DOSSIER 7.1 | 1
MAKING PEOPLE-BASED SEARCH ADVERTISING A REALITY
THE COMPLEX WORLD OF ENTERPRISE SEO BRANDING IN CHINA’S DIGITAL GOLDEN AGE
MAKING THE MOST OF YOUTUBE ADS
SUPPORTING MAJOR AD CAMPAIGNS WITH MULTI-CHANNEL DIGITAL MARKETING FIVE COMMON MISTAKES THAT MAY (STILL) BE COSTING YOU IN PAID SEARCH
PETER RANDAZZO SVP and Chief Technology Officer • Merkle Peter is the chief technologist leading the product and engineering efforts that power Merkle’s industry-leading marketing platforms and cutting-edge digital solutions. He specializes in leveraging big data technologies to achieve business goals in digital marketing, advertising, and publishing.
ADAM AUDETTE SVP, Organic Search • Merkle Adam is a recognized thought leader in the SEO industry, having worked with dozens of leading brands on technical SEO, content strategy and marketing, including: Google, Walmart, Nordstrom, Zappos, Experian, Walgreens, and Office Depot. He has been speaking and writing about search marketing for over a decade.
DALTON DORNÉ VP, Marketing • Merkle Dalton has a rich background in marketing and lead-generation, and is focused on driving growth for Merkle. She first joined as CMO of Merkle | RKG, and was previously CMO at Havas PR. Dalton spent a decade in China working in marketing and business development roles with local and international brands.
ANDY TAYLOR Senior Research Analyst • Merkle Andy is responsible for analyzing trends across the digital marketing spectrum for best practices and industry commentary. He is a major contributor to Merkle’s thought leadership content including blogs, Dossier, and the quarterly Merkle Digital Marketing Report.
LAURA SCOTT Retail Marketing Lead • Merkle Laura promotes Merkle’s Agency Services and Retail practices by supporting the strategy and development of marketing content and events. She formerly drove strategy for prospective customers, specializing in SEO.
RYAN GIBSON EVP, Marketing Strategy • Merkle With over a decade of experience in digital and SEM, Ryan is charged with advancing Merkle’s SEM solution to keep it at the forefront of the industry, and delivering optimal results for enterprise advertisers. Ryan has lent his expertise on the changing landscape of online marketing to publishers such as eMarketer, Shop. org and USA Today.
Making People-Based Search Advertising a Reality Peter Randazzo
SVP and Chief Technology Officer • Merkle
The Complex World of Enterprise SEO The Dynamics Shaping the Roles of Agencies and SEO Itself Adam Audette
SVP, Organic Search • Merkle
Branding in China’s Digital Golden Age
What Marketers Need to Know About Brand-Building on Baidu Dalton Dorné
VP, Marketing • Merkle
Making the Most of YouTube Ads
How New Capabilities and Smart Strategies Are Making It Easier to Succeed with Video Ads Andy Taylor
Senior Research Analyst • Merkle
Supporting Major Ad Campaigns with MultiChannel Digital Marketing Uncovering the Scoring System Behind Merkle’s Digital Bowl Laura Scott
Retail Marketing Lead • Merkle
Five Common Mistakes That May (Still) Be Costing You in Paid Search Ryan Gibson
EVP, Marketing Strategy • Merkle
2 | DOSSIER 7.1
MAKING PEOPLE-BASED SEARCH ADVERTISING A REALITY
PETER RANDAZZO The advertising landscape has been long split into the realms of digital and offline. While much technology, capital, and talk have been aimed at bridging this gap, the divisions are deeply entrenched – both at brands, where budgets are often similarly split, and at the agencies aligned to serve them. Many brands put millions of dollars and tons of effort into building up and maintaining a robust customer database around which their real-world marketing efforts revolve, only to find the power of that database marginalized in the increasingly competitive and murky waters of digital advertising. Our commitment to deeply understanding and managing relationships with customers had taken a back seat to optimizing distinct metrics like views, conversions, impressions, keywords, clicks, and shares. Because of the anonymity of cookies and IP addresses, digital media planning, unlike its offline counterpart, became divorced from a
thorough understanding of the individuals it was intended to influence. Marketing that should ultimately have been about people became about behaviors and devices. Merkle stood in an interesting place in this divide. Our $1.5 billion of search and display spend told us that excellent performance in digital produced results for our clients. And our decades of database marketing experience told us nothing produces success like a true measure of the customer. We knew that brands would demand a return to one-to-one, or people-based, marketing. The real question was, “When?” As it turned out, the answer to that question is, “Now.”
DOSSIER 7.1 | 3 BRIDGING THE DIGITAL DIVIDE In the last few years, the ecosystem has started to inch toward bridging this digital divide. Ad tech vendors and their data partners have gotten better at measuring digital activity through proxies for real people, such as cookies and device IDs. More recently, the larger platforms like Facebook, Amazon, and Google have leveraged their enormous scale to establish custom addressable audience pipelines, allowing brands and their agents to directly (and safely) upload personally identifiable customer records and determinatively market to customers across devices. While limited in scope to the borders of their walled gardens, these custom audiences are nonetheless an important step toward bringing digital marketing back within focus for holistic planning efforts. Taken together, the digital dots were finally beginning to connect to each other. All that remained was for those dots to be connected to the people they represented. To make people-based marketing a reality, we launched MerkleONE last year
We knew that brands would demand a return to peoplebased marketing. The real question was, “When?” As it turned out, the answer to that question is, “Now.”
as a single solution that combines our strengths in database marketing and digital performance marketing and measurement. It uses these newly connected dots to allow audience exploration, media planning, execution, and performance optimization to be born of real data about real people in the real world. It allows brands and publishers to properly leverage and hone the truths within their database marketing ecosystems everywhere. What was old is new again. People-based marketing is coming back to the marketing strategy mix, as it should, but it is not completely back yet.
