Alan Rimm-Kaufman will speaking on “New Developments In Paid Search” at the University of Virginia’s Darden School Of Business, as part of the executive education course, “Online Marketing Update At Darden.”

Rimm-Kaufman developed this executive education course in conjunction with Darden faculty members Paul Farris and Phil Pfeiffer.

The course was first offered in October 2007.

Given the overwhelming success of the first course, Online Marketing Update At Darden is being offered for the second time in April 2008.

Course link: Online Marketing Update at Darden, Spring 2008

On April 30, 2008, Larry Becker will be speaking at UVA’s Darden School Of Business on “Effective Websites” as part of the executive education course, “Online Marketing Update at Darden”.

Course link: Online Marketing Update @ Darden, April 2008

Alan Rimm-Kaufman will be speaking at MIT’s Operations Research Center on “The Business And Mathematics of Pay-Per-Click Bidding” on May 30, 2008 as part of the “OR And The Internet” conference.

Abstract:

Have you noticed those little text ads on the right hand side of a Google search result page?

Those modest links provide the lion’s share of Google’s $16B in annual revenue, and their economic clout is overturning the traditional worlds of advertising and media.

This talk will review the basics of paid search from the marketing and economic angles.

I’ll present the stochastic optimization problem which lies at the heart of paid search marketing.

I’ll discuss why the paid search bidding problem is hard.

I’ll give some flavor of the proprietary heuristics my firm uses to reach decent solutions to this problem millions of times each day.

I’ll describe our current research collaborations, one with Prof. Eddie Pinker (Rochester) and one with Profs Kirthi Kalyanam, Sharad Borle, and Peter Boatwright (Santa Clara, Rice, Carnegie Mellon, respectively).

I’ll describe the data sets we share with our academic partners, and wrap with an invitation for additional collaboration.

Q&A

Conference link: MIT’s OR And The Internet conference

Scene: A bar at a conference hotel during a marketing trade show.

Bill: Tim? Is that you?

Tim: Bill! Let me buy you a beer. How long’s it been? Four years? You’re looking great.

Bill: Thanks. I think we last met up at a catalog conference in 2008. So yes, about four years ago now.

Tim: Gee, have things changed.

Bill: Really.

Tim: And it seems to just keep getting worse. Would you ever have imagined we’d look back on 2008 as the “good old days?” Postage at 42 cents? Gas at $4? Sheesh, cataloging was easy then.

Bill: Too true.

Tim: Prospecting and acquisition mailing was still legal in ’08, right? The Never-Mail-Me Act hadn’t yet passed, the catalog co-ops were still in business and you could still rent lists. Remember the old list rental business? All those brokers and managers? And back then Catalog Choice was the only Never-Mail house.

Bill: Different times. But how’s your business?

Tim: We’re hanging in there, but it’s really tough. We’ve been shrinking — another round of layoffs — and we’re down to only three titles. We keep losing sales to the Web pure-plays, and the brick-and-mortar retailers have killer virtual reality stores online. We’re still mailing catalogs to our buyers, but the response just isn’t there. How’s your work going?

Bill: Oh great, really great. I started a new job last fall, and that’s a blast. Great people, interesting projects. Frankly, we’re struggling to keep up with the growth.

Tim: No way! What cataloger are you working for?

Bill: Oh, no, I left the catalog industry. I joined Google just after the Yahoo! acquisition. I’m managing the sporting goods category for GoogleMall. You wouldn’t believe the power of our customer-intent predictive models …

9 Predictions for the Not-Too-Distant Future

Perhaps the future of cataloging won’t be so bleak. Perhaps it’ll never get as bad as that conversation above. But it might. The catalog industry is changing at dizzying speed. Can you see your current business model working in five years? If not, the time to start making changes is now.

Here are nine predictions. All are reasonable, but taken together, they paint a bleak picture for traditional cataloging.

1. The “Never-Mail-Me” movement will grow. The organization may or may not be Catalog Choice, but there’s growing public sentiment that cataloging is unacceptably wasteful. Look how quickly the Do-Not-Call concept swept the country.

