This I believe

So for several months now I have been closely watching the industry, reading articles, PR releases, product releases, financial releases and generally paying attention.  I have been taking bits and pieces of what I care about and have been making comments and providing what I believe has been an insightful perspective on how technology can be better leveraged to improve how we advertise online.


 


I have looked at the ad servers, the networks, the lead generation tools.  I have examined the search engines, the publisher tools and the creative formats and provided you with feedback on what other reliable people have had to say.


 


And I have discussed ad agencies, their workflow, the media buying process and the tools that people use to do their jobs, however inefficiently I may believe that to be.


 


Here is my position.


 


The agencies have engaged the interactive medium completely.  Estimates for 2007 are that between 12 ½ and 20% of advertising budgets will go towards new media.  This is no longer the edgy side-project.  Engagement with technology is here.  But it is about refinement. 


 


Search is essential but everyone is coming to recognize that there is something wrong with the model.  It is extremely time-intensive and expensive to manage.  Furthermore the ROI metrics seem to slip the longer you run campaigns.  As I have said in the past, a tipping point is coming. 


 


Networks have been doing their thing the same way with some minor tweaks for a while now and people are demanding more disclosure.  Tolerance for media showing up on inappropriate sites is very low, accountability is high and additional capabilities like behavioral targeting has become an expectation.


 


That brings me to behavioral targeting – a very common topic on this blog.  I have ripped this topic up and down.  My intent has been to redefine this concept as event-based targeting and to justify that there is little about behaviors actually associated with it at all.  Just because someone took a navigation path, or saw a page means little about their behavior, the predictability of their behavior or their preferences.  All we know is something that happened.  Historical targeting is a better description but I have used event-based targeting over the last couple of months.


 


I have never tried to minimize the value of BT, only put it into it rightful place as a solid prospecting and direct response advertising mechanism.  BT does not represent the best means to capturing known individuals, in fact, it does not have the capacity to associate with knowledge about people at all.  Only with events.  But I believe that BT should be part of a comprehensive campaign.


 


People have approached me both on and off this blog about my position towards BT and some of the networks, but I think its because they have been defensive and protective of their positions as representatives of these companies.  Others have engaged me – usually advertisers, agency representatives or others who see that the evolution of practice is inevitable and being on the adaptive edge of the curve is better than the laggards edge.


 


I have also spent a lot of time plugging a concept called first party ad serving.  Forgive me for the plugs.  Obviously as a member of TruEffect I have a lot of passion for what we do here.  But I also spend a lot of time looking for other technologies that can rival or at least coexist, companion or compliment what we are doing here.


 


The patent-pending DirectServe™ Technology that TruEffect has brought to the market represents the next generation of ad serving.  It leverages the knowledge that an advertiser holds about its customers, registrants or users to re-target through ad campaigns.  This is not a replacement for other technologies out there – I have said that before as well – but a great new way of doing it.  An addition to a comprehensive advertising strategy.


 


First party ad serving is about customer re-targeting.  BT is about event-based targeting, best applied when trying to capture unknown individuals.  One is for bringing in new business; one is about farming and growing existing business.  There is no point is re-prospecting existing customers while advertising online.  DirectServe™ takes care of that.  BT leverages previous events so that you can increase the likelihood of putting the right message in front of the right person at the right time based on historical events.  DirectServe™ puts the right message in front of the right person based on known customer segmentation models, knowledge already held about customers.  This a potential marriage.


 


Now BT is largely touted by networks, so there is a limitation as to how you can use it.  I talk about TACODA a lot – which I think Dave Morgan has not be thrilled about – but its because they have been the leader in the space.  I have also talked about Advertising.com and Blue Streak and Tribal and others as well.  But ad servers offer BT too.  DoubleClick’s Boomerang does it.  TruEffect does it.  And that extends beyond networks.


 


I also talk about integration.  Agencies are not on this trail so much as advertisers.  Well, some agencies are but they are the minority.  I have strong opinions about this because I feel that they pieces of the puzzle are all here now for us to put together a great picture of our online marketing so that we can make better informed decisions about our web site compositions, product placements, online advertising and budget allocations.  But nobody has fully engaged yet.  There are leaders that are putting the pieces together, but I am advocating the full-monty and that is what you read about on this blog.