TARGETING PEOPLE IN SEARCH The elephant in the room is search. While display is beginning to surpass search in spending, search is still a digital marketing dynamo with about four of every nine digital ad dollars expected to be spent in the medium for the foreseeable future. In the US, that amounts to just under $30 billion in 2016, according to eMarketer. Search is, therefore, the last enormous pool of digital spend that is still largely peripheral to people-based marketing efforts. Jumping on this bandwagon with great and deserved fanfare, Google rolled out its Customer Match program late last year. Through AdWords and its API, customer records can now be uploaded onto Google to create audiences. Building upon its success with the launch of remarketing lists for search ads (RLSA) in 2013, search campaigns for Customer Match audiences can be managed with custom copy and bidding strategies. Our search practitioners can match real customer records to their search campaigns and create custom audiences, thereby injecting audience into search. It’s a big deal. Now, we might casually say that targeting known audience members, a staple of people-based marketing, has finally come
4 | DOSSIER 7.1
Projected US Digital Ad Spending by Format 2014 - 2019, billions | Data source: eMarketer.com, September 2015 $46.69B
*Other includes ad spend in classifieds, directories, lead generation, email, and mobile messaging.
to Google. Those of us who have been search practitioners for a while might cringe at that statement. While it is a great step forward, we recognize that “targeting” is not quite the right word because this is not how search works. Search is a problem for the people-based marketing paradigm because, by its nature, it is not a medium that directly targets the individual. It is an ecosystem built around waiting for consumers to exhibit intent in the form of keywords. In other words, it is a medium that is built around passively targeting behaviors rather than actively targeting people.
A SEARCH DSP? Now that audience has come to search, what we need is a search demand-side platform (DSP). Now, that is not meant literally, but there is something to be said for how broad the accessibility of programmatic inventory has become since the ascent of the DSP. And there can be little doubt that the paradigm it shapes lends itself to targeting people.
Say what you will about the efficacy of DSPs, but they are very popular because they make the practice of buying programmatic media for individual targeting possible for even the greenest of practitioners. They make it relatively easy to create new audiences; use proprietary data to build look-alikes for prospecting; set varying levels of campaign constraints; get immediate results; and ultimately test and re-test toward being ever more sharply focused on performance. DSPs succeed because they make the complex world of programmatic media auctions easy to navigate. The complexity in programmatic stems not from the complications of bidded media marketplaces, but rather from the ecosystem itself with its dizzying array of options and opacities. In contrast, the dynamics of display marketplaces are simple in that the primary question participants need answer over and over again is simply how much they want to bid for a particular impression. This makes them ideal candidates for participating in a people-based marketing solution.
DOSSIER 7.1 | 5 A DIFFERENT PROBLEM As the audience spigot is untightened in search, the expectation may be that the Wild West of audience targeting will come soon after, with each Google account immediately having hundreds of audiences all running simultaneously, constantly tailoring performance to microsegments. People-based search will not immediately take off like that though. Search does not suffer from the same type of complexity problem that DSPs solved for display, but it does have its own complications that need to be simplified in order for people-based search to fly. Unlike the wide world of display, the search ecosystem is narrow and largely dominated by one giant. In that sense, the ecosystem is simple to interact with. Complexity in search is more of an issue within the practice of search management itself and how the introduction of audience impacts it. Targeting an audience within search means managing a permutation of a search campaign (sometimes millions of keywords) just for that specific audience. In other words, each new, robustly targeted audience can require a full copy of the original campaign, but with its own management demands. Dimensions like copy, headlines, bids, devices, and landing pages must be maintained for every keyword. In comparison, the audience management burden in display is so low it is not worth mentioning, but it can be quite high in people-based search. It is a human limit. At what point is the work too burdensome for the return? The good news is that the paradigm for simplifying the management of audiencebased search campaigns is a well-known one, especially for the technically inclined. Trying to scale people-based search
management poses a change management problem that requires a measure of revision control to solve that should sound very familiar to software developers. As stated, each audience can be mapped to a permutation, or branch, of the master, or source, campaign. Each branch is a campaign in its own right, with varying differences, or deltas, to the source that are unique for each branch. When making a change to the source, business rules need to prescribe which changes should cascade to branches and which should not. Similarly, some deltas at the branch level should be merged back to the source and some should not. The italicized and bolded terms in the last paragraph are staples of a particular kind of revision control, called source control, used around the world to organize the work of software development teams. It is necessary to have a type of source control for search campaigns, to allow practitioners to focus on the important differences between the audience-driven permutations of their campaigns, and trust the system to manage the rest. Without it, search can only dabble with people-based marketing, as audience-based search will not scale.
Complexity in search is more of an issue within the practice of search management itself and how the introduction of audience impacts it.
6 | DOSSIER 7.1 THE FUTURE Now that the barriers to people-based search are clear, the question becomes, “Who will solve them?” Customer Match, as it stands now, is an enormous leap forward, but it does not yet offer a complete people-based solution. It will likely take some time for features like lookalike prospecting, and the optimization and experimentation it powers to be a part of Google’s Customer Match solution for search. These are barriers that Google can and will solve. If only that were enough. Solving the complexity of audienceinfused search through a type of source control is another matter. It is unlikely that AdWords will come to the rescue soon, as scale is only an immediate problem for those with large, complex campaigns. Google’s AdWords API is forced to maintain a lowest common denominator posture as it is meant to serve both as a foundation for a more sophisticated media management software, as well as for the basic and free AdWords user interface.
Because of this reality, source control will most likely be brought to search by master practitioners and the technologies they create and employ. Our vision is to help bring search out of its silo and make it a pillar of peoplebased marketing. We are connecting the dots between our clients’ search intent and database marketing, enabling our MerkleONE platform to maintain a reciprocal relationship between customer understanding and intent behavior. The roadmap for our Merkle Search Platform is particularly focused on enabling the kind of easy audience experimentation that other channels have enjoyed for years. People-based marketing is coming to search at scale. Planning for real people is the future. Just wait and see. Peter Randazzo is the chief technologist leading the product and engineering efforts that power Merkle’s industry-leading marketing platforms and cuttingedge digital solutions. He specializes in leveraging big data technologies to achieve business goals in digital marketing, advertising, and publishing.
