data profiles | MarketServ™

Posts tagged "data profiles"

Going once, going twi…too late! Smoked by Real-time bidding.

In less than 100 milliseconds, Real-time bidding (RTB) initiates and completes the process to deliver a single online ad exposure via programmatic auction to the viewer.

100 milliseconds. How fast is that? In 134 milliseconds, light can travel around the Earth’s equator, and it takes 200 milliseconds for the human brain to recognize emotion in facial expressions.[i]

We explained the role of data-driven user profiles in “Stop guesstimating and get with the program” as the key to precise audience targeting in large scalable quantities for programmatic display campaigns.

Now consider that Real-time bidding is a per impression auction for those campaigns.

When someone visits a Web site – while the page is being loaded into the user’s browser window – the site publisher sends a bid request for each of the display ad slots available to one or more ad exchanges. The ad exchange submits the details for each slot to multiple advertisers who submit their bids in real time. The winners’ ads are then displayed in the banner and other slots when the page is loaded…all in less than 100 milliseconds.

RTB and User Profile data-driven targeting work in tandem to deliver ever-increasing ROI for the campaign budget over the duration of the campaign.

The criteria that created the high-value User Profile target audience can dictate whether to bid and the value of the bid. As viewers click on ads, user and campaign data are updated, and that information, in turn, improves the performance of the next ad to be bid and served.

Since multiple RTB exchanges (and there are dozens) serve bid opportunities simultaneously, data management platforms have emerged that centralize these bid calls. At the same time, the growth of centralized User Profiles of first hand, online history for millions of anonymous users – you and me – means the best campaigns can find us and show us the ads that match us ever-more precisely.

This is where size matters to our campaign outcomes. The more first hand data in the Profiles and the more RTB exchanges available in those 100 milliseconds, the better our ROI.

Access to this size is not the exclusive province of the largest advertisers and their agencies. Any business that has significant expenditures in print, broadcast, or digital should be re-balancing their mix of channels to test and include this combination of User Profiles and Real-time bidding. The more data in the profiles and the more simultaneous auctions you’re in, the better you’ll do.

Otherwise, your business could end up, “Going once! Going twice! Gone!”

[i] http://en.wikipedia.org/wiki/Millisecond

Stop guesstimating and get with the program.

We contrasted the differences between data rich User Profiles for ad targeting and the shallower profiles used by search engines and social networks in our recent post “Getting personal” .

For targeted campaign success, data rich User Profiles can win out in a big way.

Here’s a real-world example where the same advertiser tried both – a Google-based strategy by a search engine agency and a User Profile strategy that we provided.

By every measure, the MarketServ™ strategy out-performed the Google strategy.

Google Campaign by Others
Name	Spend	Clicks	Cost/Click	Impressions
Display	$6,625.77	2,748	$2.41	614,090
Search	$5,908.56	2,461	$2.40	82,306
TOTAL	$12,534.33	5,209	$2.41	696,396

MarketServ™ Campaign by Iris
Name	Spend	Clicks	Cost/Click	Impressions
Audience Targeting	$7,500.00	6,456	$1.16	2,425,040
Content Targeting	$9,700.00	10,992	$0.88	4,741,133
Site/Search Retargeting	$2,799.00	16,962	$0.17	1,082,862
TOTAL	$19,999.00	34,410	$0.58	8,249,035

The reasons shed light on the rapid growth of Programmatic Advertising, which applies data analytics and information technology to advertising tactics and operations.

Programmatic Advertising has arrived in full force because it works. Computers replace guesswork in ad spending to deliver higher ROI.

Big sets of data and machine-driven analytics develop User Profiles for better targeting precision from campaign planning through execution. And information technology similar to the stock market optimizes the “next ad served” at the most efficient price throughout the campaign rather than human guesswork to make “wait & see” adjustments in the Google campaign.

Audience and Content Targeting:

The MarketServ™ campaign served ads to an Audience developed from proprietary User Profiles of recorded behaviors that match those of the client’s customers. Detailed online histories identified their lifestyles, shopping habits, browsing and search habits, and the web sites they visited regularly and on what type of device.

Google served ads to those who searched and discovered the client’s site regardless of any match to characteristics of its customers or objectives of the campaign.

Site/Search Retargeting/Remarketing:

Because the MarketServ™ audience was so highly qualified in the first place, retargeting to its audience resulted in a very high click through rate and correspondingly low Cost Per Click compared to Google.

