Google's FMCG tips and building loyalty in e-commerce - stories from day one of Ad:Tech Chicago

Geoffrey Precourt
WARC Online

In the opening presentation, Drew Ianni - the Ad:Tech advisory board chairman - argued that the current digital advertising industry is shaped as much by technology as by ads: “Ad:Tech started in 1996. Online advertising and interactive marketing is our heritage. Today, we’re talking about how digital is transforming all media and the business of marketing communications - media, content, consumer behaviour, branding, strategy, mobile, wifi, search, email, eCommerce, measurement, and performance.”


The digital industry, Ianni explained, is not maturing along transitional growth lines. While online advertising is projected to grow from $25bn in 2008, to $50bn by 2012, Ianni said: “We are not maturing in the standard fashion. More often, we’ve experiencing waves of consolidation and mini-business cycles. The current cycle is driven by a slowing economy and industry collaboration.”


To illustrate the former, he cited reports from the Interactive Advertising Bureau: while the online spending level for the first quarter of 2008 ($6.8bn) represented an 18% increase over the same quarter last year, it fell far short of original projections of 23-24%.


And, as an example of the latter, Ianni predicted that the top 300 ad networks probably will be consolidated to just a few dozen by 2010. “Most significantly, there will be a consolidation in the top ten,” he added, predicting the same kind of tightening in the second- and third-tier networks.


Other economic drivers for digital will be the decline of other media, Ianni added, citing specifically the vulnerability of the $48.6bn currently invested in newspapers, radio’s $19.8bn slice, and an additional $15bn spent on Yellow Pages advertising.


He said: “From the latest spending reports, that means there’s $84bn in play…. As budgets tighten, the battlegrounds have been established. Local is a huge thing for digital. The battle for those dollars will be intense.”


And, while he allowed that the Mobile Revolution has not arrived as quickly or as forcefully as industry analysts had predicted, he argued that “the iPhone revolution is well under way.”


This was buoyed by applications “currently numbering 400, with 20 additions a day.” And, he added, that while Apple sold 6 million first-generation iPhones, the company currently is ordering up 800,000 new 3G units a week, with sales projections reaching 10-12 million by Christmas.


A CPG Marketing Master in the House of Google


When you listen to Kevin Kells, Google industry development director for consumer packaged goods, it’s not so much the voice of a zealot you hear as much as the sound of a kid in a candy store.


For 15 years, Kells had worked in the CPG space, managing and growing such iconic brands as Pond’s (Unilever), Guinness (Diageo) and, until he joined Google, Revlon, where he was vp/marketing for Almay Cosmetics. It was almost as if he’d come from a completely different industry, a point of difference he was quick to recognize as he identified the three stages of digital commerce:


  • Business 1.0: Information & Communication - The Internet as Brochureware: In the mid-90s, the primary function of the Internet was to let people know what business you were in and to communicate back and forth.
  • Business 2.0: Commerce - The Internet as Sales Channel : The World Wide Web evolved into a place where you not only could find out about goods and services, but actually buy them
  • Business 3.0 Community Self Expression & Entertainment - Engage with Your Customer: Kells said of the industry in which his employer plays such a critical role: “The digital space begins to offer engagement at a deeper level. It offers better insights that provide the opportunity to do storytelling to the people who matter to you.” It’s a place, he adds, where marketers can search for connections with the right consumers grounded not in broad-based demographics but specific insights.

In this world, Kells continued, the balance of power - the focus of almost his entire pre-Google career - shifts. Marketers used to search for people; now, informed shoppers are shopping for the marketers’ products and services. As Kells put it: “In millions of millions of searchers, people are raising their hands and saying, ‘I care about what you’re saying.’”


This new Business 3.0 universe has, Kells believes, three key drivers of change: technological breakthroughs, the embracement of search as a “BIG” idea, and the effectiveness decade.


Technological breakthroughs are evident is refinements in the tools of production, increases in broadband penetration, and declines in the costs of digital storage - three 21st Century developments that have “democratized digital media,” Kells claimed.


Specifically: “Today, anyone can record music online and distribute it or shoot a home movie - complete with special effects - and have a global audience tap into it. Anyone can write a blog. The tools of production have taken the control of the production of content from a few to many - in fact, to all of us. The brands that perform well will wrap their hands around the notion that it’s not just advertising business they’re in, but the content business, full stop.”


Furthermore, broadband is now available in some 300 million households worldwide. And, because the costs of digital storage decline as the amount of available storage increases - especially in on-the-go devices - information has mobility: “Every 13 months, the amount of storage doubles,” Kells noted.