M1: Get People-Based with ONE Solution MerkleONE is the ONE integrated solution for people-based marketing. Capitalize on the shift from unknown to known audiences, panel-based to people-based marketing to reach actual people with advertising and content that’s relevant to them as individuals. Visit our site to read more about how the MerkleONE solution enables peoplebased digital marketing at scale across display, social, search and content.
DOSSIER 7.1 | 7
The Dynamics Shaping the Roles of Agencies and SEO Itself ADAM AUDETTE
What is the role of SEO in the marketing programs of enterprise organizations? I’m going to attempt to answer that question by first positing that SEO as a marketing discipline is (still): poorly understood; wildly variable in its applications and quality; and requiring of diverse skills and specialties. It’s a complicated field and one that continues to evolve. Indeed, there are several major dynamics in 2016 that are important to understand as we define the role of SEO within enterprise organizations.
8 | DOSSIER 7.1 THE DYNAMICS AT PLAY IN THE SEO INDUSTRY 1. There exist rapidly expanding skill requirements for SEO professionals, as the complexity of the field grows. Owing to its dynamic nature, SEOs must possess technical and analytical skills, user experience and content strategy expertise, and excel at writing and communication. 2. There exist two primary fields of knowledge that SEOs need to develop expertise within: technology and content strategy. For example, knowledge of how mobile applications are indexed and ranked by Google, and how that may change searcher behavior and user needs, represent intrinsically connected, yet vastly diverse areas of knowledge. 3. There still exists a myriad of bad players. The idea of link building, for example, is dated and focuses on search engines rather than users. And while its effectiveness has diminished, less sophisticated SEOs still focus there. This mix in the marketplace creates a lack of clarity and causes fear by exposing a website to penalty risk unnecessarily. 4. There is increasing belief in the commoditization of search, and SEO is no exception. Enterprise companies know they need to invest in SEO, and sometimes simply “tick the box” without intentionally choosing a partner that can provide the highest quality and sustainable strategies. This turns decisions about choosing SEO partners into a shopping exercise. Given these challenges, it is essential that today’s marketing departments invest in building SEO from the ground up, but because of the wide range of skills needed in today’s field of SEO, companies are in a war for talent. Those that can land the best minds (and keep them) will win by enabling their companies to build smarter SEO practices within the organization. This doesn’t mean there isn’t a place for search agencies in the future. On the contrary, the best internal brand teams leverage agencies more effectively by understanding how to engage them to maximize impact. This, more often than not, requires partnership skills that less experienced SEOs lack.
THE ROLE OF SEO IN ENTERPRISE ORGANIZATIONS Defining the role of SEO in an enterprise organization is a challenge, inasmuch as there isn’t a single answer that applies to all. However, we can paint some broad brush strokes to it. There are several primary capabilities of SEO, that when properly employed by an organization, will provide maximum impact on the business. Cross-channel Integration: In the enterprise, the best SEO strategies work in concert with other channels. Paid search and organic search integration are a given, but equally important is having SEO teams engaged alongside brand promotions, public communications, social media, display and email campaigns. Search is core to discovery, and aligning marketing priorities ensures SEO recommendations are integrated in the bigger picture. Platform Expertise: Search behavior exists across myriad platforms and devices. Desktop search is only one aspect of the full picture. Mobile apps and platforms, voice search on Amazon Echo, Apple’s Siri, Microsoft’s Cortana and Google Now all represent individual
DOSSIER 7.1 | 9 opportunities to optimize for discovery regardless of device type. This requires depth in technical knowledge across diverse platforms. Audience Insight: The golden chalice for marketers is executing on the promise of peoplebased marketing. The first step in SEO is segmenting audiences and aligning searcher behavior to the specific persona types a business is targeting. Content strategy is best informed by the unique audience a business hopes to target, and best implemented by understanding their motivations and creating content assets that are timely and relevant. Competitive Assessment: SEO has always been a zero sum game, and this makes competitive assessment and insight critical. Because of its function close to the middle and bottom of the funnel where intent is strongest, SEO is well suited to provide insight into a company’s customers and its competitors. This insight can uncover where gaps in the marketplace are occurring and inform strategies for a business to address them against its competitive set.
THE ROLE OF AGENCIES The role of agencies in this ecosystem is critical. In-house teams where resources are always strapped, can’t do it alone. At the same time, the best agency partners work to institutionalize SEO into the fabric of a company, creating centers of excellence and helping to build internal departments and teams. This institutionalization of SEO within enterprise marketing teams will be a growing trend (it’s already happening), as will the need to use leading agencies to help on key initiatives and to stand up integrated global SEO programs. Among many possible examples, this includes: •
Site migrations, such as moving to a new content management system, updating URLs, and migrating to ccTLDs.
Consumer journey work to help create the best possible match between a relevant result for a search, and the content that brings users onto the site and leads to conversions.
Analytics and tagging audits and strategies to ensure data is clear, precise, and accurate.
Mobile strategies including responsive web design, app indexing, and deep linking.
Global initiatives that require integrating centers of excellence with centralized best practices and brand guidelines.
It’s easy to say and hard to do, but the role of SEO in the enterprise is multifarious and complicated. However, the opportunity is clear. Programs that will succeed will focus on retaining the best people, investing in technology expertise and audience understanding, using best-of-breed agency partners, and integrating their teams and programs. Adam Audette is a recognized thought leader in the SEO industry, having worked with dozens of leading brands on technical SEO, content strategy and marketing, including: Google, Walmart, Nordstrom, Zappos, Experian, Walgreens, Office Depot, Comcast, Daily Mail, Symantec, Edmunds, Amazon, Eloqua, Intuit, National Geographic. He has been speaking and writing about search marketing for over a decade.
10 | DOSSIER 7.1
What Marketers Need to Know About Brand-Building on Baidu DALTON DORNÉ
Stats on the China market are staggering enough to get even the most cautious marketers excited – the world’s second largest economy according to the World Bank; largest mobile market (Statista); home to the fourth largest digital platform in the world, Baidu. Our last issue of Dossier talked about the groundbreaking partnership between Baidu & Merkle, when we announced Merkle is Baidu’s first US-headquartered reseller. Our partnership offers exceptional performance in one of the world’s largest but most challenging markets. In this issue, we dive a little deeper into the importance of brandbuilding in China, and key opportunities for brands across the Baidu portfolio.
