Hack81.Know How to Use Retail Analytics

Hack 81. Know How to Use Retail Analytics

While not exactly a class of its own, web measurement for online retailers entails fairly specific uses of the tools and hacks described in this book; knowing how and when to leverage these tools can significantly increase your ability to market, merchandise, and sell.

Learning how to use retail web measurement applications may equip you to sustain the double-digit growth you have become accustomed to as an online retailer. Pouring additional dollars into your marketing budget might also do the trick, but web measurement offers an inexpensive and easy way to achieve quick wins that can translate into additional sales, more active customers, and marketing cost savings.

Retail web measurement applications leverage most of the web measurement concepts, tools, and hacks you already know; extend a few of these items to better address specific retail scenarios; and apply the resulting measurement approaches and nomenclature to a fairly straightforward set of objectives:

• Attract highly qualified visitors

• Convert these visitors into repeat customers

• Generate more sales

• Do all of the above with less (fewer marketing dollars, less time, etc.)

Start with a simple retail web measurement framework that includes the following analyses. Even if you have already performed an analysis of your checkout funnel and managed to increase completion rate by a few percentage points, this hack introduces two additional exercises you can quickly perform to improve your site's ability to attract, convert, and retain customers.

6.2.1. Optimize Your Marketing Spend

Optimizing your online marketing spend is an almost failsafe way to drive additional visitor acquisition and site revenue by reallocating marketing dollars to the most efficient marketing channels and programs. If cost savings is your primary goal, you can use this same exercise to identify laggard marketing programs that you no longer need to invest in.

You need to optimize your marketing expenditures via tracking and analysis of key data:

• Track individual marketing elements (such as search keywords, banners, affiliates, and emails)

• Analyze the performance of these elements in terms of visitor loyalty, conversion and sales, and other descriptive retail metrics, such as shopping cart abandonment rate and average order value

Many web measurement vendors generate a report that describes all marketing activity using multiple fields: vendor, channel, campaign, message type, creative format, and URL, for example (Figure 6-1). You can sort the report by each of the retail metrics mentioned above or by a combination of them.

Be careful not to combine more than two or three different sorts, or your resulting analysis may suffer.

You'd be surprised at how much additional information you can glean simply by sorting columns:

• If your primary goal is to drive traffic to your site, pure and simple, sort by visitors or visits.

• If your primary goal is to attract big spenders, sort by average order value.

Optimize your marketing spend by aligning your investments with your business goals. By reallocating 30 percent of your online marketing budget to the relationships that drive the most traffic by volume, the most valuable traffic, the highest sales, or the largest order values, you can quickly recover marketing losses and better achieve the specific goals of your retail site.

6.2.2. Analyze Product Placement and Look to Book

On a retail web site, shoppers make a series of decisions before purchasing online. They look for products of interest in the places they most expect to find them. If they find them, they make a near instant judgment of whether the product is presented, described, and priced in a compelling way.

Figure 6-1. Breakdown of marketing programs

You should leverage your web measurement tool to explore how shoppers make decisions, using the following steps:

1. Generate reports that list all products viewed for a specific time period.

2. Start by looking at a time period of a week or longer.

3. Examine product metrics by product category or site location to get the "big picture" view of how shoppers are browsing your products.

4. Drill down, reviewing key merchandising metrics like product views, abandonment rate, orders, and sales for each product.

5. Analyze product placement by looking at abandonment rate and look-to-book ratios (product views divided by orders for a particular product) aggregated at the category level.

Categories with high values for product views but low look-to-book ratios may be attracting shoppers in search of products that they wrongly expect to find. Solve this problem by recategorizing products, placing products in multiple categories, and enhancing site navigation or search. Whatever the root cause, unresolved problems with product placement lead to lost sales and lost potential customers.

Analyze specific product performance by looking at abandonment rate and look-to-book ratio at the product level. Products with high abandonment rates may have excessive shipping and handling charges that are revealed only after the product is placed in the shopping cart. Products with high look-to-book ratios may be poorly presented, unattractively priced, or labeled as out of stock. Address problems with poor look-to-book ratios by enhancing site navigation, site search, product imaging and presentation, price, and shipping options.

6.2.3. Streamline the Checkout Funnel

A typical checkout sequence includes steps for:

• Shopping cart

• Login or registration

• Shipping and billing address entry

• Payment information

• Coupon or promotion code entry

• Order review

You need to measure each of these attrition points not only in terms of completion or defection, but also by shopping cart value, time to complete, and new versus repeat buyer status [Hack #52]. You can remedy many defection problems by applying web interface usability best practices to key information entry pages, such as registration and address entry. Fortunately, many of these best practices are well understood:

• Observing the differences between new and repeat buyer checkout behavior may expose an opportunity to redesign the checkout sequence to incorporate passive site registration.

• Steps with high time to complete values may be too laden with heavy graphics, excessively long (in the case of an information entry page), or text-heavy, or they may lack necessary security precautions and messaging.

• High abandonment at the login or registration steps may indicate that you haven't done enough to demonstrate reasons that new shoppers would want to spend the time to do this.

• Abandonment during coupon or promotion-code entry steps may indicate discontinuity in how you're presenting and accepting these codesfor example, you tell them the code is "123-XYZ-PDQ-2" but the form does not accept the dash (-) character.

• Higher-than-expected abandonment of shopping carts may indicate that your proceed-to-checkout button is not visible enough or is buried on the page somehow. All checkout buttons should be highly visible on the page, even if they look ugly.

• High abandonment from your shipping and billing address pages may indicate that you're asking for too much, you're not clearly indicating which fields are required, or there is a problem with how you're processing the form. The standard recommendation is to ask only for the minimum amount of data needed to complete the order.

• High abandonment from the payment information and order review steps may indicate that you're not doing enough to convey your commitment to the shopper's security and privacy. Make sure you have TRUSTe and BBB logos [Hack #26] on pages where you're collecting personal or financial information.

Combined with effective marketing and merchandising analyses, regular optimization of the checkout and registration funnels is the foundation of retail web analytics.

Brett Hurt and Eric T. Peterson