Targeted Selling Webinar Recap

This post is a recap of our July webinar Creating relevant shopping experiences through targeted selling. If you missed the live event, I encourage you to catch the replay (click that hyperlink!) which should be available shortly.

The webinar includes a number of retailer examples of targeted selling, each with their own screenshots. To keep this post to a reasonable length I have only included select images, but you can see them all in the replay.

Agenda

• What does personalization mean?
• What is targeted selling?
• Benefits of targeted selling
• How to target content
• Retailer targeting examples
• Pitfalls of targeted selling
• Building segmentation rules

What does personalization mean?

The term “personalization” is used to describe a number of things, including:

Physical Product Many products can be customized to order -- as simple as engraving a name on an iPod or adding a corporate logo to promotional products all the way to designing a personal Nike shoe. While this creates somewhat of a personalized product, it has nothing to do with the web experience.

User Interfaces Many websites allow you to personalize the look and feel, whether it's a MySpace background (perhaps Twitter is more relevant?) or adding/moving applications and boxes on a Facebook profile. Personalized UIs are rare in ecommerce since customers are on the site to shop or research products, not engage in self-expression.

I've only heard of one shopping site that allows UI configuration, and even then, it's minimal. The Mobissimo travel search engine allows you to add and move widgets:

This type of web 2.0 functionality is a very low priority for commercial websites. This still doesn't personalize the shopping experience.

Welcome Back, Kotter! A small step toward personalizing the shopping experience is to take the visitor’s name and welcome them on the home page or email subject line. Though that can have a warm-fuzzy impact, it’s a very small piece of the “Holy Grail” of online merchandising which is the full blown relevant web experience, where the commerce system is delivering content and offers that are tailored to the individual.

Personalization of the Shopping Experience Tailoring content and offers to customers based on what you know about them in their account profiles, clickstream data, referring keywords and other clues is more along the lines of the personalized/targeted selling experience.

What is targeted selling?

We define targeted selling as “the delivery of content and offers to different site visitors and customers based on what you know about them."

Is there a difference between targeted selling and personalization? Yes. As defined above, targeted selling covers both one-to-one (personalization) and one-to-many (segmentation) strategies. You can engage in targeted selling without personalization tools if you're just using segmentation.

For example, your targeted selling strategy could be as simple as "first time visitors see Home Page A" and "returning visitors see Home Page B." This one-to-many strategy is targeted, but not personalized. It could also use more complex segmentation rules like "show X offer to returning customers, who have average order values of $150 or more, who live in the US, who have browsed the “diamond jewelry” section of our site in the last 45 days."

Personalization might be "when customer returns, show him/her the last 3 products viewed, with 20% discount banner for items in last category viewed." There is no pre-determined segmentation rule.

Segmentation data sources

HTTP headers

• Referral source
• Referral keyword
• Browser
• Operating System
• Language

Cookies (*may or may not have)

• Returning visit or new
• Customer ID*
• Days since last visit*
• Affiliate code*

Geo location

• Country
• State / Province
• City
• Connection speed
• Connection type

Order history

• Purchase trends
• Gifting trends
• Average order size
• Frequency of purchases

Customer profile

• Gender
• Age
• Preferences
• Loyalty program status

On-site behavior

• Click stream
• Idling
• Search terms
• Previously viewed products
• What’s in their cart / wish list / registry
• Products they are tracking
• Promotions used
• Products reviewed

Benefits of targeted selling

• Increase average order value / items per sale
• Save money (retailer)
• Boost customer loyalty
• Improve customer experience / customer service

Types of targeted selling

Content

• Banners / images
• Product assortment (searchandising, SEOandising, merchandising zones)
• Cross-sell, up-sell, alternatives
• Custom catalog (B2B, geotargeting, "Personal Shop")
• Campaign specific landing pages

Pricing

• Currency conversion
• Price lists (employee, distributor, loyalty club member, wholesaler)
• Cart promotions / discounts / coupon codes

Experience Optimization

• Language & currency selection
• Redirects for geography, mobile devices, connection speed
• Optimized checkout workflow (credit check for new customers)

Offline Targeting

• Triggered email / SMS

Save the Sale

• Persisted information (cart contents, recently viewed items)
• Live chat prompts

Building Segmentation Rules

Once you’ve decided which data sources you want to leverage to segment your customers, you’ll need to make sure that the business can easily build targeting rules using these data points. For example, Elastic Path allows the business user or marketer to build segmentation rules via a simple rule builder:

By creating saved segments you can leverage the same rule (or segment) to deliver a variety of targeted campaigns to your customers.

What can be targeted?

