Pereira: Do you think the region is ready for data-driven marketing?
Parker: Yes, I think it is. And the reason why I'm pretty confident is two things:
1. Right now, there is a lot of pressure on marketing budgets. This is a bit of a wake up call: how are we spending our money, how are we getting maximum value out of that spend? Marketers talk about that all the time, but now with the economic crisis, this is a review point for many organisations, and they're re-evaluating how they go to market. We are seeing a move towards database driven marketing as a direct result of that. There's a lot more interest in digital marketing, how the database can be leveraged, so they can spend less on ATL marketing
2. I've been in the region for 10 years and what I did for eight of those years was helping organisations get their data right. I think a lot of companies now understand the value of getting good data, and really want to use it intelligently in marketing. Lots of companies have collected data on their customers; very few have actually used them. While the market is a little immature compared to the US or UK in terms of database-driven marketing, it is at the right stage now, looking for that next step.
Pereira: What kind of data do you look at?
Parker: Different products use different data. We don't use our credit data to do marketing data. That would be wrong. Where we do have credit information we need to be careful around not using that in marketing. We collect a lot of data to build our models, so often we'll collect data that is anonymised -- things like addresses which don't have a name attached to them. We collect demographic information, information about the consumer household, how many people, what's the age, income, lifestyle, how they like to be sold to, which channels? We remove the names and build models from that data. Most of the data we have in Asia Pacific is that kind of information.
We mostly work with customer-owned data, which is a goldmine of insights, and we add to that some third party data like Mosaic, an advanced consumer geo-demographic segmentation solution to get another level of insights and profiling. We do have some of our own consumer data, but more frequently we'll partner with other companies that we trust that have an opt-in data source of consumers that our organisations can market to.
Pereira: How is customer privacy safeguarded?
Parker: Unlike many other companies that are in the marketing business, we have a very large part of our business vested around the trust the companies put in us to manage their credit data.
As one of the leading credit bureaus globally, we have thousands of banks that give us their accounts details everyday. So we can't afford to make mistakes in the way that we operate on the marketing side of the business in managing privacy. We take a lot of care around making sure that we're working with trusted data providers. When we work with our customers, we mostly work with their existing customer database. If they're doing acquisition, then we're very selective about the partners that we work with.
There is an opportunity here for a much more scientific approach to marketing. And I would continue to encourage marketers to re-evaluate how they're using their data, and look for opportunities to try and leverage that, because they'll find very rewarding results from investing in their data and database-driven marketing.
Faster = better, better = faster, faster + better = the new standard
For a lot of big brands today, a well-considered web marketing campaign can take six weeks just to tune the messaging. Sample messages are sent, results are analyzed, and messages are tweaked. Then the process repeats. What emerges is pretty solid messaging -- assuming the market hasn't moved on in the meantime. More resources won't make it go faster because the serial process is the constraint.
Think of this as the static model of advertising.
The static model can't keep up with today's fast-moving trends. Fortunately, within reach is a new approach that can accomplish in days or hours what used to take weeks. Welcome to real-time advertising. Where more money can't compress the static model schedule, the real-time model inherently compresses both time and money.
Putting this level of capability into action takes more than new technology and tools; it takes new thinking regarding risk, objectives, and measurement.
The static model is carefully and methodically tuned to minimize downside risk because few brands can afford a big failure. At the same time, there is little upside in overachieving to those supporting the brand. (Not really a recipe for game-changing behavior.)
If you want to change the game, you have to change the rules. Near real-time development combined with real-time data encourages lots of short experiments rather than fewer big ones. Managing the downside risk need not be a function of cautiousness. Real-time data and control means a campaign can be terminated at any time. Brands can now do and try things they would never have dreamed off before. The economics also are changing to reward upside performance as well. The combination is going to rock the advertising world from top to bottom and everywhere in-between. But you have to play by the new rules.
The new rules are going to drive new thinking throughout the planning and execution process. This extends beyond the how to the why. With new ways to engage customers with fast-acting, interactive campaigns, everyone from the brand manager, to the creative director, to the media buyer must be re-aligned in how they work together.
Traditional metrics only tell us that we've touched someone. The new standard of engagement calls for new measures as well. New types of engagement measures allow us to differentiate between eyeballs and true engagement.
Better data ultimately drives better decisions. To leverage the new data, however, we need to re-think the overall process of how campaigns are developed and deployed.
