After several years of vendor consolidation the
web analytics market has a new enterprise-level vendor. Following a summer of
innovation and product announcements, Google finally launched Google Analytics Premium in September.  

Premium aims to address the needs of large corporations
that cannot rely on the free standard version Google Analytics (GA, with all
inclusive package, a powerful analytics tool, migration and implementation
consultancy, training, account management and 24/7 support.

However, switching web
analytics tools is rarely an easy decision or a straight forward process.

If you are thinking of
upgrading from GA or a switching from another analytics vendor, here are the pros and cons…

What is GA Premium?

GA Premium includes
some key corporate governance, financial and analytical benefits.

Contracts and ownership of data

First, you
get a signed contract rather than the generic Google Terms of Service. Google is assuming the legal accountability
corporations expect from their vendors. The contract includes strong service
level agreements regarding data processing, interface availability and uptime
levels. It is a guarantee that companies will have access to their data
whenever they need it.

Google is
also providing dedicated account management and consultancy services. It wants
to ensure Premium is implemented to a high standard and used by clients to
avoid underutilisation and disillusionment.

Google has
also defused a key corporate objection by assuring clients that ‘you own your
customer data and control how that data is shared’. Legal and risk departments
can now be appeased.

Costs

Costs
are fixed at $150,000 per annum, irrespective of traffic volumes. It is a very
attractive proposition especially if your websites generate over a billion page
impressions per annum. You no longer need to estimate your annual traffic volumes
upfront or reserve any budget for server call overconsumption.

$150,000
will be prohibitive for many companies
. However, it does create some
exclusivity and an implicit assurance that the quality of service would not be
diluted on the altar of additional sales.

The price
point isn’t arbitrary. I believe Google has priced Premium to cover the projected
human resource cost required to support Premium clients. Having dominated the
analytics market for small to medium (with some large) companies it is now
going for the really big fish.

I expect Google to modify the price (my guess is
down) once it has had time to assess its success attracting new Premium clients
and evaluate the costs associated with running the service.

Finally, analysts
will have 50 custom variables available to them (up from five); greater
processing power; more data export flexibility and advanced attribution
modelling tools. The list of additional features is only bound to grow in
coming months.

For a full
features list have a look at the GA Premium fact sheet.

Who should consider switching or upgrading
to GA Premium?

Companies
that already use the free standard version will obviously find it relatively
painless to upgrade. You would not need to make any modifications to your implementation.

Custom variables

That being
said, I strongly recommend using the extra 45 custom variables as much as
possible. Custom variables allow you to capture data dimensions unique to your
websites, industry and business requirements.

You would be able to further segment
your customer base by data captured on the site (e.g. type, preferences, age,
gender etc.); classify pages by type/function; and cluster user groups by
interaction with the website (e.g. Signed
up for newsletter, reviewed a product, used an online tool such as a loan/mortgage/saving
calculator etc).

Premium has
plugged a major shortcoming of the standard edition with the additional custom
variables. It gives a web analyst the freedom to explore customer behaviour in new
ways and discover the necessary insight for optimising both the customer
journey and your ROI from the online channel.  

Investment in staff and resources

This brings
us to the next point, having a dedicated web analyst. Whilst you do not need
to be a web analyst to use Premium, you would be wise to spend $60,000 to
$90,000 on a dedicated and experienced analyst
to extract the added value Premium
offers.

You should
also consider devoting some technical development resource to support your
analyst. Web analytics implementations, like websites, are live organisms.
Modifications and enhancements are essential. And with GA’s aggressive
development roadmap it is almost inevitable that new features would require
more implementation. 

In most cases this development
resource does not need to be dedicated to web analytics and can also be used
for other projects.

Unique features

Premium includes several features currently unavailable, not
easily configurable or cost extra (tick as appropriate) with other vendors.  

The most significant is the advanced attribution
modelling tool. The tool not only allows marketers to assign conversion values
to all channels in the conversion path but also custom-weigh those values. 