DOSSIER 7.1 | 11 It’s no secret that Chinese consumers are among the world’s most brand-conscious buyers. In fact, they’re willing to pay more for branded products compared to their pricedriven peers in other markets. According to McKinsey, 45% of Chinese consumers believe a higher price point equates to better quality, compared to just 16% among US consumers.1 A recent study found 49% of young generation Chinese will advocate for brands personally and online, compared to only 34% among their peers in the US.2 This trend is also reflected in Chinese searcher behavior. Chinese consumers are more likely to use search to research and validate brands, while US searchers are more likely to search non-brand products and look for price comparisons. As such, brand value is extremely important in China. Sounds like a brand marketer’s dream. But why should this matter to digital marketers? It matters because China is the world’s largest mobile market. Its rising middle class is extremely comfortable purchasing across digital devices, not to mention nearly 90% of Chinese consumers use Baidu for search. Simply put, the stakes are high for brands in China’s hyper-competitive digital ecosystem. B.A.T., or the ‘Big Three’ digital players in China (Baidu, Alibaba, and Tencent) are known to dominate the landscape in search, e-commerce, and social media, respectively. While e-commerce is big business in China, opportunities for brand-building remain limited. Increasingly popular Alibaba competitor, JD.com, offers little to no brand
control, as the experience is like being in a virtual department store. While you can have some brand control in the form of a brand store on Alibaba’s Tmall.com, maximizing the robust brand opportunities on Baidu is attractive to advertisers and allows the most brand control in China’s digital landscape. Considering the critical importance of brand building in China, this makes search, and in particular Baidu, a key part of a brand’s China marketing strategy.
BUILDING A BRAND ON BAIDU A key opportunity for companies marketing on Baidu is to leverage Baidu’s Brand Zone. Brand Zone allows advertisers to own a sizable amount of the search engine results page (SERP) with brand content and information, not only ensuring top placement, but essentially owning all the space above the fold. The result can be akin to a SERP takeover experience, which is completely unique to Baidu and not available on other search engines. It might sound obvious, but Brand Zone works for brand searches only, and must be exact match. Baidu’s philosophy is that for exact brand searches, the consumer is specifically looking for a particular brand and therefore enjoys the rich brand experience that Brand Zone delivers. The results back up this belief. Engagement with Brand Zone is high, with an average click-through rate of over 50% compared to 25% for brand searches on Google.3 There are a variety of formats and options within Brand Zone, and the experience is very customizable based on content, assets
McKinsey Research, China’s New Pragmatic Consumers, 2010
Boston Consulting Group Study, The New China Playbook, 2015
Merkle numbers from Q4 2015
12 | DOSSIER 7.1 and consumer value-adds that the brand can provide. BMW does a great job of leveraging multiple ad formats within Brand Zone, including images, videos, and the ability to scroll through a carousel of the different BMW series. Beyond including images, music, video, and social media tie-ins, some savvy luxury brands have even used Brand Zone to live-stream their fashion shows from Paris and New York – effectively taking their brand experience into one of the world’s most voracious luxury markets via search.
In addition to Brand Zone, advertisers should ensure that they are taking steps to protect their brand keywords. Designing and implementing a brand protection strategy is one of the first things Merkle does when onboarding new clients and is a form of trademarking your brand on Baidu. Now look at this brand search for Topshop. Without brand protection, Topshop is not able
Unsurprisingly, Brand Zone comes at a price
and is best for enterprise brands looking to ensure a rich brand experience. The thing for advertisers to remember is that Brand Zone is the only way to ensure top-of-page placement for your brand.
DOSSIER 7.1 | 13
to secure the top placement on the Baidu SERP, with competitors like GAP and Shopbop appearing above them using their brand name. Compare this to when a company has brand protection and is also employing Baidu’s Brand Zone, like the BMW example, resulting in a very different experience.
INCREASING DISCOVERABILITY & CAPTURING NON-BRAND CLICKS Another opportunity for marketers is Baidu’s Mini Zone. Unlike Brand Zone, Mini Zone is triggered by non-brand search. Mini Zone lives on the right rail and can bump out. Advertisers can buy keyword packs from Baidu for up to a year for Mini Zone. The example above shows some savvy Mini Zone work by Chanel for the term “lipstick.” You can see the bump out trumps the paid and organic results for lipstick and successfully grabs the searcher’s attention.
BEYOND THE SERP While Baidu is China’s largest digital platform and the world’s fourth most trafficked site, there’s much more to Baidu than the search engine alone. Baidu has numerous successful and important properties, including Baidu
Knows (community forums), Baidu Baike (wiki-like encyclopedia), Baidu Music, Baidu Map, Baidu Cloud, Baidu Space (social network), and more. Baidu’s Brand Zone Matrix is yet another way to maximize visibility across Baidu properties. Why stop at the SERP? Brand Zone Matrix offers advertisers “6 plus 1” – which is the ability to have Brand Zone on up to six Baidu platforms, plus be mobile-optimized across them all. The consistent visibility and brand experience across multiple platforms ensures strong brand performance on Baidu, at massive scale. In a fast-growing brand-conscious market like China, smart marketers will maximize all high-performing opportunities to capture market share and drive performance. Merkle makes this easy for our clients. Visit merkleinc.com/Baidu to learn more about how we can help with your China marketing performance. Dalton Dorné has a rich background in marketing and lead-generation, and is focused on driving growth for Merkle. She first joined as CMO of Merkle | RKG, and was previously CMO at Havas PR. Dalton spent a decade in China working in marketing and business development roles with local and international brands.