In months, 50 million people opted out of telemarketing, and Congress quickly passed the Do Not Call Registry in 2003. A similar thing happened to e-mail with CAN-SPAM.

If you don’t think the same thing could happen to unsolicited catalog mailings, your head is in the sand. How will your business attract new customers when you can no longer mail prospects?

2. Postage will continue to rise steadily. As bulk mailers mail less, there will be less bulk fees to fill USPS coffers, leading to additional rate increases.

3. Fuel will continue to soar, driving up freight and delivery costs.

4. Product pricing will get extreme. There will be room at the bottom for retailers crazy enough to compete head-to-head with Wal-Mart. There will be room at the top for ultra-luxury, high-cost goods. The middle ground of pricing will evaporate.

5. There will be no room for poor service. Customers will expect all Web sites to behave like Amazon.com and eBags, all call centers to behave like Lands’ End and Crutchfield, and all customer service departments to behave like Nordstrom and L.L. Bean.

This level of service will be expensive. Search engines will present retailers’ ratings right alongside their product listings.

6. “Push” marketing channels like e-mail will die. They’ll be killed by spam. Marketers will need to compete via “pull” marketing channels like RSS. Permission-to-market will become the direct retailer’s most valuable asset.

7. Merchandise depth will recede. General merchandise catalogs and gift catalogs will flounder, losing share to specialty retailers — merchants who understand the long tail.

8. Google will own the top of the purchase funnel. Online retailers will need top-notch paid and natural search to survive.

9. Consumers will expect free shipping. And they won’t be fooled by retailers who roll those costs into their pricing.

These visions sound bleak, but I’m not a pessimist. I believe the future of direct selling is bright. The Internet lets us communicate our message to the entire world at essentially zero cost. The only trick is saying something the world wants to listen to.

The availability of detailed, predictive data is soaring, and the cost of computing is plummeting. That helps us. And, from a total energy footprint, direct selling may be more environmentally friendly than traditional retail.

I’m seriously concerned that the traditional cataloging model may not survive. I’m simultaneously bullish about the “new” direct marketing. The future belongs to the brands and business models that evolve fastest. Are you ready?

This article assumes that insights gained from Web testing lead to effective print catalog presentations. That is our “Big Assumption.” If you disagree, skip to the next article. No hard feelings; see you next month.

Still reading? Good. There are two decent reasons to accept the Big Assumption.

First, the web is so prevalent that the demographics of web shoppers are pretty much the same as the overall demographics of all direct shoppers. The folks experiencing your online tests are probably somewhat similar statistically to your catalog prospects, making the Big Assumption reasonable.

Second, the web offers catalogers our first chance to do real testing. While many catalogers think they’ve been testing for years, they really haven’t been able to do so effectively. In-the-mail testing is too slow. It’s also inflexible and expensive.

All of that makes the Big Assumption attractive.

What You Can Learn

Accepting the Big Assumption, here are five starter ideas on using online testing to improve your catalog:

• Use online testing to determine which element of your Unique Selling Proposition is most powerful. Suppose your firm offers fair prices; cheap, fast shipping; a liberal return policy; great phone support and decades of experience. All well and good. But which of these is most effective at getting prospects to buy? Web testing can answer that, quickly and inexpensively.
• Use online testing to determine product breadth. Should you sell the single best product in each product category? Should you present a three level good-better-best selection in each product category? Or should you sell every SKU your merchants can source? This a big question with big implications, and with different answers for every retailer. You can use online testing to run head-to-head tests of product selection. If tests across different categories yield similar results you can take those insights back to your paper pagination.
• Use online testing to develop tag lines and copy for your “hero” products. I know at least two best-selling authors who used online testing to pick the best titles for their books: Tim Ferriss, “The 4-Hour work Week: Escape 9-5, Live Anywhere, and Join the New Rich”; and Ian Ayres, “Super Crunchers: Why Thinking-by-Numbers Is the New Way to Be Smart”.
• Use online testing to evaluate promotions. Which holiday promotion works best: “Free Shipping On Orders Over $100” or “Buy Three Widgets And Get The Fourth FREE”? I don’t know. You don’t, either. Web testing lets you observe the response lift and the discounting costs of each promotion before committing to something in print.
• Use online testing to for pricing tests. For the brave, but possible.