The iPhone that currently holds thousands of songs will give way to a new device in 2012 that will carry a full year of video with no repeat performances. And, a few years later, that same sized handheld device will store a lifetime of video.


The bottom line of technology breakthroughs: everyone has access to information. And everyone is in the content game.


And, because he’s moved from the CPG space to the quantitative environment of search, Kells backed up his notion of change with a number of hard numbers:


  • 1.4bn people are online all over the world.
  • 500,000 people were online in 2003.
  • 188 million people are online in the U.S.
  • 64% of U.S. households have broadband.
  • U.S. Internet users typically spend 15 hours a week online.
  • $420bn was the total of 2007 eCommerce sales totaled.
  • 83bn emails sent every day.
  • 2,000 emails are sent every second.
  • 71% of US Internet users view video online 11bn videos are streamed online every month in America
  • 1bn mobile devices were sold in 20 years.
  • 1bn mobile devices were sold in the next four years
  • 1bn mobile devices were sold in the following two years.
  • 6% of sales are completed on the Web
  • 50% of all sales are affected by online search before purchase

Search is a BIG idea and a core behavior: “Consumers use search to discover what they care about,” said Kells, “and the digital landscape affords [marketers] the opportunity to track their behavior. And that knowledge, he continued, increases the sophistication of insight and the ability to “get more insight and tell more and better stories more often.”


Because the insight that not one marketing program fits all sizes of consumers, Google has continued to refine and facilitate the use of its search product, offering specific results in such categories as video, news, images, and news archives.


The result? Consumers connect to the information they care about. And marketers connect to the customers they care about. The intersection of the two is the sweet spot for marketers.”


For those CPG marketers who believe that search results in a low involvement in their product categories, Kells had a sharp response: “I couldn’t disagree more … low-involvement brand marketers” are the only reasons for low-utility search results. And, again, he used facts and figures to support his cause:


  • “Recipes” yield 20 million searches a month, or more than 200 million a year. “Customers are saying, ‘I care about the very thing you might be selling.”
  • “Chocolates” drives 10 million searches a month, and “consumers are raising their hands and saying ‘I want to learn more.”
  • “Body odor” and “sweating” generate more than 2 million searches in a category that is not quite as engaging as recipes and chocolates.

A CPG company that doesn’t maximize the search return on all its marketable assets with the maximum relevant search distribution (ads and free links) is not taking advantage of its full network of consumer connections. And those links can happen in unlikely ways. Kells related the story of Heinz using the Internet and YouTube for a “Top This TV Challenge.’  

Heinz's Top This TV Challenge helped connect the brand with consumers

The idea was for consumers to create their own spots for ketchup in a digitally-driven contest. Some 1.18 million Internet users responded to the initial online call-to-action video. And, the marketing circle started with that challenge and moved to search ads, a microsite dedicated to the spots, product marketing (including label), the winning ad airing on the Emmys and display ads announcing a second challenge.

According to Kells' quantification of results:


  • There were 4,000 qualified video entries (8,000 actually entered, but half didn’t make the first-round cut-off).
  • Consumers spent 80,000 hours watching the videos on YouTube.
  • Consumers spent a total of 105,000 additional hours interacting with the brand.

Actually, the most important reaction was more qualitative: “We’ve already seen a pick-up in the [ketchup] business,” the Heinz North America president/ceo told the Pittsburg Post Gazette.


The Effectiveness Decade: The third of Kells' Business 3.0 drivers finds its support in the positive response to two key questions: Can online campaigns drive brick-and-mortar sales? Can search help build my brand? If I have a great video asset, should I put it online as well as TV?


Kells said that some major marketers need to tape a “step of faith” to overcome the uneasy feeling that a misstep on the Internet might make them look foolish. Can digital advertising drive in-store purchases?


Ask Dove, which used flash banners, small text search ads (“Try Dove Deodorant”, “Dove Deodorant Sample”), click-to-play online videos, and flash banners on travel, food, health and family sites to produce a 25% increase in in-store dollars spent on the brand and a powerful ROI: for every $1.00 invested with Google on the program, consumers spent an incremental $1.43.


For question number two - “Can search help build my brand?” - Kells pointed to a test for a food/snack brand that appeared in paid search positions. Unaided awareness increased 14 points; purchase consideration jumped 11 points. And, every bit as importantly, those same measures dropped 5 points in each category for the product’s competitive brands.


