Friday, November 26, 2010

From Web Analytics & SEM To Business Intelligence

From Web Analytics & SEM To Business Intelligence

The roles of people involved in both web analytics and PPC management is changing. Both are becoming more important and mainstream within many organizations. And the skills needed to master them are becoming more complex. Here’s a look at the how things have changed, and what to expect going forward.

The Evolution Of Web Analytics

Web analytics was originally an IT function, a tool that could be used to measure website activity and report basic traffic stats to the organization. Only later, the web became a very important marketing channel, and responsibility for website management and analysis in many organizations moved to the marketing department. Google facilitated this trend by branding web analytics as a marketing tool, providing an easy-to-use interface that could be installed once by IT and providing all the info necessary for analysis by non-technical people in the organization.
Now we might be seeing the web analytics industry shift once again. We have seen many acquisitions in the last year and some of them hint at where we are heading. First, we saw Omniture’s acquisition by Adobe, which was a bit controversial in the eyes of many industry experts such as Yahoo’s analytics guru Dennis Mortensen.
However, IBM’s purchase of Coremetrics and Unica was more intuitively obvious. Unica provides a robust web analytics tool that can populate a database and be queried using a browser-based interface, which led many people to ask: will web analytics become just another team inside business intelligence departments? That is a plausible option, as the web is just another source of data.
These changes are reshaping the way web analysts are seen in organizations too, and also to whom they report. Google’s analytics guru Avinash Kaushik discusses this in depth in his postWho owns web analytics? and proposes a framework that helps organizations to decide where web analytics should sit based on their maturity level. Some companies should have it under sales, others under marketing and yet others under the CMO himself (see ex-Googler Brian Clifton’s view on web analytics and marketing). And, if else fails, why not having it under the finance department as proposed by Jim Novo? As he says: “At least mission and thought process are aligned, and some degree of influence / ability to act is present.”
So, what should we expect in 2011? Will the market change its direction? See a hint on what is waiting for us next year in Brian Clifton’s web analytics Predictions.

Search Marketing, Meet Business Intelligence

In the same spirit, PPC management has also evolved as SEM managers and optimizers are increasingly drowning in data. Originally, PPC management was about building nice creatives and managing a campaign’s ability to bring people to the website without spending too much of the total marketing budget. However, as paid search becomes a significant part of marketing budgets, advertisers can’t rely solely on online conversion data, and they need to optimize SEM performance according to actual sales or client lifetime value (CLTV) metrics, that typically appear in a separate, isolated database—within a CRM system like, for example. With customer sales data sitting in a different database than AdWords/Facebook/Yahoo analytics, you may find yourself trying to integrate data within a big IT hassle. Following the data retrieval and integration at the granular (e.g., keyword/ad) level, you will most likely find yourself in an “Excel hell.”
Advanced advertisers that have started to tap into business intelligence tools will agree that these are superior tools in comparison to spreadsheets or various in-house mechanisms that require various SQL queries or communicating with the “IT guy.” The results of using a good business intelligence tool to analyze pay-per-click advertising data are mainly increased campaign ROI (based on optimization of client values) and increased work efficiency (less time wasted on integrations).
Besides in-house tools that can be tailor-made to organizational needs but might be costly and buggy, SaaS solutions that attempt to solve the SEM spreadsheet nightmare are emergingt. One prominent provider, Edge.BI, is a start-up company solving that “Excel hell” problem for SEM’s. The company has tailored-made its SaaS for in-depth PPC analysis and decision making, offering professional search marketers to enter the next stage in the SEM optimization evolution. The company’s SaaS is in Beta and offering trials to selected advertisers: I was quite impressed with the Edge.BI functionality and ease of use.

The New Skill Set Needed By SEM Managers

To illustrate the skill set required from SEM analysts nowadays, here is a random job description I picked up from
Job Title: Search Engine Marketing Analyst
“Numbers are your life! If this statement describes how you see the world, we want to speak to you ASAP… success is built on our ability to access and analyze results and trends of the hard data.”
  • 1-2 Years in Search Marketing and Web analytics on the Client or Agency side.
  • BS in Statistics, Math, Marketing or Economics.
  • A solid foundation in statistical analysis and marketing research.
  • Demonstrated ability to creatively develop analytic solutions
Within the above SEM job description, note the occurrences of the words of “numbers,” “analyze results,” “hard data,” “statistics,” “analysis,” “analytical” and so forth. It seems that the SEM industry will soon be filled up with data crunching, Business Intelligence geeks that will be quite savvy in Fishing From a Barrel…! And those SEM optimizers that are not analytical? Are not using the right software efficiently? Well, they’re going to have to reinvent themselves or be left in the “Excel hell” and out of the job.
Unrelated to where we sit and who is our boss, one thing is certain: web analysts and SEM managers increasingly need the business intelligence mindset. As we evolve into website personalization, multichannel campaign management, database integrations, data mining and predictive analytics, our minds need to switch to a highly analytical mode.
In summary, data crunching is increasingly important. The web is a wonderful laboratory and it enables us to make data-driven decisions with a very high level of accuracy. We can test hypotheses and prove them right or wrong by the numbers. So, if you suffer from data-phobia, you can start your journey by either reading The Cartoon Guide to Statistics or Innumeracy; both very entertaining and enlightening.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.