Of course not all online insights generalize to print. For example, I wouldn’t use homepage layout to inform catalog cover design; they’re simply too different. But if you grant me the Big Assumption, you can bring web messaging insights back to the print page.

To get going, all you need are some ideas about what to test. After being shackled for years by the chains of in-the-mail testing, you and your team likely will have ideas in abundance.

You need testing software. You can get great Web testing software and spend five or six figures. You can also get great Web testing software for free (see below).

And you need someone with enough basic HTML chops to execute the presentation of your different ideas. While decent HTML will suffice, someone with Javascript and DOM skills will really let you turn on the magic.

That’s it. You could be running tests on your site by the end of next week.

How It Works

There are many fine testing platforms available, including as Offermatica, Optimost, and SiteSpect. Let’s say you choose to use Google’s powerful Website Optimizer (GWO), which is free for AdWords advertisers. You don’t even have to install software or buy additional hardware; GWO runs on Google’s servers.

To start with GWO, insert a small Javascript snippet on your site’s order confirmation page. This is the page that comes up just after the user clicks “submit.” The script informs GWO when an order occurs.

Next, you set up a test by selecting a page or a region common to multiple pages, such as top navigation. For a split test, create multiple versions of the whole page. More interesting are multivariate tests (MVT), where you choose several area on the page to vary at the same time. After loading these regions and versions into GWO, you turn on the test.

Each visitor will randomly experience one of the different page combinations. GWO will keep each visitor in that specific treatment for their entire visit and subsequent visits so users won’t experience any inconsistencies while they surf. GWO tracks which tested treatments produce that highest conversion rates. It provides statistical confidence intervals on its results. It also parcels out the contribution from each tested region.

For a sample test, click on “The 3×3 Experiment” under related content to the right of this story.

Make Better Decisions

You can run tests with several MVT regions and several versions of a region. You can run tests across multiple pages or tests that compare site processes, such as one-page vs. multipage checkout. And you don’t have to use GWO; similar experiments can be run with most paid MVT testing tools.

Imagine the power of settling catalog design disputes by watching your customers. If small tweaks proven to increase conversion rates, the same winning formulas could help your catalogs.

For multichannel merchants, the age of online testing has arrived. The tools are powerful, easy-to-use, and - at least in one case - free! Catalogs aren’t going extinct any time soon. But, as in-the-mail response rates continue to fall and postage and fuel continue to rise, savvy mailers must lean on online testing to give their books every advantage.

Alan Rimm-Kaufman is CEO of the Rimm-Kaufman Group, a Web agency providing pay-per-click search management and site effectiveness testing for online retailers. You can reach Alan online at www.rkgblog.com.

George Michie will be moderating a panel on affiliate marketing at IRC 2008.

For many years, George has been (and remains today) highly critical of the affiliate channel, so this promises to be a lively session!

Joining George on the panel will be Barbara Hurd, VP of Business Development, Harry and David; and Vicki Updike, VP of Merchandising and Marketing, Miles Kimball.

Affiliate fan or foe, come watch the fireworks!

Alan Rimm-Kaufman will be speaking at the Shop.org Marketing Workshop in Scottsdale, April 7 — 9.

Alan will be leading roundtables discussing off-site testing via paid search (copy, landing pages, matchtypes) and on-site testing via MVT (layout, flow, headline, value proposition, design).

Thanks for listening to my talk at NEMOA, 13 March 2008. Here’s a list of links discussed.