There is also a qualitative component to good decisions. Social media expert Tara Hunt likes to point out that social capital is what makes online communities flourish, not money. Those that help you strum the social graph are very conscious that their tweet or stream of your content needs to build their social capital or they won't engage with you. Use data to understand when and how people are engaging with your content in ways that continues the conversation. If you are getting eyeballs, but not this type of engagement, it's time to shut it down and try something else.
Anatomy of a successful social media campaign
Since the goal of social media campaigns is to engage with people in ways that benefit a particular brand, traditional metrics like impressions and conversion rates fall short. If we want to encourage viral conversations, we need to find better metrics. The better you can measure what people actually do after the initial contact, the more you know about what actually works in the social graph. The better you know what actually works, the better direction you can provide to the creative director in terms of target audience, preferred media, and actionable campaigns.
With knowledge of the campaign's objectives, the next stop for the brand owner ought to be the media buyer. The data-driven media buyer is a tremendous resource that should be engaged early in the process and long before the creative is spec'd out. Since what works for a print ad is very different from what works for a social media campaign, engaging the media buyer and/or publisher early in the process lets the creative director work within a tighter focus.
Gartner echoes this but from the perspective of media companies, advising them to build core disciplines around understanding and predicting consumer trends by mining social media.
The media buyer knows what works and how much it costs. If social media is included, the media buyer can select the appropriate platform and set requirements and expectations. Platform capabilities are constantly evolving, so the media buyer is likely to loop the platform into the process. Since the media buyer is going to need metrics around the specific activities the campaign targets, these will get baked into the requirements as well. Requirements and expectations developed this way allow the creative process to serve the practical goals of the campaign.
One of the interesting developments is that a few (soon to be many) media buyers are demanding performance guarantees… and getting agreement.
Here's how it works and why it makes sense:
Big Brand is running a campaign to where users create a customized ensemble using a Flash application. They are willing to pay $250,000, but only if they are assured that they will get 75,000 people to engage with the application, or $3.33 per user. They drive a hard bargain. They want people who take the time to complete an ensemble.
These days, it makes sense to say OK. My company, for example, drives traffic to the application with paid placements, paying distributors per user. This is pretty traditional except for the development savings -- but that savings already puts us ahead. Where we win big is that many of the initial users will recommend the campaign to friends, either by forwarding a personalized widget, or posting it in activity feeds for friends to see. Not only is this a deeper level of engagement -- so Big Brand is happy -- but we don’t pay to recruit these additional users. Our profit increases because Big Brand pays for additional completes that we source for "free." The greater the viral leverage, the bigger the upside is for both Big Brand and us.
Making your own serendipity: increasing the odds that your campaign goes viral
No one can predict the next viral campaign. Historically, viral campaigns generally turn on a fluke of some sort. A campaign planned over six months that serendipitously reflects a breaking news event is great, but can you increase your chances of being in the right place at the right time? Have you ever had a great idea for a campaign, but it would only work if you could do it now?
Real-time advertising capabilities won't necessarily make your ideas successful, but for the first time it makes them actionable -- and at little cost or risk. A hypothetical example will make the point.
In October 2008, a disgruntled Iraqi legislator threw his shoes at President Bush during a visit to Baghdad. The story made news worldwide, largely because the quirkiness of the "attack."
Image yourself as a brand manager for specialty shoe maker Shoe Co. Within hours of waking up to the news, you use the Sprout Platform to deploy an interactive, Flash-based widget, and an engaging social media application that allows users to "throw" shoes at a world leader of their choice. Users get to:
- Choose from a selection of world leaders
- Upload their own face as the thrower
- Choose from a selection of shoes to throw
Users make their selections, throw a few shoes at their preferred target, and then are encouraged to forward it to their friends. The level of interactivity drives viral adoption, and within hours the news coverage shifts from the original incident to how your campaign is sweeping the internet.
For a very small investment, the brand manager succeeded in:
- Getting tons of free publicity for Shoe Co.
- Subtly exposing enormous numbers of people to the Shoe Co. brand for an average of 5-8 minutes. Reports tell how many people engaged, for how long, and how many forwarded it to friends or added it to their Facebook stream. The brand manager can also see demographic data on participants (via Google analytics and from social networks) so they know who the key influencers are on social networks.
- Reinforcing the quirky, non-conformist personality of Shoe Co. in ways millions of dollars in advertising couldn't match.
- Gaining real-time insights into user reactions by monitoring campaign reporting, Twitter, and similar services. The brand manager can even update the campaign in real-time in response to the buzz.