 


Tying it all together will enable an advertiser to make the best possible decisions regarding allocation of online media spend.  It will promote the best utilization of technology, improve product placement on web sites, increase the value of existing customers, the initial value of new customers and enhance the likelihood of increasing the utilization of interactive media as a channel for marketing.


 


See my ten-step recipe for full-integration of all the technology pieces of an online advertising campaign.


 


First let me redefine that a third party cookie is a vendor’s cookie and a first party cookie is an advertiser’s cookie.  Here is the recipe.

_uacct = “UA-980395-1″;
urchinTracker();

Ten Step Recipe for a Fully-Integrated Online Marketing Initiative

In my next post, This I believe, I state my position on agencies, networks, ad servers, writers and pretty much all things online advertising related.  In the end I offer up a reciper for putting together all of the technologies into a holistic, comprehensive marketing initiative.  Here it is.

First let me redefine that a third party cookie is a vendor’s cookie and a first party cookie is an advertiser’s cookie.  Here is the recipe:


 


(1)            Starting with the tagging of a web site so that cookies can be set (first party cookies of course) when someone is on the site. 


(2)            Then add in Ad serving – first party ad serving (like the patent-pending DirectServe™) to promote products or services on the web. 


(3)            Mix in the search advertising and be sure to use the ad server’s first party cookie and leverage a redirect so that the search term can be embedded into the cookie so that when the user lands on the advertiser’s web page that search term is associated with that user.


(4)            Deploy customer re-targeting advertising whereby you leverage customer segments from the eCRM database to recognize and distinguish customers through ad campaigns with specific banner advertisements, rich media and video.


(5)            Deploy BT for anyone that visits the web site directly and cookie that user with a first-party cookie so that the re-targeting mechanism can work when they encounter that individual online (on a network or web site).


(6)            Engage with a site-side analytics provider that will use a first party deployment – like WebSideStory – and take full advantage of tracking anonymous user behavior across your web site.  Track all navigation patterns, entrance and exist points, product position preferences, sales cycles, etc.


(7)            Leverage site-side analytics to write the first party cookie and segment the cookie value based on user preferences in association with CRM data.


(8)            Feed the customer preference cookie value to generate the customer segments and associate the customer segments with creative target groups that the ad server will serve in the re-target campaigns (back to #4)


(9)            Have the DirectServe™ write to the cookie during the ad serving process the details of the ad serving history to the cookie so that when a prospect of customer comes to the site, the site-side analytics software can internalize all of the external activity and use it for further analysis on how someone became a customer or how returning customers were reacquired.  First party ad serving will give site-side analytics outside perspective of web marketing.


(10)       Complete the circle by leveraging the holistic view by analyzing the reports on how you acquire customers through BT, Search and Banner advertising, how you re-acquire existing customers through DirectServe™ first party ad serving and how both types interact with your site using site-side analytics.  Determine which messaging strategies, campaign combinations of banners and placements, search engines and terms and which technologies are delivering the greatest source of new customer and returning customer yield and make future media buying decisions based on that analysis.

_uacct = “UA-980395-1″;
urchinTracker();

Apply Direct Mail Knowledge to Online, and Then Do it


Dave Wilson, president of Wilson Relationship Marketing Services wrote in iMediaconnection today bout extending offline direct mail practices to the online world.  His incite about the established world of customer segmentation and modeling in the offline world stimulates the mind about the possibilities online but then he only scratches the surface with regard to what’s out there.


 


Dave says that unlike offline, “…in the online world marketing activities are geared towards “pulling” prospects from an audience of identified potential customers. To understand and identify the right targets, marketers need to know the key attributes of prospects and customers online just like in the offline world. As a result, customer segmentation becomes imperative online, whether the media channel is search engine marketing (SEM), banner ads on portals or email campaigns.”


 


So media selection is a key component of segmentation.  That is common practice today.  Vertical sites, networks and search are all widely accepted.  Dave talks about subject-matter verticals like news or entertainment-rich content sites being capable of delivering a specific demographic while specific networks may be capable of delivering another demographic type.  Search engines can deliver keyword targets and some even go so far as provide vertical search on a site-by-site basis.  LookSmart is a vertical search engine provider for example.  This is all good when you are looking to drive traffic which, of course, is a huge component of online advertising.  But with online marketing – like offline direct marketing – there is more.


