First time accepted submitter jonez450 writes “Times are tough in the advertising business. But PCWorld publisher Ziff Davis has come up with a new plan to gain a competitive edge: Paying other tech sites $1 CPM to place tracking code on their sites in return for data about their users via JavaScript. The company is also offering free content in return, but the ‘private’ Ziff Davis Tech Co-Op doesn’t want anyone to know what they are up to.”
Edit Note: This is a guest post by Dane Jasper, CEO of ISP Sonic.net. The post can also be found on the Sonic.net CEO blog here.
Recently a number of ISPs have been caught improperly redirecting end-user traffic in order to generate affiliate payments, using a system from Paxfire. A class action lawsuit has been filed against Paxfire and one of the ISPs.
This is a serious allegation, but it’s the tip of the iceberg. I’m not sure if everyone understands the levels of sneakiness that service providers can engage in. So, while I’m no expert (as we are an ISP who doesn’t do these things), but as a broad overview, here is my quick guide to the five levels of ISP evil, and the various “opportunities to monetize customers” that we’ve passed on:
5: Improper NXDOMAIN handling, also known as “Domain Helper” applications. When a customer attempts to visit an invalid site, instead of returning the RFC standard “no such domain” response, the servers provide a search result which includes sponsored links. Sometimes the results are not well matched to the mis-typed domain, and they promote ads instead with broad commercial appeal like insurance, which will generate a high payout if the customer clicks. Extra evil points for making it difficult to opt out of this, requiring opt-out via a cookie or browser setting rather than providing “clean” DNS servers. (Paxfire’s system is positioned as a search/helper application, but these systems can be easily converted, even without the ISP’s awareness, to an affiliate pumping system.) Evil score: 2 evil points, somewhat evil, but now every major access provider provides helpful results for address typos.
A diagram showing how Phorm's “Webwise” system creates copies of its tracking cookie in each domain the end-user visits, based on the report published by Richard Clayton. Wikipedia.
4: Clickstream Tracking. An ISP is in the unique position as the point of traffic origination, creating the opportunity for very in-depth analysis of Internet usage behavior. Tracking the user’s Clickstream, the site to site to site movement as they browse using a set of tools like Phorm allows service providers to create cash out of information about private use of the Internet. Clickstream data buyers are generally ad targetting; if you visited Ford.com and looked at F-250 trucks, then CNN.com, it might make sense to place ads for large Chevy trucks on the CNN page rather than an ad for fabric softener. Absent this prior knowledge that you were a potential truck buyer, the ads might be for something of less interest to you, and thus less likely to be clicked, to “monetize”. Over time, analysis of the complete Clickstream can provide lots of insight to advertisers. Extra evil points for selling the Clickstream data without telling customers. Evil score: 5. What you do online is private!
3: Ad Swapping. Transparently proxy all web traffic, and when ad banners are in transit, perform real-time swaps of the ads for other ads for which the ISP is getting a cut of the revenue. Legitimate advertiser ads are sometimes fetched so that no one notices the decline in impressions. The pitch to ISPs from companies like NebuAd sometimes included claims of “partnerships” with content sites to better target ads. Extra evil points for ISPs who provide demographic data to the firm running the ad-swapping system. Evil score: 6.
Our reply: “No, not interested, thanks. -Dane” Email reply to Mark Lewyn, President, Paxfire Inc., Wednesday, October 29, 2008 3:35 PM
2: Affiliate Program Pumping. As alleged in the Paxfire scheme, ISPs or their accomplices take incomplete or incorrect domain entries into the URL bar and direct them to an intermediate page, which redirects transparently to a URL which includes an affiliate tag. So, a consumer types “amazon”, and rather than returning an NXDOMAIN, or even a search result, the ISP DNS server directs them to an IP address which does a content reload toward a URL of the form amazon.com/affiliate-id=XYZ. Purchases made subsequently are compensated as if it was legitimate traffic from an affiliate. Evil score: 8, with a bonus point for poisoning the affiliate ecosystem.
1: Rolling Over. In an attempt to avoid costs or under pressure from government or content creators, ISPs have handed over customer information, and even subjected customer traffic to broad snooping. Allegations range from service providers simply quietly handing over customer info to law firms with improperly filed lawsuits and incorrectly served supoenas, to the physical wire-tapping of major fiber optic lines. We’ve got your back. Evil score: 10. Potential for human rights violation.
