Gilt Groupe is a private
e-commerce shopping site based in New York City. The store focuses on flash
sales to create the fastest most exciting shopping experience with unparalleled
customer service from the moment they enter the site to the delivery of the box
at the doorstep. Gilt provides insider access to today’s top designer brands
and labels, up to 60% off retail price to members only (Gilt Groupe, 2014). All
sales are first come, first serve so members have to be sure to log on early
for the 36-48 hour sales.
Because the company is
innately digital it is vital for Gilt to have integrated web analytics in
place. This allows the company to focus on taking care of their customers and
growing the company and site.
Gilt Groupe implemented
Google Analytics in 2011, but has since upgraded to Google Analytics Premium.
The company’s main incentives to upgrade were to have access to unsampled data
and capture more detailed information at the user level. Having this
information would help Gilt make every customer interaction count as much as
possible and allow critical business decisions to be made based on
statistically sound and stable data.
The first task Gilt took on
with Google Analytics Premium was to ensure the information they were gaining
was from their full audience, not just a sample of the data. Unsampled reports
from Google Analytics Premium has allowed Gilt to obtain a clear view of
results from both tests and actual campaigns. The unsampled data results have
given Gilt the opportunity to remove uncertainty and act on test and campaign
results with confidence (Google Analytics, 2012). This directly reflects the
goal to provide impeccable customer service because Gilt can see exactly how
their customers are reacting without having to hope or assume a sample of the
data is accurate.
In order to examine a wider
variety of key metrics and gain a more holistic view of customers, Gilt
implemented custom variables. Gilt is currently using more than 20 custom
variables, to enable more opportunities for comparison and analysis, as well as
A/B testing (Google Analytics, 2012). With the newest release of Universal
Analytics by Google, custom variables are now referred to as custom dimensions,
but function the same way as custom variables to help answer new questions
about how users are interacting with content and devices (Google, 2014).
The custom variables Gilt
has been tracking include user IDs, partitions for testing, time stamps, page
types, demographics, testing variants, hit times, and more. Gilt has been able
to take this data to construct a clickstream and to reconstruct visitor pathing
across their different domains, analyze both onsite and external sources of
traffic to sales, do site personalization, and view test results in Google
Analytics Premium (Google Analytics, 2012). This allows Gilt to track different
types of visitors and create content based directly on a specific customer.
Gilt has seen an increase in sales and higher site engagement as a result of
these tests.
Gilt is also using the
clickstream data to create decision models to predict buying behavior of their
customers. The detailed data collection about each session related to timing,
pricing, sale position, and others allows Gilt to feed theses variables into a
model and predict the possibility of a purchase (Google Analytics, 2012). This
creates more time through efficiency and more time for analyzing the site,
customer behavior, and making decisions rather than strictly spending time
collecting data.
Being an ecommerce site,
Gilt really focuses on sales funnels to ensure success. To fill their sales
funnel from top to bottom, Gilt has implemented attribution modeling. Google
Analytics Premium’s attribution modeling tool allows users to quickly view and
compare the results of different attribution models, as well as understand how
users pass between different marketing touch points (Google Analytics, 2012). Ana Kravitz, Gilt’s web analytics senior
manger said, “GA really shines here.” And she further explained the feature was
easier to use than competing tools available (Google Analytics, 2012).
Using attribution modeling
has helped Gilt learn the last click model brings in a high amount of affiliate
revenue. But, using the first-click model, affiliate revenue is much lower. In
the aspect of business and marketing decision-making, Gilt can see the danger
of basing marketing efforts solely on last-click results. The tool has showed
reducing other marketing expenditures to focus on affiliate marketing would
reduce the ability to acquire new users and Gilt would slowly lose customers
through attrition. But, marketing programs such as referral generate new
customers and rarely get credit for last-click purchases (Google Analytics,
2012). Currently, Gilt’s site and marketing campaigns are built heavily on
recommendations and trying to gain new customers. It is beneficial for Gilt to
see in the funnels where visitors are coming from for this reason, whether it is
email campaigns, link networks, social media pages, or others.
To help sell products more
efficiently through the flash sales, Steve Jacobs has implemented software that
could meet the nature of the company’s constantly changing inventory. The
software created helps Gilt adjust inventory by analyzing customer clicks on
links to brands, colors, sizes, styles, and other categories users have
purchased and matches that information to the characteristics of the day’s
merchandise and predicts which products consumers might be interested in. Then,
members get customized alerts about product offers based on the predictive
software-driven recommendations (Boulton, 2012). Customer data is collected
from the time the customer interacts with the web site until they make a
purchase. This software really helps Gilt with personalization and helping
people find and discover what they are interested in.
In theory, Gilt could use
paid online media to increase their sales and members, and ultimately enhance
their analytics. This would allow the company to see if they really need to
spend the money on paid media, or if their referral programs and social
interaction are enough. The analytics portion of the paid search media could
show keywords and categories that Gilt is missing out in. Or, specific brands
that could show for a successful flash sale. But, overall Gilt has a strong
analytics program and strong implementation plan company wide.
References:
Boulton, Clint. (2012). Gilt
groupe’s flash sales directed by analytic software. Retrieved from http://blogs.wsj.com/cio/2012/06/18/gilt-groupe%E2%80%99s-flash-sales-directed-by-analytic-software/
Gilt Groupe. (2014). About
gilt. Retrieved from http://www.gilt.com/company/main
Google. (2014). Universal
analytics usage guidelines. Retrieved from https://support.google.com/analytics/answer/2795983?hl=en
Google Analytics. (2012).
Gilt groupe embraces the advanced functions of google analytics premium.
Retrieved from http://static.googleusercontent.com/media/www.google.com/en/us/analytics/customers/pdfs/gilt.pdf
No comments:
Post a Comment