Cheers –

Alan

PS: Add us to your feedreader: http://feeds.rimmkaufman.com/rkgblog

Links & Resources

NEW! Watch the video: 45 Web Marketing Ideas For Online Retailers (Video)

• NEMOA
• 120′ water balloon stunt
• 120′ water balloon analysis
• Cluetrain
• Google Reader
• Google alerts
• Andy Sernowitz WOM book
• Scoble & Israel Naked Conversations
• Debbe Weil Corporate Blogging Book
• Wordpress
• WordPress plugins
• coComment
• Feedburner
• Technorati
• Seth’s Blog
• GapingVoid
• John Chow
• Shoe Money
• Neil Patel
• Occams Razor
• 4Q
• High Performance Web Sites
• Google Webmaster Tools
• Firebug
• Google Analytics
• Sitemaps
• Google Website Optimizer
• Yahoo Pipes
• Semoz Factors
• BulletProof Web Design
• Simple Bits
• PPC Into Call Center
• MineThatData
• Isabella Oliver
• VAT19
• Wine Library TV
• WillItBlend
• ShopInPrivate
• Heinz Top This
• ShareThis
• Ning
• ThinkGeek
• PayPal
• BillMeLater
• HackerSafe
• Do It Wrong Quickly
• rkg blog

Add rkgblog to your feedreader: http://feeds.rimmkaufman.com/rkgblog

Slides

Powerpoint slides: 39 Ways To Improve Your Web Batting Average (large: 8meg ppt)

Video

Watch the video: 45 Web Marketing Ideas For Online Retailers (Video)

Start the new year off right by getting up to (search) speed

Are you on top of today’s hottest ideas in free and paid search? Here are 14 easy-to-implement ways to get your site to the top of everyone’s results. Each could support a full article in its own right, so I’ve also provided additional links to help you dig in further.

Free Search Engine Optimization (SEO)

1. Social media sites drive links; links drive rankings. Get familiar with Digg (digg.com), StumbleUpon (www.stumbleupon.com), Netscape (netscape.com) and Reddit (reddit.com), because these social-media sites can drive huge traffic. More importantly, that traffic leads to numerous inbound links, which are the rocket fuel powering your organic rankings.

Not all of your content is Digg-worthy, but see if a few times a year you can develop content so compelling that it warrants attention from the social-media tribes.

2. Markup matters. Inbound links are critical, but markup still matters. Use clean, well-formed, validating HTML. Pay special attention to your title tags, and give each page a unique title. Use the robots-nocontent tag to tell the spiders to ignore your navigation.

3. Embrace cascading style sheets (CSS). These separate presentation from content. Well-designed CSS achieves the following:

● your pages load and render more quickly;

● searchbots better understand your page; and

● your site will be friendlier to disabled users and mobile devices. When using JavaScript enhancement, make sure your site degrades gracefully for older browsers. For a thorough explanation of CSS, I suggest you check out www.csszengarden.com.

4. Blogs and RSS. Search engines love fresh content presented at regular intervals in machine-friendly format. Blogs and really simple syndication (RSS) provide just that. With proper publishing pings and tagging, the engines will discover and index your RSS content rapidly. Blogging is another great way to communicate with customers and prospects. There are many good blogging platforms; WordPress (see wordpress.org) is particularly popular given its rich set of plug-ins.

5. Google’s Universal Search offers new opportunities. Launched last May, Google’s Universal Search strives to merge relevant results from across Google’s specialized engines (news, images, video, blogs, etc.) into the core Google search results pages.

Universal Search offers savvy multichannel marketers back-door, first-page results on competitive terms. Provide the engines fresh and relevant images, video, blog posts and press releases that are suitably tagged and marked up.

6. Great free apps. Among the many free tools Google provides for AdWords advertisers, Google Analytics (google.com/analytics) is a powerful application for understanding how visitors reach and use your site. Google Website Optimizer (services.google.com/websiteoptimizer) is handy for multivariate testing.

Paid Search Engine Marketing (SEM)

1. Brand vs. nonbrand. Many multichannel marketers enjoy large sales with little cost on searches for their brand name. These searches aren’t incremental, however. Analyze your campaigns both with and without your branded terms. Challenge your paid search agency or in-house team to drive profitable sales growth out of your nonbrand portfolio.

2. Beware of broad match and content networks. Broad match allows the search engines some leeway in matching your ads against similar search queries. While broad match is a convenient tool and appropriate in some circumstances, often that additional traffic is of lower quality.

Test what happens to your pay-per-click (PPC) sales and your PPC costs when you remove broad match from your top terms. You may find considerable benefit.

As with broad match, syndicating your ads to the content networks can offer you greater click inventory. But again, this additional traffic can be of lower quality, increasing your PPC costs with minimal sales benefit.