I’ve got more to say on this last topic, but there is a clock that must run out before I am permitted to write. Tick-tock, a couple days to go.
Related research and analysis from GigaOM Pro: Subscriber content. Sign up for a free trial.
Growing Mobile Data Use Turned Up Heat on Carriers in Q3U.S. Wireless Data Market: Q4 and Year-End 2008Flash analysis: Collaborative consumption – a first look at the new web-sharing economy
Jamie Beckland is a Digital and Social Media Strategist at Janrain where he helps Fortune 1000 companies integrate social media technologies into their websites to improve user acquisition and engagement. He has built online communities since 2004. He tweets as @Beckland.
Marketers have built a temple that needs to be torn down. Demographics have defined the target consumer for more than half a century — poorly. Now, with emerging interest graphs from social networks, behavioral data from search outlets and lifecycle forecasting, we have much better ways of targeting potential customers.
The rise of mass-produced consumer goods also brought the rise of mass-market advertising. In the 1950s and 1960s, the goal of television was to aggregate the most possible eyeballs for advertisers. In order to convince consumers that an advertising message was relevant to them, consumers had to buy the idea that they were just like everyone else.
Marketers created that buy-in by bucketing people into generations. When you lump 78 million people into one group called “Baby Boomers,” it’s much easier to sell them stuff, especially when consumers accepted their generational classification.
But now, that entire system has broken down. The year that someone was born will not tell you how likely he is to buy your product.
Fragmentation is now the norm because the pace of change is accelerating. Generations have been getting smaller because there are fewer unifying characteristics of young people today than ever before:
With the recent rise of the social web, people self-select into groups so small, so fragmented, and so temporal, that no overarching top-down approach could be successful at driving marketing performance.
Marketers have responded by adding more demographic information to the mix, but even that is a losing battle. I worked with one client who was introducing a technology product, and had identified a target market of “connected consumers.” Connected consumers were 34-55, had a household income over $120k, and read technology publications regularly. This target market represented 14 million consumers.
They were targeting 14 million consumers to sell 50,000 units — that means they were hoping for 3.5 sales for every 1,000 people with whom they connected through their marketing.
What if, instead, you could get 500 sales from every 1,000 people you marketed to?
It’s possible through psychographic profiling. Psychographics look at the mental model of the consumer in the context of a customer lifecycle. Amazon.com has long been a leader in this space, through innovations like “recommended products” and “users like me also bought.” Its algorithms have learned to predict its users, and what they are interested in. And now, there are a number of tools that any business can use to leverage psychographics.
Here’s how a psychographic profile might look different from a traditional marketing profile target for a childcare provider:
Psychographics provide much more useful information about users. There are multiple data sources making this possible today. Social profile data, behavioral data and customer lifecycle data can now finally be leveraged to contact people who are ready to buy.
Social Profile Data
Profile data from social networks consist of all the fields users grant permission for brands to use on their behalf. Most things that users track on social networks can be leveraged to create a closer relationship with a customer. Fields like relationship status, alma mater, interests and occupation can all be managed through social profile data management tools.
Social profile data is the critical cornerstone of psychographic insights. The level of nuance and insight provided by social data, when compared to standard demographics, is the difference between performing surgery with a scalpel or a butter knife. Previously unimaginable questions are now routine:
Are customers who kayak more likely to buy water shoes than those who canoe?Who is more likely to spend over $100 on an order: Seattle Seahawks fans or Seattle Mariners fans?Are your customers more likely to purchase when they move across the state or across the country?
In addition, companies such as GraphEffect are measuring purchase intent by doing semantic analysis on Facebook status updates. This type of qualitative analysis can move users into specific marketing funnels from their very first online experience with your brand.
Behavioral Data
Retargeting advertising messages is gaining popularity among marketers, but its very success has jeopardized its effectiveness. Ads that follow users around the web have been implemented — usually poorly. Every ad network quickly incorporated the ability to place cookies in users’ browsers, and display specific ads to them any time they visit a site that’s part of their networks.
The next generation of ad targeting will focus more on telling the customer a story over time, based on specific behavior triggers. That means ad networks and clickstream data aggregators will work together to trigger when a customer moves forward in a mental model toward a purchase event.
Site content and product recommendations will also be informed by clickstream analysis. Companies such as RichRelevance, Certona, Baynote and Monetate all offer the ability to personalize information to specific visitors based on their behavior. Leveraging those alongside a payload of social profile data can turbocharge those services from the first moment a new user visits a site.