 


The web is the newest greatest one-to-one communication mechanism of our time right?  The greatest thing since the TV, only now you can communicate and not just broadcast.  So segmenting an audience is about generating a response.  Segmenting prospects is about generating customers.  Segmenting customers is about driving repeat business.  Direct marketing covers both.  Not just one.  When we examine the direct to consumer communication capabilities of online marketing we have to go beyond the geo-targeting, day-part targeting and demographic preferences of a site, search engine or email list.  We have to also look at behavioral targeting.  We have to look at psychographic segmentation.  We have to look at cross-channel customer knowledge and segmentation data.


 


Okay-okay.  Where is the lunatic going?  First of all behavioral targeting.  My good friends at TACODA and Advertising.com offer healthy targeting capabilities based on event-based behaviors.  Pixel your advertisers’ sites and you can target users across their networks based on people who have been on your site in the past.  Or, can you do it better?  Yup – but that way works well if you are targeting their network.  You can use an ad server to pixel your web pages and target users web-wide, including users across a network who have conducted specific events on your advertisers’ sites.  So if they have seen a product page in the past and then you encounter them on the web, you can target them with specific ads.  Better prospecting.  The network model works great is you are primarily advertising on the network.  The ad server model works better if you are casting a wider net across the internet.  Check out DoubleClick’s Boomerang or TruEffect’s DirectServe for basic behavioral target prospecting.


 


Now, direct marketing is also about going after existing customers too.  Catalogers don’t just go after prospects, they hit the hell out of existing shoppers.  And so should you.  Creating customer segments of your online customers is vital.  Remember I said cross-channel segmentation?  If your business has an offline component and an online component like – say Target or LL Bean – then you should be marrying your offline and online data.  The deeper your customer database the more you can segment your customers for online direct marketing.


 


Online direct marketing should be conducted in two places simultaneously, site-side and web-wide.  Site-side is customer targeting while they are on your web site.  Using first-party cookies in conjunction with user logins is paramount.  Know your customers and leverage that information to guide them through to buy-buy-buy.  Okay, maybe not that aggressive, but your content management system should be geared towards displaying optimized products based on the segment membership buckets you have placed them.  Categorized behavior and predictive modeling should enable you to increase the likelihood to buy and to increase the amount per purchase per visit.  Once someone logs-on you can access their account and user personally identifiable information.  Prior to logging on you can use first party cookies and anonymous profile data.  Both are very powerful.


 


And here it is, my little plug.  Web-wide direct marketing to existing customers may be accomplished through customer re-targeting through your online advertising campaigns.  If you are advertising online, and you frequently advertise on the same sites because you perform well on those publisher sites you have to know that your customers frequent those sites too.  That means that you are paying to re-prospect your existing customers.  What % of your advertising audience is comprised of existing audience.  How many impressions are you wasting on people who already shop with you?  Would your dollars be better utilized direct marketing to existing customers on those sites while you simultaneously prospect-message to non-customers?  Then read anyone of my many posts that talk about DirectServe and first party ad serving.  Because that technology will tie it altogether for you.


 


One last tid-bit.  TruEffect offers something called TruTags™.  Neat little patent-protected item they threw out for 2007.  It is a site tag that you put on advertiser web pages that will simultaneously fire other tags.  So you only need to tag your advertiser’s web pages one time and you can add as many other tags as you need to the TruTag™ repository.  Its really nice for those advertisers who are difficult when it comes time to adding tags to their web pages for additional tracking mechanisms.  You simply TruTag™ their site and then as an agency you control the addition of extra tracking tags.  So you can use behavioral targeting of other providers and things like that without having to re-tag.  Nice.


 


Reactionary with Insight.

_uacct = “UA-980395-1″;
urchinTracker();

Web Site Analytics as a means to Target Behavior on the Web and Drive Opportunities

We all know that a practical use of web site analytics like Web Trends, Omniture or Web Side Story is to track behavior on your web site so that you can determine ways in which to improve navigation pathways, content placement and positioning that will increase desired results (e.g., purchase habits, spend per purchase, registrations, applications, etc.). 


