The fastest growing bed company in America

Goals

This “fastest growing bed company in America” is a Fortune 1000, NYSE-traded multichannel retailer who has a large wholesale footprint and a growing direct to consumer channel (website and call center).

Delve was initially brought in to help with Website Analytics and Testing/Optimization. As the direct business grew, we were tasked with helping start a Marketing Business Intelligence capability. Our mandate was to turn data from variety of sources into insight to explain ‘what happened’, ‘why did it happen’, and ‘how do we learn from the insight going forward’.

Approach

We approached the project in several concurrent phases:

1. Establishing a Marketing Data Mart
We started by identifying all data sources and establishing API, FTP or e-mail “feeds” from all sources. Sources included:
- Website Analytics (Google Analytics)
- In-house Call Center DNIS-level data (from a 3rd party vendor)
- Outsourced Call Center DNIS/campaign data (from a 3rd party vendor)
- Financing data (from an outside financing bank)
- Merchandising and Transaction detail (ShopVisible, the e-commerce provider)
- National Television (from a media agency)
- Local Television (from a media agency)
- Paid Search (from an interactive agency)
- and other sources

2. Data Audit
While collecting the data in a Marketing data mart, we were auditing the data as it came in to establish “100% trust” in all numbers.

The client released a new version of the website just before the assignment kicked off; the new e-commerce website included many AJAX-enabled functionalities. One part of the data audit was based around setting up custom tracking in Google Analytics. We also went through an extensive process of comparing Web Analytics data to the e-commerce platform and the back-end warehouse to ensure consistent reporting.

3. Data Mart Automation and Data Publishing
Where possible, we automated the data feeds — any source that had an API fell into that category (such as the client’s e-commerce platform provider or Google Analytics). Some sources didn’t have an API, but we were able to receive a daily automated e-mail with the relevant details and unpack the e-mail attachments to append the data to our DataMart.
Several select sources (such as marketing agency data, or financing data) could only be sent to us manually. In those cases, we copied/pasted the relevant data into the DataMart manually.

Once we had a recurring process for aggregating the data, we established a web service (OmnichannelBI) and a set of Excel documents to allow both the Client and Delve teams to quick access all the variables.

4. Reporting and Analysis
We released a variety of recurring and AdHoc reports:
- Recurring: Daily, weekly, monthly and quarterly reports, where each one focused on different elements of the business. For instance, we developed a detailed SKU-level merchandising report that is released daily; a KPI report that focuses on the relationship between visits, orders and AOV; and we guided the release of a comprehensive quarterly report that looked at the impact of seasonality, promotional activity, merchandising activity, television, online media and financing on call center and web sales.
- Ad Hoc: Besides an array of ad hoc web analytics reports that focused on different areas of the website, one of the more interesting analyses was around a local TV test across 30 DMAs. We were able to calculate the incremental impact from TV on website traffic and aid the client in understanding the importance of local TV on direct sales.

Results

Today the client has a well-established marketing business intelligence capability. We are intimately involved in day-to-day reporting and marketing efforts and work together with the client’s team to analyze sales drivers.