14 | DOSSIER 7.1
How New Capabilities and Smart Strategies Are Making It Easier to Succeed with Video Ads
YouTube has long been the world’s most used video sharing website, and its more than a billion monthly users constitute a huge share of worldwide internet users. The YouTube audience is vast and diverse, ranging from music fans streaming song videos to DIYers looking up how-to videos. As such, brands have the opportunity to get in front of many existing and potential customers through the platform. However, such a wide net requires deep knowledge of how the different types of targeting and ad formats available can be used to optimize YouTube campaigns for their intended audience. Fortunately, advertisers now have more options and targeting tools at their disposal than ever before to do just that, some only recently made available.
TrueView Gives Advertisers Confidence Their Ads Aren’t Totally Ignored Launched in 2010, YouTube TrueView instream ads allow advertisers to show users
video ads before or during other videos. These ads can be skipped by viewers after they’ve played for at least five seconds. Since users are typically most interested in the video they’ve selected, it’s fitting for advertisers to worry about whether spend will go to ads that might just get ignored. However, TrueView in-stream pricing works such that advertisers are only charged if viewers watch the duration of an ad if it is less than 30 seconds long, at least 30 seconds of the ad if it’s longer than 30 seconds, or engage with some element of the in-stream creative. Engagements which qualify as chargeable interactions include clicks to the advertiser’s website and clicks on call-to-action overlays, cards, and companion banners. Google changed the click behavior of in-stream ads in April 2015 to reduce unintentional clicks. Formerly, a click anywhere on the YouTube player would count. This gives advertisers assurance that the views they’re paying for meet some reasonable standards to ensure they aren’t just being skipped over.
DOSSIER 7.1 | 15
YouTube in-stream ads have also been bolstered with many additional features over the past few years which advertisers can use to make the most of their video ad investment.
Shopping Ads Enhance TrueView Video Ads Google Product Listing Ads (PLAs) have grown markedly over the past few years, with PLA share of retailers’ total Google search ad clicks reaching 38% by the end of 2015. Google’s success with this ad format and the strong performance it has driven for advertisers has even led competitor engines Bing and Yahoo to develop their own product ads. In November 2015, Google sought to expand the reach of these ads by rolling out TrueView for Shopping to all advertisers. This format allows advertisers with instream video ads to show product ads along the right side of videos while they play, complete with pricing and product
title information, and clicks take customers directly to product pages on the advertiser’s website. In order to take advantage of this format, brands must have a feed loaded to the Google Merchant Center. During the campaign creation process, advertisers can narrow down which products will be allowed to show using custom filters to feature the most relevant products possible in these ads. While many YouTube campaigns focus on branding rather than direct response, these ads provide users with the ability to quickly navigate to products that might be featured in or related to themes in video ads. Of course, most users aren’t looking to click out of YouTube when they’re waiting for a video to play, but this format gives users the ability to immediately find out pricing and title information about the products relevant to video ads even without clicking. Advertisers can also improve the efficacy of
16 | DOSSIER 7.1
Shopping Ads for YouTube allow advertisers to feature product ads alongside video ads with pricing and product title information. REMARKETING LISTS
this format by layering on audience lists of users known to be familiar with the brand.
Remarketing Lists for YouTube Ads Google remarketing lists date back several years now and are built based on users who have recently visited an advertiser’s website. Using the Google Analytics or AdWords remarketing tracking tags, brands can create custom lists based on the pages users visited and the actions they took on-site. These lists can be layered onto YouTube ad campaigns in order to reach only users who have recently visited specific pages or taken actions, such as signing up for newsletters or converting. Advertisers can target these users with content specific to what they’re most likely to be interested in. These lists can also be used along with TrueView for Shopping in order to filter products for these audiences with knowledge of what they might want to purchase. For advertisers that don’t want to manually select the products shown to users in remarketing lists, Google’s dynamic remarketing setting will automatically show users products which they’ve viewed on the brand’s site. Aside from targeting users who have recently visited a brand’s website, advertisers are also now able to target users for whom they’ve obtained email addresses firsthand.
YouTube advertisers can use remarketing lists to target in-stream video ads to users who have recently visited or interacted with the brand’s website. CUSTOMER MATCH Google’s Customer Match targeting gives advertisers the ability to create targetable audiences from email lists obtained firsthand.
Customer Match for YouTube Google’s Customer Match product was released in September 2015 and allows advertisers to build user lists out of email lists in order to target those users in Google search, Gmail, and YouTube. As such, advertisers now have the ability to bake deep customer insights into their targeting, such as buying patterns and interests. Customer Match lists must contain at least 1,000 identifiable users in order to be used in targeting, a minimum aimed at protecting consumers’ privacy. This limits how specific an advertiser can get with their targeting, but advertisers can still get creative in deciding which video content to show to which users on their lists. Advertisers can also apply different performance goals and budgets based on this audience information.
DOSSIER 7.1 | 17 Further, because brands have these users’ email addresses, it’s easier to attribute YouTube ad view impact on those users’ propensity to convert, since orders can be linked back to the list regardless of whether these users actually click on the YouTube ad(s) they view. Another potential opportunity for brands to explore with Customer Match is using similar audience lists created by Google whenever a list of at least 5,000 emails is uploaded. These audiences are comprised of users who display similar search interests to the users included in the email lists uploaded, and can help brands to expand the reach of audience targeting and bring in new-to-file customers.
The World is a Big Place – Target Regions Strategically Ad formats and audiences can help to supercharge YouTube campaigns. A more fundamental setting can be hugely important as well: geographic targeting. Approximately 80% of all YouTube views occur outside of the United States. As such, ads have the potential to reach viewers across the world who are looking at a huge variety of content. While this provides a large-scale global opportunity, brands must be intelligent in their targeting to reach those users who stand to be the most impacted by their ads. For example: •
A local Charlottesville, VA musician may only want to feature pre-roll YouTube ads for viewers in the Charlottesville area.
A retail brand that ships only to the United States and Canada probably wouldn’t want ads to appear in countries outside of those two.
A national retailer running promotions at its brick-and-mortar stores might only want to target users viewing videos in close proximity to their store locations.