Content

• Banners / images
• Product assortment (searchandising, SEOandising, merchandising zones)
• Cross-sell, up-sell, alternatives
• Custom catalog (B2B, geotargeted, "Personal Shop")
• Campaign specific landing pages

Pricing

• Currency conversion
• Price lists (employee, distributer, loyalty club member, wholesaler)
• Cart promotions / discounts / coupon codes

Experience optimization

• Language & currency selection
• Redirects for geography, mobile devices, connection speed
• Optimized checkout workflow (credit check for new customers

Offline

• Triggered email / SMS

Save the sale

• Persisted information (cart contents, recently viewed items)
• Live chat prompts

Retailer examples of targeted selling

Home pages

Amazon recognizes if you're a logged in account holder or not and welcomes you with your name, and a link to recommendations. Using geo-IP location, it recognizes your country. Using personalization, Amazon remembers products you have viewed or purchased previously, and merchandises the site accordingly. It also uses a persistent shopping cart, keeping contents in your cart indefinitely until you purchase.

You can also show different home page designs to different types of visitors. This is an idea for Danskin - show the left image when referring search term includes the word "plus-size" or the right when the term includes "rythmic," "gymnastics" or "leotard."

SAmazon recommendations change daily, based on profile and site behavior.BannersCustomer profiles

Profiles - feedback

Customers can tell Amazon which recommendations are relevant and which are not.

Profiles - self-identification

Savvy retailers like Disney Shopping (left) and eBags (right) gather optional customer information when the account is created.

Profiles for custom catalogs

MyShape.com delivers "Personal Shops" to each customer by gathering detailed information including exact body measurements, style, color, fit and brand preferences.

Product Pages

Cross-sells, upsells and alternative recommendations are common on retail websites.

Landing Pages - SEO-andizing

SEO-andizing refers to showing something different based on the referring keyword from a search engine or paid search ad. This is what you’d normally see on the Cable Organizer site if you browse to this cable turtle page:

If you were to find Cable Organizer in a search engine for the term “cable turtle” it serves up a “BEST MATCHES FOR CABLE TURTLE” so that if you don’t like the landing page item you got, you immediately see relevant alternatives based on keyword relevance. It's like a mini search results page, baked right into the product page:

Category Pages - Geotargeting

Overstock does a great job detecting my Canadian location. It shows the Canadian flag for instant recognition, Canadian prices and indicates when an item will not ship to Canada with a "US Only" badge in category page results.

Search

Could be as simple as remembering my sort by preference in all future searches. So if I search for bird cage, sort by price low to high, and then large bird cage, when I perform my second search, I don't need to re-sort again.

Drugstore.com "searchandises" results when it doesn't carry an item.

If you're logged into your Personal Shop at MyShape and you search for "cocktail dress," you're only going to see results in your size, that matches your shape and your color, brand and style preferences that you identified in your profile.

Live Chat

I spotted 123inkjets.com using a live chat prompt with a 10% offer to return to the site after the visitor closes the window/tab. Other retailers use rules before triggering a live chat prompt because CSR resources are so expensive. You might want to trigger a chat if someone bounces back and forth within a category, or performs X number of searches and still hasn’t added anything to cart, or is looking at an item that is high dollar value. You wouldn’t want to offer chat for low dollar items, or for visitors that are browsing the clearance section, or live outside the country you ship to.

Providing CSRs with cross-sell capability is important to industries like financial services, travel or B2B. Baking targeted selling into the Customer Service portal is also a profitable idea.

Cart Summary

Showing cross-sells and up-sells on the cart summary page can increase average order value. Most personalization tools allow you to use rules to make the recommendations more relevant. For example, when a silver item is in the cart, suggest silver polish and one other complementary silver item.

Cart Summary - Cart Promotions

Cart promotions depend on what’s in the shopping cart. So you can set rules like “when XYZ product is in the cart, show a reminder for a warranty upsell." Or "when cart contents less than $100, show notification with exact dollar amount the customer needs to add to cart to get free shipping or a free gift."

Customer Service

Amazon notifies customers when an item in a saved cart has a change in price.

Mobile

Amazon mobile pulls information from your account so your recommendations the same on both devices.

Triggered Email - Post Purchase

Amazon sends a triggered email asking for a review after purchase, and special offers based on past purchases.

Avoiding Common Mistakes

• Irrelevant items and offers - don't use your recommendation engine right out of the box. Understand the "rules" you can use to control your assortment
• Blacklist out of stock items
• Don't target if you're not prepared to measure
• Have a process understood by all who touch the website
• Perform quality assurance

Next Webinar

From Manufacturer to Retailer: Expanding Your Brand through Ecommerce

Consumer expectations and the popularity of online shopping can be tempting growth opportunities for manufacturers. Learn about the unique opportunities, options, and challenges facing branded manufacturers as they make the migration to ecommerce.

Presented by Ecommerce Consultant Sally McKenzie

Date: Tuesday, August 25, 2009
Time: 9am Pacific/ 12pm Eastern
Register today!