- Managing their risk because if it didn't take off -- or worse yet, created a negative reaction – the brand manager could have pulled the campaign before it got wide awareness.
While this story is hypothetical, the premise is completely plausible and all the required services and tools are commercially available today. Even if Shoe Co. had to buy paid placements to initiate the campaign, the viral leverage makes the overall cost extremely attractive.
How the new thinking trumps the old ways
A leading movie studio came to us to create an engagement campaign to reward the 500,000 fans of the film on Facebook. The community had been building since the release of the first movie of the series in 2001. But there was low activity in the group and little evidence they were doing more than just becoming fans of the movie and then leaving the page. There was no way to harness this group and help them spread their love of the movie online.
The studio didn't have a clear idea of what it wanted. The studio reps knew they wanted to give their fans a fun, branded experience but didn't assign goals for the campaign since they saw it as a loyalty versus awareness campaign. They didn't do a marketing spend against the campaign since it wasn’t meant to drive awareness or do more than give existing fans something fun to do.
At the same time, they were spending 25-50 percent of their marketing dollars on social media sites, but with traditional online ads and takeover ads. They used the static model to reach and engage fans, even though they had the tools to do more.
The campaign we rolled out allowed fans to customize their favorite car, add music, videos and photos from the film, or the fan’s collection, and share the personalized widget with friends to see and share on leading social networks.
The results impressed the studio’s creative team. By the time the movie launched, about 60,000 people visited the campaign. There was a 26 percent conversion rate, meaning that more than a quarter of the people who entered the campaign portal published a personalized widget on either their Facebook or MySpace page for their friends to see. What’s more, the average engagement time for the 38,000 people that entered the campaign was almost two minutes.
The other online media that the studio placed performed at a far more "industry standard" rate. There’s no doubt that more money spent driving traffic to the engagement campaign would have made it even more effective, since each activity spurs more friend activity when the widget is posted. Traditional media can’t come close to a 25 percent response, and it can’t spur viral activity.
Economics are changing for both brands and agencies. Brands need to do more with less because recessions are no time to go dark and let your customers stray. In fact, recessions are a great opportunity to build share and awareness, so smart brands are looking to do more -- even if they don't have a bigger overall budget to work with. In practical terms, more with less means that agencies get squeezed, and media buyers have to show concrete results rather the soft metrics such as CPM or page views. Activities are a flexible and extremely useful metric for measuring social media engagement.
With clear objectives from the brand, savvy media buyers are ascending in dominance because they are best positioned to help agencies design focused and effective social media campaigns that guarantee results.
Guaranteeing results is all about leverage and data. Actually making money with a guarantee is about driving engagement and leveraging the interconnectedness of the social graph -- and then being able to measure the results.
Below are eight best practices that will help ensure you get the most from your social media campaigns. These are above and beyond the basic rules of working with social media, such as don't talk down at people, be honest, tell a story, etc.
- Use the right metrics by measuring the activities that represent the type(s) of engagement you seek. If you want people to spread the word, then good metrics are posts to social networks, forwards, streams and activity feeds, tweets, etc. If the goal is brand affinity, use metrics that measure length of engagement and number of activities.
- Bake data and metrics into the early planning stages. If you can't define the metrics, you shouldn't be talking creative yet.
- Leverage the media buyer to help define the creative.
- Guaranteed engagement pricing models align the interests of brand and agency. Alignment is a prerequisite for doing more with less.
- Plan for quicker and more frequent campaigns. Social media users are fickle, trend-driven, and shift focus faster than static model planning cycles can follow.
- Be prepared to turn on a dime. The technology available today allows you to develop and manage campaigns in real-time. But can you keep up?
- If it's not working, cut your losses rather than waiting for overwhelming or "final" data.
- Be viral-ready. No one can guarantee a viral campaign, but it's your job to be ready when serendipity strikes, or better yet, manufacture your own serendipity.
The effectiveness of social media campaigns are changing dramatically as new technologies and thinking empower brand managers, media buyers, and creative directors to do more with less, compared to traditional campaigns. In addition, they engage in ways that traditional campaigns simply can't. The know-how, tools, and program design is within reach to regularly "strum" the social graph to maintain nearly continuous engagement with the target audience.
Carnet Williams is the founder and CEO of Sprout.
A pretty lousy branded YouTube channel in the auto and vehicle category
To take my new lousy scale out for a test drive, I started in YouTube's automotive category. To my surprise, here's what I found:
Based on this data, Buick's branded YouTube channel is pretty lousy. It has the lowest number of subscribers per video in the autos and vehicles category. By comparison, the Mercedes-Benz channel has generated 137 times more subscribers per video.