Notes from the panel 'eCommerce Marketing: The Art of Customer Acquisition and Building Lifetime Value'


Panelists: Teresa Caro, strategy director, Avenue A I Razorfish: Joe Megibow, vp/customer experience and online marketing,; Stephen Governale, senior vp/director, SMG Search; Julie Clark, senior account executive, Revenue Science; Araceli Arroyo, avp/eBusiness, HSBC Card and Retail Services.


Thesis: Online retail sales in the US will increase 20% this year to $150bn this year. Most growth will be driven by veteran online buyers. New technologies and services offer retailers the chance to “engage” with prospects and customers in new and ways. Tools such as online video, blogs, podcasts, social networks, search, and word-of-mouth help connect with consumers.


The panel's observations included:


Behavioral targeting best practice: targeting type


  • Know what types of targeting are available from your Behavioral Targeting vendors and the differences between them: Behavioral Segmentation, Site Re-targeting, Search Re-Targeting, Creative Retargeting

Behavioral targeting best practice: data sources


  • Know who/what is populating your segments!
  • What type of data is utilized: publisher, search engines, other
  • Is the data paid for? What are the vendor’s rights to the data? Are the providers aware?

Behavioral Targeting best practices: privacy


  • Do your Behavioral Targeting vendors have a policy that describes how they use behavioral data?
  • Do they adhere to NAI, TrustE, IAB and BTSC?
  • Do users have a way to opt out of being tracked?

Where to begin?


  • Know your product and know your customers
  • Target multiple personas/customer profiles
  • Create a customer lifecycle and market to them accordingly
  • Now that you know your customers get to know where to find them

Connecting with your customers


  • Reach out and acquire new and existing customers in a way that makes sense
  • Be genuine, your customers can see right through your intentions
  • Market your brand in a positive way
  • Make sure you have the time and resources to commit to your marketing initiatives
  • Measure and adjust

Getting customers to your site


  • Users and search engines crave fresh relevant content
  • SEO/SEM: Search is big. Maximize your exposure within search engines where your customers are looking for information
  • Beyond Search: Increase the doorways to your site via multiple marketing channels: RSS, blogs, microblogs, social networks, podcast, the options keep increasing

Maintaining customer loyalty

  • Listen to your customers needs
  • Customer usability should be at the top of the list
  • Customer service should service via all contact channels
  • Remind your customers of the benefits you offer
  • Be creative with incentives
  • The customer may not always be right but they sure have control

Is search getting too much credit?


  • In some cases, search will convert an acquisition without influencing the purchase decision. In this scenario, search is meets/facilitates pre-existing demand and commonly transacts through brand or brand URL searches
  • Conversely, search also has proven ability to influence purchase decisions, and drive brand demand. Demand generation occurs through search advertising against upper funnel searches (“hybrid vehicles”, “mortgage lenders”)
  • The vast majority of ad serving and measurement platforms are flawed in that they attribute a conversion to the last exposure event in a string
  • Because search is often the “go-to” tool for consumers in the final stage of a purchase process, a disproportionate number of conversions are attributed to search
  • This is evidenced by conversion rates for brand terms (often double-digits), relative to the conversion rates for category level terms (sometimes less than 1%)
  • While search’s ability to convert brand believers shouldn’t be discounted, it’s important to understand what purchases search marketing is influence or facilitating  

Search: good news


  • Many leaders in the space have begun to improve accuracy and drive understanding of multi-exposure attribution. Today, both Atlas and Google/DoubleClick have ability to record multiple exposures in a string
  • Yahoo! also provides “assist” data against search queries to identify terms that may have been previously clicked before the final click that resulted in a conversion. This data sheds light on how upper funnel searches can influence conversion through subsequent (brand) searches
  • Many search bid/campaign management tools also track and allow for optimization within multi-query search strings  

Bad news


  • Multiple exposure reports are often not synced with general performance reports and must be looked at separately
  • The data offered by multiple-exposure event reports show which events happened before the conversion-driving event, but do not statistically determine influence. Influence is still largely left to open interpretation
  • In most cases reports limit number of traceable events based on cookie windows or a finite number of events, which may prohibit accurate attribution - particularly for high-consideration purchases
  • Ability to optimize based on accurate attribution requires simpler outputs that assign weights to specific keywords and display ads. Search is able to assign attribution weights based on recency, keywords, ad groups and randomly No system assigns attribution weights to both display and search in an automated fashion  

Next generation attribution will:


  • Identify real statistical correlations across all exposures
  • Weight attribution based on historical performance
  • Learn dynamically, and adjust inter-ad influence on an evolving data set
This paper is an edited version of Geoffrey Precourt's coverage of the Ad:Tech Chicago 2008 Conference. All his reports direct from the event can be found at