For one thing, know if you’re running ads on the content networks. If so, test to determine if the additional costs are warranted by additional sales. There are good clicks to be found hidden in the content haystack, but finding them takes effort and care.

3. Bid for profit, not position. If you’re concerned with impressions and branding, bid your pay-per-click campaigns to hold your ads in the top positions. If, on the other hand, you’re concerned about profitable sales, bid your ads based on your economics.

Long-Tail Term Bidding

Track cost and sales at the ad (not ad group) level, and use statistical clustering approaches to help you bid your long-tail terms efficiently. All bid management systems are not created equal. Top-notch bid management algorithms pay for themselves many times over in incremental profit.

4. Comprehensive term lists. There’s great value in the long tail of search terms, and comprehensive term lists are your ticket to capturing that value. As a rule of thumb, test a number of terms equal to four to six times the number of pages on your site, expecting about a quarter or a third of those terms to have sufficient traffic and performance to merit ongoing funding in your campaigns.

For example, an online marketer with 4,500 SKUs and a 5,000-page Web site should test 20,000 to 30,000 terms, expecting 5,000 to 10,000 of these terms to work long-term.

Keyword counts (adwords.google.com/select/KeywordToolExternal) can be inflated artificially with no benefit — for example, by pre-pending “buy” or post-pending “online” — your object is real unique phrases with actual traffic and marketing opportunities.

5. Build rational ad groups. Your Google campaigns will perform better when you group ads semantically. Sensible ad groups help your performance on the content networks and also can improve your quality score. (See adwords.google.com/select/guidelines.html.)

6. Quality score matters. Google ranks ads by a combination of maximum bid and quality score. Quality score is a function of clickthrough rate, landing page relevance and additional secret factors. To avoid quality score concerns, write targeted and compelling copy, ensure that your destination URLs tightly match the ad phrase and ad copy, and don’t launch new campaigns with artificially low costs per click.

7. Conversion is the ultimate PPC battleground. In the long run, multichannel marketers who get the most from each paid click to their site can afford to pay the most for those clicks. Do everything you can to increase site conversion.

Offer great merchandise, fair pricing, free or reduced shipping, strong guarantees, shopper-friendly return policies, user testing, multivariate site testing, Web-effectiveness projects and so on, and you’ll do fine.

Increasing your sales-per-visit metric allows you to invest more in advertising. Higher bids place your ad higher on the page, scaling your sales and profits considerably.

Both SEO & SEM

1. My SERPs aren’t your SERPs. Increasingly, search engines serve customized search engine results pages (SERPs) based on profiles and prior searches. To view search results without this user-specific “contamination,” use Firefox and install the CookieSwap add-on. Also, try Google’s Ad Preview tool to see your paid results as a normal searcher does.

Hopefully some of these tidbits spark new ideas for you — best of luck with all your search marketing efforts!

Alan Rimm-Kaufman is CEO of the Rimm-Kaufman Group, an online paid search and Web-effectiveness consulting firm. You can reach him online at www.rkgblog.com.

11 steps to get you up to speed

Is your company blogging yet? With the soaring cost of postage and pay-per-click advertising, blogging offers catalogers a powerful channel to get their message to prospects and customers worldwide. The cost is low. The impact is high. Here are 11 steps to get you started.

1. READ LOTS OF BLOGS. Before you start your own blog, spend at least two weeks voraciously reading. Choose an RSS reader; I recommend Google Reader. Get familiar with blog search engines, such as Technorati. Find bloggers writing about niches related to your company, products, industry, customers, etc.

Add bloggers you like to your RSS reader. Follow their link-outs to find more bloggers. Intentionally oversubscribe to feeds. Aim for 50 feeds in your RSS reader by the end of the second week. Oversubscribing helps you understand which blogs keep your attention. It also lets you experience how readers decide to subscribe and unsubscribe. After you hit overload, pare down your reading list to a comfortable level.