Since Buick has customized its branded YouTube channel with its own logo, custom background, and other branding elements, I can only conclude that its channel sucks because most of its videos are really lousy.
"It's not my Grandfather's Buick, that's for sure"
For example, one of the newest videos uploaded to the channel is "Craig Zinser Engineering Manager, Infotainment Systems Buick Design."
I'm sure Zinser's "passion for photography" contributes to his work at Buick as an engineering group manager for infotainment systems. But watching the video doesn't make me want to subscribe to Buick's channel.
The oldest video uploaded to the channel is "Buick Regal Remix Event in Chicago Luxury Sport Sedan."
One of the people interviewed in the video says, "It's not my grandfather's Buick, that's for sure." That's not very auspicious if you consider that Oldsmobile proclaimed that it was "not your father's Oldsmobile" in the late 1980s, and that brand was phased out in 2004.
And the most popular video on the channel is "2012 Buick Verano."
This 16-second long ad has 56 likes and 367 dislikes. That's not a good ratio.
This begins to explain why Buick's branded YouTube channel has generated only 4.7 subscribers per uploaded video.
Don't bore your audience
So, what shouldn't you do when creating your brand's channel? Don't upload a bunch of boring videos.
A study conducted at the University of South Australia's Ehrenberg-Bass Institute for Marketing Science found videos that evoke positive emotions (exhilaration, hilarity, astonishment, happiness, and inspiration) were more likely to be shared. I suspect these emotions are also the most likely to engage your customers and prompt them to subscribe to your YouTube brand channel.
For example, take a look at "The Force: Volkswagen Commercial," if you haven't already seen it. This is the most popular video on the Volkswagen USA channel with almost 53.8 million views. It also has more than 205,000 likes and less than 3,400 dislikes. That's a good ratio.
A pint-sized Darth Vader is astonished when he uses the Force on his father's new 2012 Passat in the driveway. This award-winning commercial first aired during the Super Bowl in 2011. According to the Mashable Global Ads Chart, it is also the most shared video ad of all time with more than 5.5 million shares.
Creating a laugh out loud (or a "rolling on the floor laughing") response is a key element to creating a viral video hit according to recent research conducted by the Ehrenberg-Bass Institute for Marketing Science. I think it is also more likely to turn viewers into subscribers.
Dr. Karen Nelson-Field, who conducted the research, says, "Marketers should try and up-scale the degree of arousal so if they think something is amusing, they need to get it to a point where someone will actually physically laugh out loud."
She adds, "I think the opportunity missed for marketers is they send out a video that misses the sharing opportunity but perhaps maybe worse -- bores their audience."
A pretty lousy branded YouTube channel in the entertainment category
Next, I looked at the branded YouTube channels in the entertainment category. Here are the shocking results unveiled by my new lousy scale:
|Epic Meal Time||2,604,612||100||26,046.1|
|The Ellen Show||1,203,418||3,523||341.6|
|Glee on Fox||623,941||154||4,051.6|
|The X Factor UK||459,771||1,331||345.4|
Based on this data, the BBC's branded YouTube channel is pretty lousy. It has the lowest number of subscribers per video in the entertainment category. By comparison, the Epic Meal Time channel has generated more than 687 times more subscribers per video.
I've watched the BBC's content, and it "informs, educates, and entertains." So, I can only conclude that BBC'schannel sucks because it suffers from an embarrassment of riches.
Suffering from an embarrassment of riches
A quick review of the BBC channel's feed tab demonstrates why the largest broadcaster in the world's batting average is so low: The content being pushed to subscribers is so high.
The feed tab shows all of the action that's happening on a channel, and the BBC channel's feed tab is pushing an average of 40 videos and two playlists a week to its subscribers. These videos and playlists include "highlights" from BBC One, BBC Two, BBC Three, BBC Four, and Radio 4 programs.
In contrast, the Epic Meal Time channel's feed tab is pushing an average of one video a week to its subscribers. These videos from the YouTube cooking show feature extremely high-calorie meals, generally created out of meat products (with particular emphasis on bacon) and alcohol (especially Jack Daniel's).
Too much of a good thing
So, what shouldn't you do when creating your brand's channel? Publishing content regularly and often is a good thing, but too much of a good thing is generally a bad thing.
A good level to aim for is a minimum of one video per week, but the right amount of content depends on your audience, your goals, and your content. One or more weekly uploads will provide new content to your audience with enough frequency to keep them coming back.