Fortunately for advertisers, all location targeting options that are available in AdWords for search campaigns can also be used for YouTube ads. This means campaigns can be targeted to a number of different types of locations, including countries, regions, cities, zip codes, and radii around businesses or other landmarks. Further, Google is consistently getting better at pinpointing user locations to more specific location types, giving advertisers greater confidence that more granular location targets will work for most users.
Conclusion Google’s YouTube audience continues to grow, and along with it the opportunity for advertisers to reach wide audiences with video ads. The toolset advertisers have available to segment and target users is also growing, making for smarter in-stream ad investments that have a bigger impact on the audiences they reach. As such, brands are becoming increasingly willing to explore what YouTube in-stream ads can do for them, and this trend will likely only continue in the years to come.
Andy Taylor is responsible for analyzing trends across the digital marketing spectrum for best practices and industry commentary. He is a major contributor to Merkle’s thought leadership content including blogs, Dossier, and the quarterly Merkle Digital Marketing Report.
18 | DOSSIER 7.1
Uncovering the Scoring System Behind Merkle’s Digital Bowl LAURA SCOTT
DOSSIER 7.1 | 19 Companies invest big bucks in branding efforts. Celebrity-supported campaigns and aspirational ads build a brand’s image, exposing potential customers to their name. These branding plays are a great strategy for creating visibility, but in an omni-channel world, it’s not enough to run a creative campaign through a single channel. Brands must create a cohesive message across touchpoints and coordinate marketing efforts across channels to create the best customer experience and drive the most value from their branding investment. Each year, Merkle evaluates the king of all branding plays - Super Bowl ads. During the Super Bowl, a team of Merkle employees huddle in a conference room to compare social, paid search, SEO, and programmatic campaigns supporting Super Bowl ad buys. We use criteria that represent bestBrands must create in-class digital marketing practices that enable brands to a cohesive message have the highest digital visibility. Each set of criteria has across touchpoints and an objective, scaled point system. The next morning we release a report ranking brands’ digital performance and coordinate marketing we award a Digital Bowl champion. efforts across channels While this scoring system was created to judge Super Bowl advertisers specifically, it can be used to evaluate the effectiveness of any omni-channel marketing campaign. Here are the fundamentals of the criteria behind The Digital Bowl, and some essential strategies you should understand in supporting big branding efforts with digital marketing.
to create the best customer experience and drive the most value from their branding investment.
SOCIAL MEDIA Social media gives brands the opportunity to amplify their message and continue to capture the attention of prospective customers. Brands should take advantage of this by creating social content that complements big branding efforts and engages customers.
During the 2016 Super Bowl, Amazon mastered this by having brand ambassadors Alec Baldwin and Missy Elliot engage in live, ongoing social conversation.
Great brands extend the conversation on social by using calls to action and personal, authentic responses to user posts. Aside from building relationships with real customers, these interactions can help to provide brands with promotable content as well as strategic ideas for future content based on user interests.
20 | DOSSIER 7.1 Brands should also amplify their message on social by creating fresh, complementary content that is unique to the platform and engages customers in real time. This approach gives social its own distinct flavor, providing users with new content that’s relevant to the campaign, but which they won’t find anywhere else. Campaign-specific hashtags are also a must for branding efforts. Customized hashtags should be used at various touchpoints to promote and organize the conversation.
SEO A good SEO strategy often sets the stage for your other digital marketing tactics. Creating a campaign-focused landing page is the first step in optimizing for organic and gives all other channels a target location to send prospects and customers. Well-coordinated campaigns even send users to their chosen landing page directly from their main branding initiative in an attempt to continue to engage them. Curated content and structure on that landing page is a must for both SEO and customer experience. Advertisers should include campaign messaging in the structural elements of the page, like page title and headers, as well as in descriptions and the core page content. Landing pages should also be discoverable by search engines for queries relating to the ad campaign. A strong technical foundation on your site and the landing page will ensure that things like directives in the site’s robots.txt, redirects, and meta tags aren’t keeping your page out of search. Campaign content also needs to be accessible to users on any device. Mobile-friendliness is a must for good organic performance and user experience.
In this year’s Super Bowl, Doritos shined in SEO. Their campaign landing page re-used the same URL from last year to build on its link equity. With this strategy, Doritos doubled its backlinks, and increased search visibility for Super Bowl 50.
DOSSIER 7.1 | 21 Linking is another important tactic for optimizing your campaign for search. Both internal and external linking are important signals to search engines about the authority of your landing page. Natural internal links on your site pointing to your landing page send crawlers to your branded content and communicate its value to search engines. External linking communicates the authority and trustworthiness of a page. Brands must accrue backlinks naturally from reputable sites as a way to communicate the value of the content and ultimately gain visibility in organic search.
PAID SEARCH & EMAIL Advertisers can control messaging and maintain visibility in search and inboxes during campaigns with paid search and email. Successful campaigns run supporting paid search ads on the first page of search results for relevant queries like the topic of their campaign and branded searches. Since mobile devices now account for more than half of all search queries in the United States, it’s also important that ads appear on the first page of results on mobile..
Since mobile devices now account for more than half of all search queries in the United States, it’s also important that ads appear on the first page of results on mobile as well.
Paid search gives advertisers full control of messaging through ad copy on the search results page and on landing pages. Ad copy should tie in relevant campaign elements like hashtags or other branded content. Content on landing pages should present a cohesive campaign message that relates to the ad campaign. This extends the reach of the branded campaign and presents users with a coordinated message across touchpoints. Email gives brands the power to activate customers with well-timed outreach. In addition to capturing searchers, advertisers need to be reaching out to existing customers through email during big campaigns. Content needs to be viewable no matter how users are accessing it, and mobilefriendly emails are a must for driving direct response.
Taco Bell did an excellent job with coverage during Super Bowl 50 by displaying PPC ads leading up to, during and after the big game.
