far back as 90 days. Usually, you do this if you have created new custom reports, content groups, change the filters, and so on.
- There is a software version, called On Premises. (The SaaS version is called On Demand.)
- It has real path analysis ... it does not daisy-chain individual steps as others do. The paths it displays are what happens in actual visits, up to 20 steps long. It has forward and backward paths (one visit can appear several times depending on how many times the node was hit), paths-from-entry (one visit, one path), content group paths as well as page paths. Its one lack (that I care about) is SiteCatalyst's Pathfinder report which allows you to identify wildcard pages in a 3-step hypothetical path.
- This isn't going to ring a bell for a lot of people, but it handles list variables much better than its competitors (basically, parameters that hold multiple values such as "choose as many as apply" kinds of variables.
- It handles the tabulations of parameters really well. It deals with three kinds of parameters: those in the pages' URL, those placed in the WT.meta's (I don't think any competitors use this approach and it is fantastic for easily keeping URLs clean for SEO purposes), and those collected automatically by the standard tag. When tabulating parameters, its competitors require more up-front work, lots more configuration time, or severe limits on the quantity.
- Having recently tried out Google Analytics' new Content Groups feature, I was reminded of how powerful it is in Webtrends. There's really no comparison. Furthermore, the content groups can be configured p in the UI as well as hard-coded into the page. Content Group paths can be up to 20 steps long, and are not daisy-chained.
Cons
- The big downside, the elephant in the room, is that it does not (as of right now) have on-demand segmenting, drilldowns, etc. You have to think of what you want in advance and create those reports then analyze some data. This is huge. You can, of course, re-analyze old data after creating new reports but you still have to wait. (This deficiency may become obsolete with the release of Webtrends Explore later this month (May 2014).)
- It has fewer mature integrations with other products and databases than competitors do, although I'm told it works with SharePoint better than anything else does.
- Its attribution modeling capability is behind Google Analytics'. In my humble opinion, this could be changed quickly if Webtrends would make some tweaks to its standard visitor history files (i.e. preserve the order in which past visits were sourced beyond the single most recent one, rather than storing all those past sources as a randomized list).
- It doesn't incorporate statistical tests, confidence intervals, or statistical associations. However, this same criticism can be applied to its competitors (other than A/B Testing products). It's a tabulation program, as they all are. In this respect, web analytics tools as a group are relatively primitive. Sorry to bring this up as a criticism of Webtrends but it's my pet peeve about the whole industry and I just have to say it. (p.s. take advantage of the heavy-duty Webtrends Scheduled Export functionality to get really granular data that you can feed to a stats program to get significances.)
- Although the documentation, help screens, phone support and the knowledge base have improved tremendously in recent years, there is still a pretty steep learning curve because it is different from the tools that entry-level users may have already been exposed to. This can be a shock and many users are alienated at first because they just don't get some of the fundamentals at first. I'd like to see much better help screens that are thoroughly interlinked with the KB and documentation. Having superb online support would make a world of difference with the adoption of this basically powerful tool.
- Any analytics product can affect ROI if used well. The impact depends on the analyst far more than the product. Having said that, I've seen companies go from data frustration to data addiction (and big ROI) once Webtrends is working properly for their situation.
Using Webtrends Analytics
Webtrends Analytics Implementation
- Implemented in-house
Comments
Please log in to join the conversation