According to comScore Video Metrix, the average viewer in the U.S. watched 118.3 videos at YouTube.com for 484.4 minutes (8.1 hours) in June 2012. That's an average of 27.5 videos a week for 112.7 minutes (1.9 hours).
So, uploading an average of 40 videos and two playlists a week to your YouTube brand channel is probably counter-productive. It's more content than the average viewer will watch.
To determine that for yourself, you should use YouTube Analytics to regularly to assess your channel's performance. What you learn can help inform programming decisions for your channel.
For example, examine the dates or videos where there was a high gain or loss of subscribers to learn more about what resonates with your audience. And identify and analyze videos that drove the most subscriptions relative to how many views they received to learn what caused more viewers to become subscribers.
And if you just happen to be the marketer who is responsible for the BBC's branded YouTube channel, then you might consider using the network template on your existing channel to showcase five new channels for your BBC One, BBC Two, BBC Three, BBC Four, and Radio 4 programs.
Now, I realize that a Brit isn't likely to take advice from a Yank who wouldn't know "a wicked googly" if he saw one. But you should still choose the right template for your channel to create the best experience for your subscribers.
A pretty lousy branded YouTube channel in the science and tech category
Next, I examined the branded YouTube channels in the science and tech category. Here's what my new lousy scale discovered:
|Google Tech Talks||99,226||1,688||58.8|
Based on this data, Intel's branded YouTube channel is pretty lousy. It has the lowest number of subscribers per video in the science and tech category. By comparison, the Apple channel has generated almost 862 times more subscribers per video.
But even if it sucks, Intel's channel isn't the elephant in the room.
The elephant in the room
I'm sure you noticed that half of the branded YouTube channels in the tech category belong to Google. As a matter of fact, there are more than 100 official Google YouTube channels.
Now, Google acquired YouTube in November 2006 for $1.65 billion. So, maybe it isn't surprising that the company has more YouTube brand channels than you can shake a stick at. But, is this a smart strategy that you should adopt? It depends. It's "wicked smart," as we'd say in Boston, if each channel is targeted at an ideal market segment that meets all of the following criteria:
- It's possible to measure.
- It's large enough to earn profit.
- It's stable enough that it doesn't vanish after a short time.
- It's possible to reach potential customers via promotion and distribution channels.
- It's internally homogeneous (potential customers in the same segment prefer the same product qualities).
- It's externally heterogeneous (potential customers from different segments have basically different quality preferences).
- It responds similarly to a market stimulus.
- It can be cost-efficiently reached by market intervention.
- It's useful in deciding on marketing mix.
Market segmentation 101
Google has more than 101 ideal market segments worldwide. So, creating more than 101 YouTube channels that feature videos on a wide range of topics, from product demos to visiting speakers to Google's offices, is a wicked smart strategy.
If you look at Google's YouTube directory, you'll see that 42 channels are focused on users and advertisers in 36 different countries and regions. That's wicked smart. Why would you create just one channel for users and advertisers who speak 27 different languages?
Another 30 of Google's YouTube channels are focused on users of different products, and four more are focused on developers. That's also wicked smart. If you want information on Google Chrome, just go to the Google Chrome channel. Why force users to search through videos on other products?
Finally, 20 of the official Google YouTube channels are company-wide. That's probably smart, too. If a user wants to subscribe to Google Search Stories to be informed as soon as a new one is released, why make them receive updates about other topics they're not interested in watching?
Unless your company has only one ideal market segment, then it makes little sense to have just one YouTube brand channel.
Improving the batting average of your branded YouTube channel
Hopefully this article has helped you to take your channel to the next level. If nothing else, I hope my new lousy scale has demonstrated what not to do when creating your brand's channel.
If we've learned anything from these three lousy branded YouTube channels, it's this: Don't jump to conclusions.
When YouTube launched its new channel layouts in December 2011, it made it easier for all channel creators to organize and showcase exactly what they wanted. This included a more streamlined and consistent design as well as new, more flexible layouts for featured content.
So, incorporating banners and background images to match your brand's look and adding links to other sites is necessary but not sufficient to creating a very good YouTube brand channel. To avoid ending up with a lousy channel, you also need to generate more subscribers per uploaded video than other channels in your category.
This means you need to improve the batting average for your branded YouTube channel. And it's worth nothing that "batting average" is a statistic that measures the performance of batsmen in cricket as well as hitters in baseball.
On Twitter? Follow iMedia Connection at @iMediaTweet.