22 | DOSSIER 7.1
DIGITAL MEDIA/DISPLAY Successful digital marketing doesn’t just capture interest, it also generates it with strategies like prospecting with display advertising. Strong digital strategies use programmatic for both creating interest in campaigns and re-capturing visitors to targeted pages. During big ad campaigns, advertisers can purchase premium spots on top sites to ensure banner ads are front and center for all site visitors. Advertisers can also create interest by featuring teaser content. For a big TV ad, for example, this might be hosting commercials on YouTube paired with links and companion banners. Display campaigns also need to re-engage site visitors with retargeting. Brands should place remarketing pixels on their site prior to a big branding push. These should certainly be included on the campaign landing page, but should also be added to any page relevant to that campaign like the homepage, related product pages, or complementary blog posts. This allows advertisers to show curated content across the web to site visitors after they’ve interacted with that key content.
In Super Bowl 50, T-Mobile placed more retargeting pixels on its site pre-game than any other Super Bowl advertiser, allowing them to continue to show campaign-related content to visitors even after they had left the site.
CONCLUSION Big branding campaigns are a great opportunity to get your company’s name in front of prospective customers. A strategic approach to digital marketing can squeeze extra value from the effort and investment you’re putting into that campaign. Whether your company is running a Super Bowl ad or working on a slightly smaller endeavor, employ our best practices for digital marketing to create a coordinated and effective marketing strategy. Laura Scott promotes Merkle’s Agency Services and Retail practices by supporting the strategy and development of marketing content and events. She formerly drove strategy for prospective customers, specializing in SEO.
DOSSIER 7.1 | 23
RYAN GIBSON In addition to managing search campaigns for our clients, we spend a lot of time evaluating enterprise digital campaigns for non-Merkle clients, including deep evaluations of paid search programs. We see all varieties of strategies, theories and ideas on how people think search campaigns should be managed to get the best results. Some are successful, while others are…. Well, let’s just say they’re creative, but they may be missing the mark when it comes to maximizing opportunity. To help our readers avoid a similar fate, we’ll break down five common mistakes that we’ve seen across multiple campaigns so far this year and the alternative approaches that yield better results. Some of these mistakes date back as far as the paid search industry itself, while others are more recent and typically reflect missing opportunities to leverage new capabilities.
24 | DOSSIER 7.1 MISTAKE #1: Having thousands of paused keywords that are relevant to your products or services. In many paid search programs we assess, it’s clear that keywords are being paused simply because they don’t produce a certain level of data over a given timeframe, whether it’s 30 days, six months or a year. There are always going to be a good number of keywords in your account that don’t produce enough traffic or conversions on their own to yield statistically significant predictions. But so long as you’ve chosen keywords that are relevant to your products or services you sell (and we’d assume you would) there’s no such thing as a bad keyword. Just bad bids. That means that every keyword associated with your business will have some value. That value may be modest or significant, but it’s going to be rather specifically correlated to the intent that is represented by the keyword. The key is to make sure you’re aligning the keyword’s bid to its actual value. That could be $.01/click for lower performers or a $20 bid for your best terms. It all depends on the value that each keyword drives. Determining that value accurately is where most folks get hung up, typically making one of two key mistakes that result in good keywords getting paused: 1. Looking at a window of cost and conversion data that’s too short for their business or 2. Using poor methodologies for aggregating data across low traffic keywords, or failing to do so at all. Understanding the right window to evaluate click, cost and conversion data is critical. Many systems default to a rolling 30 day window. While that may seem like a reasonable timeframe, it’s actually a bit arbitrary. Advertisers should be looking at a window that at least covers the typical click-to-conversion timeframe of their customers, and they may benefit from using much longer data windows, so long as they can properly account for seasonality effects. Secondly, how advertisers aggregate data across keywords that still have low traffic even with an appropriate data window is critically important to paid search success. It really is the ‘secret sauce’ of PPC bidding and management.
As you’ve chosen keywords that are relevant to your products or services you sell, there’s no such thing as a bad keyword. Just bad bids.
Many bid tools offer one folder as a way to group keywords. Typically, these tools ‘auto-magically’ group keywords by performance, rather than through a sophisticated understanding of the keywords themselves and their associated attributes. Just because “red stiletto pumps” and “Levi’s men’s boot cut jeans” have produced a similar sales-per-click recently, doesn’t mean they have much else in common or that they will trend similarly going forward given seasonality and other effects. And this approach doesn’t offer much help for keywords with thin data.
DOSSIER 7.1 | 25 Enterprise accounts need to establish additional attributes by which keywords can be evaluated. This way, we can look at the data at a level beyond rudimentary keyword groupings and make smarter decisions. These attributes may or may not be directly present in the keyword themselves.
In the example below, we see that attributes like gender, high heels, shoes, price point and brand specificity are not intuitively understood from the keyword itself, but known from the product or category being advertised. It turns out that these are imperative to understanding the value of the intent that lies behind those keyword searches, allowing us to gain insight into the attributes that are most predictive of performance for low traffic keywords, regardless of campaign structure.
red leather stiletto pumps
levis men’s boot cut jeans
$50 - $75
Through sophisticated aggregation of these attributes, we can be more confident that we understand a keyword’s true potential. At the right bid, we can keep it running with the knowledge that it and its close cousins, will generate the results we need.
MISTAKE #2: Little to no geographic targeting in the account. Even if campaigns have geographic modifiers set up, bid variation is minimal. Many people still believe that granular geo-targeting is only for brands with a brick and mortar presence, but one of the best predictors of the value of a paid search click is where it takes place. Even if you don’t have a brick and mortar location, the location of the searcher may indicate that they’re more or less likely to represent a value to your brand and become a customer. Many advertisers don’t implement geographic bid modifiers at a specific enough level to properly account for these value differences though. We often see modifiers set at the state level, but the possibilities extend much further than that. For example, one of the key components of geo-location that helps predict value is
26 | DOSSIER 7.1 Income Bracket: % Difference from Overall Non-Brand Conversion Rate
[11 - 20%] +1% [top 10%] +15%
[31 - 40%] –6%
[ lower 50%] –25%
[21 - 30%] –4%
Data source: subset of Merkle paid search clients
[41 - 50%] –20%
household income (or HHI). Within AdWords, Google has an HHI bid modifier lever tied to the zip code of the user. The performance differences across the various HHI segments is often very large, so using this tool to better allocate ad dollars can be an easy win. Importantly, for the cases where Google can’t determine the zip code of a search, you’ll want to make sure that you still have fallback targets set up at the US or state level to ensure you’re not missing out on that traffic. (Google has gotten
very good at understanding where a searcher is located in city-level targeting, however they’re still missing some understanding of which zip a searcher might be in - understanding roughly 70% of those locations.)