Customer Lifecycle Data
Social profile data can also be used to predict customer lifecycle. Imagine knowing not only if a customer has children, but the exact ages of those children. In addition, key indicator purchases, like buying diapers for the first time, indicate a customer entering a new lifecycle. Other key indicators, like shipping address changes, first purchases of furniture, or first purchases of substantially higher-value goods can all indicate the start of a new customer mentality and behavior pattern.
These patterns are predictable, so you know the future behavior of high school seniors by looking at the current behavior of college freshmen. By using demographics alone, all high school graduates would be marketed to identically. Using psychographics, we know who is likely to be interested in specific product or content recommendations at a specific time — such as when they actually start their first day of college.
This vision is starting to gain traction among serious marketers. At the 2009 Internet Strategy Forum, Xerox’s VP of Interactive Marketing, Duane Schulz, said that a 1% clickthrough rate was a huge failure — even though it is 10 times the industry average. In his mind, a successful campaign would never waste 99% of its impressions. Using psychographic data, you don’t have to waste any impressions.
We have seen a similar upheaval in marketing before. In the 1960s, marketers who embraced the power of television, broad-based insights into psychology and demographic data created world-class brands and billions of dollars in value. At that time, if you didn’t advertise on TV, you lost. Today’s new tools offer a similar choice: Build a deep understanding of your customer, or risk irrelevance.
Image courtesy of iStockphoto, porcorex
More About: advertising, business, data, demographics, MARKETING, social media, trending
For more Business & Marketing coverage:Follow Mashable Business & Marketing on TwitterBecome a Fan on FacebookSubscribe to the Business & Marketing channelDownload our free apps for Android, Mac, iPhone and iPad
If you weren’t convinced about Twitter’s advertising platform, new statistics from the company’s CEO Dick Costolo might have you looking back into putting some money aside to advertise on the service.
Spreaking to Walt Mossberg at AllThingsD’s D9 event, Costolo said that Twitter was a “remarkably successful business” that currently has over 80% of its advertisers choosing to renew their campaigns on the service. Referring again to Twitter ads as “orders of magnitude higher” than traditional web adverts, advertisers including Volvo have seen 50% engagement rates via one advert on the Twitter website.
Radio Shack are reported to have upped their instore exchanges and purchases by double digits from the day before they ran an advert on Twitter, keeping the campaign exclusive to Twitter users.
Internally, Twitter said it would have 100 advertisers by the end of the year, in reality it had 150. This year, the company is working with over 600 companies. Costolo says the problem isn’t finding advertising partners, it’s making sure that it “doesn’t jam” the service with too many ads.
When asked whether Twitter was profitable, Costolo successfully managed to dodge the subject.
Impressive figures, but of course the company is going to concentrate on the more successful campaigns. With Twitter about to introduce its own photo sharing service with Photobucket, we could see more targeted advertising campaigns in the very near future.
Internet advertising revenues in the U.S. hit $7.3 billion for the first quarter of 2011, a 23 percent increase over the same period in 2010, according to figures released today by the Interactive Advertising Bureau and PricewaterhouseCoopers. This is the highest first-quarter revenue level ever for the industry.
Considering the pretty woeful hit rate of most Internet advertising, that may seem surprising, but as traditional ad destinations like broadcast TV and print publications continue to lose ground, advertisers probably don’t have many options. The technology continues to improve, however.
Google’s AdMob advertising network recently announced new formats for tablet advertising, mobile advertising for small businesses and improved in-app advertising. AdMob is receiving 3.5 times as many ad requests as it did a year ago.
Startup LocalResponse has introduced check-in advertising. The company aggregates check-ins and mentions of merchants across multiple social networks including Facebook, Twitter, Gowalla and Foursquare. Merchants can view the check-ins and develop a response, such as a coupon, which is sent via Twitter to the check-in customer. The company claims a 60 percent click-through rate, which is spectacular in comparison to the click through rate of most online advertising.
Google Chairman Eric Schmidt, a man who should know a thing or two about advertising, made a personal investment in Eyeview, which allows online video advertisers to show personalized ads based on location, demographic, time, weather and browsing history.
While this is all interesting technology, I can’t help thinking of former Facebook data guru Jeff Hammerbacher’s comment in BusinessWeek: “The best minds of my generation are thinking about how to make people click ads. That sucks.” he said.