 


But sophisticated marketers also integrate web site analytics with eCRM to recognize individuals, track behavior and database preferences so that knowledge gained can be used for future messaging, product placement and site-side advertising.  Content management tools can be leveraged to push specific products at a user who is tracked by web site analytics and identified by eCRM to elevate the chances of putting the right product or message in front of the right person at the right time.


 


Don’t shiver, this is all in common practice today.


 


Now what about outside of the immediate domain of a marketer?  What about on the web?  Here is where things normally get bland with anonymity and event-based targeting.


 


But site-side analytics providers like Web Trends and Web Side Story use first party cookies, the cookies of the marketers they work with.  So when someone lands on the web page of the marketer – say Disney – Web Side Story drops a Disney cookie rather than a Web Side Story cookie.  This is exciting because it means that the data asset is readily available to Disney in a format that they can easily access and extend across other aspects of their infrastructure.  They append the cookie written by Web Side Story with other information that is useful to them by, say their eCRM system.  If the site side analytics data was wrapped up in a third party cookie like an Ominiture cookie, than the marketer would have all their data wrapped up in the third party cookie and would have to conduct synchronizations to get access to the data for other purposes outside of site analytics.


 


Either way, let’s think about DirectServe – first part ad serving.  Is it possible to marry the information generated by a first party ad serving platform with a site side analytics platform and what would there be to gain?


 


Trueffect’s DirectServe platform serves ads from clients’ domains.  So cookies written and read are the advertisers’ cookies.  When someone is encountered that has the Disney cookie (following the earlier example), they can be identified based on the contents of that cookie and be targeted accordingly.  If the segmentation methodology of the cookie data dictates a specific type of banner storyboard, that message methodology will be followed.  If the individual encountered does not have the Disney cookie – which was incidentally set by Web Side Story – than TruEffect will set the cookie.


 


Back up, what am I talking about?  The site analytics software is setting the first party Disney cookie in this example.  Well, Disney can also set the cookie too.  So Disney can append the cookie to include customer segments that would represent target profiles for ad serving – things like family traveler frequency or store shopper frequency.  The cookie file will be interchangeable between Disney’s other systems and the site analytics software because the site analytics software is working in the first party domain.


 


When a new cookie is set by TruEffect using DirectServe, all of the acquisition marketing data is associated with that cookie.  So all the ads seen and all the sites those ads were seen on are associated with that cookie.  When the individual clicks on an ad and goes to the advertiser’s site, they anonymously carry with them the acquisition marketing data.


 


Let’s go back to the concept of integrating a DirectServe ad serving implementation with a site analytics platform.  When someone comes through to a site with a first party cookie that has been written by TruEffect, that cookie will be readable by the advertiser’s site analytics software.  All the data, including how that person arrived at the site will be associated with that anonymous individual and can be aggregated together with the rest of the data that gets collected by the site analytics software.


 


Go the next step.  If that user registers or somehow becomes a record in the eCRM system, then all the knowledge about how that customer was acquired becomes associated with that record.


 


Think about all of my postings on re-targeting now.  The advertiser can use customer segment models to write cookies to identify known individuals based on preferences.  TruEffect can recognize and distinguish those individuals from unknown individuals based on those cookies through an online advertising campaign anywhere on the internet in real-time.  Then the advertiser can message to each preference type specifically and re-target customers while simultaneously prospecting unknown individuals.  Click-thrus will be diverted to the appropriate destination pages like any other campaign and existing customers will be driven by way of a content management system to the ideal transaction-promoting section of the web site.


 


Site side analytics will capture all of the activity.  eCRM will record all customer behavior and build an expanding record base that includes anonymous customer acquisition profiles.  eCRM will be leveraged to create customer target profiles for DirectServe customer re-targeting and content management software will be used to optimize the delivery of product and content placements that optimize desired transactions to the most appropriate individuals to achieve maximum ROI and lowest CPA.


 


It’s all in there.  Patented into a nice little package. 

_uacct = “UA-980395-1″;
urchinTracker();