2. READ BOOKS. To understand why blogs are revolutionary, read these books: “Naked Conversations: How Blogs are Changing the Way Businesses Talk with Customers” (Wiley, 2006) by Scoble and Israel, and “The Cluetrain Manifesto: The End of Business as Usual” (Perseus Books, 2001) by Locke, Levine, Searls and Weinberger. For stories and tactics, check basic blogging books, such as “Clear Blogging: How People Blogging Are Changing the World and How You Can Join Them” (Apress, 2007) by Bob Walsh.

3. MAKE COMMENTS. Before you start, get comfortable commenting on blogs in your niche. Commenting offers a gentle way to ease into the conversation.

4. PICK YOUR PLATFORM. Decide if you’ll download software or use a hosted application. I’m partial to WordPress because of its rich ecosystem of plug-ins and themes. I suggest running this free software in-house, if your IT team can accommodate.

5. PICK YOUR WRITERS. Decide who in your organization will blog — and how frequently. Some organizations have top management blogging: Check out Bob Lutz at GM (http://fastlane.gmblogs.com/archives/bob_lutz) or Richard Edelman at Edelman (www.edelman.com/speak_up/blog). Other organizations hire bloggers for their content, such as Ice.com’s blog Sparkle Like The Stars (www.sparklelikethestars.com/) and BzzAgent’s “90 Days” experiment last year (http://90days.bzzagent.com). Blogging is a serious time commitment. Make sure your designated writers can produce solid content, consistently.

6. PICK YOUR TOPIC. A blog needs a focus and an objective. Decide what your blog is about — and why.

Firms blog for many reasons:

• to attract prospective customers;
• to communicate with vendors;
• to establish expertise;
• to attract potential employees; and
• to communicate with Wall Street.

Different objectives dictate different writing styles and topics. For example, Google maintains one official blog in multiple languages (English version at www.googleblog.blogspot.com) for announcements and news. It has another 50 or so blogs for specific products. An example: Inside AdSense at www.adsense.blogspot.com. Google officially recognizes another 50 or so employee blogs. Check out Matt Cutts at www.mattcutts.com/blog or Christopher Sacca at www.whatisleft.org. While Matt Cutts can blog about the Spicy Chicken Sandwich at Jack in the Box, such a post would be out of place on Google’s main blog.

7. START WITH A PRIVATE BLOG. No need to practice in front of the world. If you’re hosting your blog internally, don’t allow outsiders from beyond the firewall to reach it. If you’re using a hosted blog, try marking posts as private, restricting readership to folks within your company.

Treat the posts as public. Use this internal test period to work out technical, branding, visual or legal hurdles for a month or so. See what frequency of blogging your organization can support. I recommend two posts a week. Choosing a posting frequency and posting consistently builds readership.

8. GO PUBLIC. After a test period of private blogging, commit to a public blog. Launch your public blog with the posts you made during your test period, so your shiny new blog enters the visible blogosphere with a hefty amount of established content.

9. LEARN MONETIZATION. Never run other people’s ads on your blog. Your team should read bloggers who cover blog monetization, however, because they often provide great tips for increasing readership, engagement and inbound links. I regularly read John Chow (www.johnchow.com). He offers great advice on gaining blog traffic, and is far less evil than he purports.

10. MOD YOUR BLOG. Most blogging platforms allow you to customize your design, change functionality, embed video and add widgets. Enliven your basic blog with brand-consistent gadgets and bling. A little bit goes a long away. (For a great, intentionally over-widgeted blog, check out Fred Wilson’s A VC blog at www.avc.blogs.com).

11. TRACK YOUR SUCCESS. You’re doing something right if your blog gains readers, subscribers, comments and inbound links. Track these metrics with FeedBurner, Google Analytics and Technorati. True success is when your blog gets into the mainstream press, is mentioned positively by vendors and customers, and starts generating sales.

Read Hugh MacLeod on increasing revenue at Stormhoek winery five-fold through blogging (www.gapingvoid.com/Moveable_Type/archives/003577.html).

Best of luck, and see you around the blogosphere!

Alan Rimm-Kaufman is CEO of the Rimm-Kaufman Group, an online agency helping catalogers manage paid search and improve site conversion. You can reach him at (434) 970-1010 or at www.rkgblog.com.