Information Technology and Management:

The software that managed the MarketServ™ campaign is used by the biggest brands for their campaigns. Its algorithms continuously optimized the campaign to serve the next ad based on predictive analytics of preceding responses in the campaign.

Experienced ad traffic managers directly at the data management platform interpret statistical results and manage adjustments.

Google campaigns require intensive management time by the advertiser (or at least a surrogate). Its software does not have optimization capabilities. The human interface has to “best guesstimate” everything from keyword refinements to what bid price will win to expose the ad.

Will every advertiser get the same results? Of course not.

But rather than replace the human element, Programmatic Advertising helps advertisers make educated decisions to best achieve marketing and campaign objectives to drive traffic to your web site whatever your goals – such as generate leads for your funnel, build sales, or create awareness to support other advertising channels.

To maintain competitive edge, businesses need to get on board to the transformational opportunities of Programmatic Advertising.

A traveler’s guide to Geo-targeted Advertising

Adventure travel and geo-targeted Internet advertising have a lot in common.

Consider this definition from Wikipedia: “Adventure travel is a type of tourism, involving exploration or travel with perceived (and possibly actual) risk, and potentially requiring specialized skills and physical exertion.”

Well, maybe not the physical exertion because you’re doing the heavy lifting at your desk, but you’re certainly risking dollars with your specialized skills.

Geo-targeted display ads are the rage promoted by the major tour operators – Facebook, Twitter, LinkedIn, Google, Yahoo, Bing. They’re all clamoring for your ad dollars.

And while the Web can take your ad anywhere in the world, your campaign may not come back with the souvenirs like you hoped when you planned your adventure.

So, here are some travel tips before you set out on your journey.

Know the technology that’s being used to deliver your ad. And remember – your ad is targeted to devices, not the users and profiles of those users (more on that later).

IP-based targeting is the underlying technology – the compass – used by the search engines and social media ad programs. The limitations of IP mapping can be like magnets that throw the compass’ needle out of whack.

IP-based targeting delivers your ad to the computer’s geographic location based on the Internet Protocol address assigned by its ISP (Internet Service Provider). The IP address does not have a direct correlation to exact geography like a ZIP code. Private networks, corporate networks, rural and exurban areas (from where ISP’s reroute users to their server buildings) are just some of the intervening factors.

Travelers Tip: IP mapping accuracy is not 100%. The more granular you get, the less accurate it becomes so your ad will be delivered to viewers who aren’t in the territory. It’s pretty good down to the City/Metropolitan area level. It’s offered at the zip code level, but approach cautiously.

A/B test until you get it right. For example, bundle a few neighboring zip codes rather than using just one or expand/contract the radius around your center point.

Mobile or Location-based targeting is the other underlying technology, but despite the huge gains in mobile access by users, it has not gained traction with the advertising programs offered by search engines and social networks.

Mobile or Location-based targeting gets its geographic information in one of two ways: Confirmed or Derived. Confirmed location targeting depends on the user turning on the device’s GPS capabilities or when the user “checks in” at a location. Derived relies on inferences such as tri-angulation of cell towers and mapped Wi-Fi hot spots.

IP-base targeting to devices is much less sophisticated than targeting actual users that fit an actual profile of your target audience. Since search engines and social networks don’t maintain the depth of user information to support user profile-based advertising, IP targeting works for them.

Travelers Tip: IP-based geo-targeting is a good way to get your campaign into the general area for most of your audience, but it’s not a precise map to the pot of gold at the end of the rainbow. You need to include other targeting factors when planning your campaign.

So user information comes into play. You need to know who are you targeting.

The search engines and social networks aggregate their data and also apply some analytics to create modifiers to refine the Geo-targeting for better ROI. Google, for example, uses the search histories of their users. Facebook uses Profile information and aggregates information on users’ friends to better define the individual.

User-profile based targeting is highly effective for hyper-targeted campaigns because the profiles have much more direct 1st party online data that have been gathered from multiple sources. All kinds of behavioral and shopping history data are in each profile. Geographic information – both IP and Location-based – are just a couple of the hundreds of attributes within each profile. This allows the data managers to create very specific campaign Audiences from millions of profiles with hundreds of attributes on each profile.

Geo-targeting and User profile targeting can both lead to great campaigns. Like travel plans, your campaigns cost thousands of dollars. Plan your budget to try both, and scale up what works best over time.

Happy Trails…