Models include First Click, Last Click,
Linear (aka equal weight), time decay (where more recent channels in the path
get higher value) or U-Shape (where first and last click have higher values).
So if you engage in multiple marketing activities and struggle to determine the
value of those activities, Premium could be your solution.

Premium will also include greater integration with other
Google products such as DoubleClick (expected to include auto-tagging for
display ads similar to the AdWords integration).

These integrations, which will
always be unique to Google, save significant time and cost by eliminating the
need to set up campaign tracking and reduce the chance of tracking errors
caused by human intervention.

Another feature that is currently available in the free GA
version but is unique to Google (as far as I’m aware of) is Analytics
Intelligence. Intelligence is a simple statistical analysis algorithm that
automatically identifies significant data pattern change and alerts you to that
change.

It is one of my favourite analytics features. I would like to see this
feature enhanced as I believe it is the future of web analytics – where machine
provides human with threads leading to the needles in the haystack.

Why shouldn’t you switch or upgrade?

Whilst
removing some key barriers to corporate adoption of GA Premium, there are still
some strong arguments against a switch to Premium.

No customer-level data

GA is not
really designed to capture and process customer-level data. Yes, custom variables can be configured to visitor-level but overall the tool
is not very accommodating for such analysis.

That is not a tool drawback per se
but a conscious decision by Google. Therefore, it is unlikely to change in the
foreseeable future.

If you are
looking to engage in customer-level analytics for customer targeting and
personalisation or if you wish to tie web analytics with CRM and other offline data
then GA is probably not ideal.

Google does
provide a little more leeway under the new Terms of Service for GA Premium with
respect to the collection of Personally Identifiable Information (PII) – it is okay
to do so as long as Google cannot identify the data as PII).

However, it is
unlikely to allow the required level of data granularity for full customer
analytics.

It is the
biggest drawback of GA
. I strongly believe that, as it grows in significance, web
analytics will merge into traditional business analytics. This is a vision
currently pursued by companies like IBM which purchased Coremetrics and Unica
in 2010.

GA Premium needs resources

As I
mentioned above, companies without dedicated analytical and technical
development resources would struggle to take full advantage of GA Premium. But that
is likely to apply for all enterprise-level web analytics tools. No dedicated
resource then stick to the free standard version.

The pain of switching

On the other
scale, If you have already integrated your web analytics tool with other systems
such as multivariate testing, content management, personalisation platforms, marketing
or CRM databases etc. you might find a switch to GA Premium (or any other tool
for that matter) not very straight forward, protracted and even painful.

Will GA Premium stifle development on the
free GA version?

I actually expect
the free version to benefit from Premium features gradually trickling down into
the standard version. It is also important to note that no existing features were
removed from GA and made exclusive to Premium (Google should slap a ‘no
features were harmed making GA Premium’ tagline on the standard version).

Final thoughts

There are no
hard and fast rules for selecting a web analytics tool. It is a matter of
corporate appetite for change, the availability of qualified resource to drive a
migration process, the skill to execute the migration process successfully and
the ability to utilise the tool for analysis and optimisation.

If you are already
using GA and finding yourself constraints then Premium should be your next
move. However, if you are using another corporate-level tool then you should
start with a very thorough gap analysis that would include:

  • How current requirements are met by incumbent
    vendor vs. Premium.
  • How foreseeable future requirements are met by
    incumbent vs. Premium.
  • Annual running cost differences and cost of
    migration.
  • Which internal resources would participate in
    the migration process and their availability.
  • Will a new tool such as GA Premium actually
    deliver the added value to justify a migration project.

The last question is probably the most
important one to explore. Often companies lack the understanding and skill to
drive real value out of their web analytics tool. They get frustrated and
assume the problem is with the tool rather than an internal issue.

They seek to
resolve the issue by purchasing a new tool only to be disappointed again when
the same internal deficiencies emerge.