Subscribe to DSC Newsletter

It looks like Web Analytics is hot and there is a lot of jobs in this field now. But currently the analytical techniques employed by Web Ananlytics are really simple: the most sophisticated one is fractional factorial design; most analysis is done by summarization and segmentation. The entry barrier for WA therefore is kind of low. Everyone with average analytical ability can get in. Data Mining/Predictive Modeling definitely is a more hard-core analytical field and suit better for people with strong analytical aptitude. But is there any real demand in job market for this skill?

Views: 1438

Reply to This

Replies to This Discussion

There two types of web analytics:

  • reporting: simple in terms of statistical technology (but sometimes not in terms of architecture)
  • mining insights: typically more sophisticated, requires text mining, NLP and all the more advanced data science techniques

The fractional factorial is a higher end breed, more of a GLM traditionally , but there are choice modeling approaches as well ( for these designs.

So, the research design does no always categorically dictate the mathematical technique, it does give it a nudge in a certain direction though. If you can apply high end methods to age old research designs, then you can raise the bar on the way way way low barrier for web analytic human resources (which is usually just reporting anyway). Text mining is cool, but there are automated packages for around 3k that are easy to use, but inflexible, thus not really harnessing the potential, so we are back to the low barrier of entry because vendors decide what is best for their clients and not the data scientist.

Yes - there is a great deal of demand for data mining/predictive modeling people. Not only that, but there's competition between employers for the best talent & salaries are much better for heavy-duty quant people than they are for entry level analysts. 

While you're correct about the entry level web analytics professional - most need only to generate reports out of Google Analytics or Bing, there is still substantial demand for folks to use machine learning techniques to program recommendations engines, to construct PPC bid optimization models, and to generally mine web data for patterns.  If you're managing thousands of search terms and trying to optimize your bidding strategies across multiple web properties, then a great analyst would be your best friend.


On Data Science Central

© 2020   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service