3. Know your organization
Are you looking for reporting or analysis?
• Decentralized decision making?
• Cover your back culture?
• Risk adverse?
• Distribution of knowledge?
4. Factoring in price
Free tools aren’t free!
• Implementation can be a huge cost
• Cost for custom tags and reports?
• What about added cost for dashboards etc. (example: SQL,
Tableau)
• Incremental time and cost now added to making changes on
the site (tag management)
5. Implementation
How strong is your IT?
• We’ll cover implementation in more detail later.
• There are consultant firms that pretty much just do implementation
• Get it wrong and all of your metrics can be off!
• Support for tagging, custom tags?
Note: If you have any interest in learning or know JavaScript and get certified in one of the top
analytics tools you’ll have multiple job offers the next day. There is a big shortage in the market
place right now for people with these skills, Razorfish has really struggled to fill these roles.
6. Clickstream vs. Web Analytics 2.0
How sophisticated are you really looking to be?
• A/B and multivariate testing?
• Heat mapping?
• Advanced segmentation?
• Most of the big tools do all of the above, but some better than
others.
7. Reporting
How easy is it to pull out the data from your tool? Most of the time
your data will finally land in Excel or PowerPoint.
• Easy to use API feeds?
• Data exports?
• Automation and hooks directly through Excel?
• Most of the big tools do all of the above, but some better than others.
8. Razorfish POV
Web Analytics Tools Comparison
Platform Pros Cons
• External data integration • Complex interface
• Excel client • Learning curve
• Retroactive meta-data integration through SAINT • Complex Tags
• Strong segmentation
• Cross-visit attribution modeling through Discover
• Best in class alerts, dashboards, and target features
Adobe • Focus on ad hoc reporting
SiteCatalyst v15 • Flexible pathing analysis
and Discover 3 • Custom reports (variables/events)
• Multiple levels of persistence
• Conversion deconstruction
• Genesis integrations with email service providers, Salesforce,
and others
• $$
• Highly intuitive, refined and easy to navigate interface • Report building backend similar to V8.5
• Storyboard mode • Reports not very flexible
• Well-organized profiles through Spaces • Difficult to perform ad hoc reporting
• Meta-data integration – Translation tables • REST API inferior to Excel plugin
Webtrends • REST API access through Excel • External data integrations mostly custom
Analytics 10 and • Good export features • Segments tool inferior to Adobe Discover
• Custom reports (variables/events) • Complex Tags
Segments
• Multiple levels of persistence
• Unlimited variables
• $$ Pricing (based on
30MM PVs)
$ < $25K
$$ $25 - $50K
$$$ $50 - $100K
$$$$ >$100K
9. Razorfish POV
Web Analytics Tools Comparison
Platform Pros Cons
• Comparative benchmarks and intelligence • Comparative benchmarks limited to content sites
• Tabbed reporting (report toggling) • Limited segmentation capabilities
• Social media monitoring • Clogged interface
• Visitor profiling – LIVE • Sampled data
• Powerful metric attribution features • Limited MS Office integration
IBM CoreMetrics • Most complex tagging required
• Unclear data integration capabilities
• Difficult implementation (12 tag types)
• $$$
• Easy to use interface – fast learning curve • No raw data feeds available
• Rich dashboards and visualization • Very limited external data integration
• Strong segmentation • No forecasting
• Click-based attribution modeling • Segmentation is limited to the visit
• Real-time reporting • No metadata upload capability
Google Analytics • Custom reports (variables/events) • Very limited user provisioning
Premium • Multiple levels of persistence • New tagging required
• Social assist feature • Limited unsampled data (no dashboards/scheduled
• Simplest tagging reports/interface)
• Intelligence / automatic alerts • $$$$
Pricing (based on
30MM PVs)
$ < $25K
$$ $25 - $50K
$$$ $50 - $100K
$$$$ >$100K
11. Visitor Acquisition
The big 3 (Omniture, Webtrends, and GA) break out visits
by:
• Direct
• Search
• Referring
• Other
12. Visitor Acquisition (cont.)
Some things you should keep in mind with Direct traffic:
• Direct traffic is not free!
• If you don’t set up your campaigns right it will look like direct
traffic
Webtrends example:
ww.mywebsite.com?WT_id=MyCampaignName
13. Watch out!
Your Web Analytics tool ad traffic number will never match
your ad serving (Atlas & DoubleClick) counts!
• Web analytics tools won’t match either, they all have their own
business rules.
• Tags are in different locations on the page.
• 3rd party servers have issues to.
14. Simple segmentation
You can learn a lot by segmenting on just the acquisition
traffic source:
• Bounce rates by source
• Conversions by source
• What are the top key words?
• Top converting referring sites
• Top performing media?
15. Watch out for averages
Averages like, average time on site, or average pages per visit can sometimes be
a bit deceiving. It’s definitely worth taking the time to look at the distribution.
17. A page load FYI
Be carful when using page loads as an engagement metric.
Sometimes one page acts like multiple pages…
A lot of new sites now use tabs
and accordions to navigate and
display content, if the url
doesn’t change it’s not a new
page.
Tab example:
http://office.microsoft.com/en-us/make-it-great/
Accordion example:
http://www.microsoft.com/office/cxm/en-us/small-
business-premium/index.html
18. Time to convert
Another watch out! Most of the web analytics tools tend to
focus most of their reports on a single visit.
Number of purchasers
28% of total
10 min.
43% of total
45% of total 68% of total
62% of total
30 min.
1 hr. 6 hrs.
2 hrs.
Time between first Visit & purchase
(Min.)
19. Tracking multiple visits
Same as time to convert, here is another example…
I recently needed a new Found the cartridge I Epson’s cart experience
printer cartridge, so I needed, and then started sucks so I ended up buy the
Googled Epson price shopping. new cartridge on Amazon
…but had Epson’s site been better I would have converted on my second visit.
20. Another watch out
In the last example I used Google to find the Epson site, but on my second
visit when I went to purchase the cartridge I just typed in the url… so am I in
the search segment or the direct traffic segment? which channel would have
gotten credit for the conversion had I actually purchased the cartridge on the
Epson site?
21. What are converters doing?
FPP Upgrade Conversion Rate
By Page Load (FYQ2 2011)
22. Click density reports
It’s a real pain to put these reports together but site managers love
them:
24. Internal Search Insights
• Can shed some light on what content is important but
hard to find or missing.
• How successful are those internal searches?
25. Focus on the critical few
Its all about a handful of KPIs, don’t get overwhelmed by tracking and
interpreting every possible metric.
• Example – share of search, what can you do to improve that metric? We’ll
talk more about this one next week.
26. Macro and Micro Conversions
Another great method taken from the book on how to decide what the
critical few should be
Traffic to
Website
Conversion
Traffic to
Website
Support Research Careers $
Micro
Micro Conversion
Conversio Macro Conversion
n
27. Steps to purchase
Another great approach is optimize the funnel, i.e. the different steps
a customer typically takes before making a final
purchase/conversion.
28. One more approach
Another great approach is focus on the customer’s lifecycle.
Customers will have different needs and will use your site differently
based on where they are in the customer lifecycle. This approach is
great for service type products and sites.
Decentralized decision making? (T-Mobile very much this way, all needed to see numbers before making a call)Cover your back culture? (Microsoft is very much this way, don’t want to make a bad call and need someone to blame)Risk adverse? (Microsoft is this way, why I have lots of work… agency makes the risky recommendations so the client has someone to blame) Distribution of knowledge? (T-Mobile was like this, knowledge was power people kept info for themselves, Analyst level had very little view into what changed)