MISTAKE #3: PPC is driven to a Cost-Per-Lead or Cost-PerAcquisition model when the ultimate goal is revenue or profit. You wouldn’t consider a customer who only buys from you when there’s a sale as valuable as someone who regularly purchases your premium product, would you? Then why would you allow that type of thinking to steer your paid search program? The reality is that keywords with similar conversion rates can produce vastly different value to the advertiser, whether that ultimate value is a sale, qualified lead or total profit. Driving a paid search program by a simple cost-per-lead or cost-per-acquisition target overlooks a huge part of the equation for the vast majority of advertisers. Very simply, if your goal is to maximize profit at a given return on investment, your paid search bids should be based on the profit that each keyword produces. Sales are a very good proxy for profit, but they don’t reflect the different margin rates that your products may have. Even if you’re only able to look at a CPL or CPA value in the immediate term (for instance if you have a long decision cycle after you meet a customer), you should still work to tie the ultimate value back to the keyword that drove that initial connection. For example, if you’re in the banking industry and using paid search to generate new savings account sign-ups, your initial conversion metric may be leads in the form of application completions. Longer term, though, you will want to determine how many of those applications passed muster and led to funded accounts, and at what levels, even if it’s 90 days beyond the open date. Tying this level of intelligence back into the ad program, at the keyword level, will yield smarter bids and better results.
DOSSIER 7.1 | 27 MISTAKE #4: All keywords are set on exact or phrase match only. The logic starts out with good intentions: “If we can create a keyword list that encompasses all the ways that people may be searching for our products, we should be able to put the entire program on phrase or exact match and ensure that we don’t capture any irrelevant, more broadly matched traffic. Plus, we know that more exactly-matched query traffic is going to convert at a higher rate than more broadly matched traffic.“ While it is ideal to funnel search queries to exact and phrase match keywords, to more accurately set bids for those terms, it isn’t practical for most advertisers to do this without leaving conversions on the table. Even with the engines matching exact and phrase match keywords to close query variants like minor misspellings and stemmings, more liberal match types like broad and broad match modifier still have an important role to play in most large-scale paid search programs. Similar to our earlier points around paused keywords, broad match keywords may bring in queries that don’t convert as well on average as more exactly-matched keyword phrases we’ve directly added to our account. However, as long as we set bids appropriately, they can provide incremental value, efficiently. With appropriate negative keywords in place, broad and broad match modifier keywords serve an important role in surfacing important high-traffic keywords the advertiser may have overlooked and capturing the still vital long-tail of search that is all but impossible to foresee entirely.
MISTAKE #5: All mobile bid modifiers are set to –100% or clustered at arbitrary intervals. We know that mobile traffic typically produces lower online conversion rates than traffic from other devices, but rarely do we see a case where mobile should be effectively turned off across the board. We still see that happening in some accounts. More common, though, is to see large clusters of a program being bid down by somewhat arbitrary amounts on mobile, including -100% or -75%. Just as different keywords produce a different value from one to another, different keywords will have different relative performance on mobile compared to desktop. As such, mobile bid modifiers should be actively evaluated and set at the ad group level (the most granular option in Google AdWords) based on the performance of individual keywords. If the advertiser finds that the keywords composing a given ad group have very different relative mobile performance from one another, it may be worth separating the keywords into new ad groups where mobile bid modifiers can be set more precisely.
28 | DOSSIER 7.1 Typical Mobile Bid Modifiers in Audited Accounts
More Sophisticated Approach with Each Ad Group Bid to Value
Furthermore, some types of searches are less likely to convert directly on a mobile device, but may convert later on a larger device or even offline. Advertisers need to make sure they’re including cross-device considerations into bidding, as well as potential offline lifts. Google’s cross-device conversion estimates are now available to advertisers at the keyword level, so that you can get a directional perspective on what’s converting elsewhere AND they can be more easily integrated into keyword-level bidding. We’ll typically see mobile devices get a 20-30% lift in conversions from including cross-device activity, so be sure you’re factoring these values into paid search. For brands with a brick and mortar presence, Google also offers free tools to estimate the offline impact of AdWords ads, and this lift can be even more substantial.
Closing Thoughts While paid search is now a well-established industry, it remains a dynamic one, and there is still a wide gulf between good search programs and great ones. Whether it is a failure to capitalize on the latest industry developments or a flawed approach to the more fundamental aspects of the business, we continue to see the non-client accounts we audit making costly mistakes. This is not an industry where you can succeed simply by following a strategy guided by legacy assumptions. It’s essential that you frequently evaluate and reevaluate your paid search strategies, and always let the data be the guide and proof that what you’re doing is working (or not). Ryan Gibson has over a decade of experience in digital and SEM. He is charged with advancing Merkle’s SEM solution to keep it at the forefront of the industry and delivering optimal results for enterprise advertisers. Ryan has lent his expertise on the changing landscape of online marketing to publishers such as eMarketer, Shop.org and USA Today.
China is one of the world’s fastest-growing economies and is home to the largest online population. In that market, Baidu is a giant among giants – with over 80% market share for Chinese consumers. Merkle is Baidu’s first US-headquartered reseller. We’ve been working in China for over seven years and have more than 400 staff members in offices in Shanghai and Nanjing. Visit our site to learn about Baidu’s robust offering or contact us directly to talk about the advantages of partnering with Merkle on your Baidu program.
30 | DOSSIER 7.1
Merkle is a global data-driven, technologyenabled performance marketing agency and the largest independent agency in the US for CRM, digital, and search. Contact us to talk about how Merkle can help you achieve your marketing goals: [email protected]