BEHAVIOURAL
TRUEFFECT COO - SCOTT NELSON
Scott Nelson, COO of the sugar free, diet, low calorie, cookie-less TruEffect, welcome to Goallover. You must be thrilled that someone outside of Colorado has actually heard of you. But seriously, it’s a pleasure to have the chance to discuss your non-cookie adserver solution, and with that in mind let’s get on with a little friendly queue and hey! Tell me a little about the background of TruEffect Mr COO
Scott/TruEffect : TruEffect was originally formed by a group of people from MatchLogic, a seminal ad server that was acquired by Excite@Home in 1998. The team and intellectual property acquired from MatchLogic were the genesis of TruEffect in 2002. We began a market evaluation and development of the technologies that set us apart from other ad serving, targeting and measurement companies serving clients since 2003. Last year, we launched beta clients on the technology platform that is the power behind our Relationship Advertising approach, preparing us for the broader launch this year.
Goallover : So you were involved in the acquisition from Excite and the launch of TruEffect itself?
Scott/TruEffect : Yes. Ron Hill and I were the two founders of TruEffect along with a small but critical group of folks that are still with us today.
Goallover : When did you realise that the old style dataset model was “broken”?
Scott/TruEffect : Actually, we saw the writing on the wall around 2000 when the promise of anonymous profiling was still quite exciting. We were involved in the burdensome process of enticing consumers to volunteer information about their interests and characteristics so we could then re-use that data for targeting. The process of driving traffic and gathering accurate, meaningful information is expensive, and we struggled to see the lift sufficient to overcome the cost. Since then, there has been genuine progress, particularly for the publishers and ad networks who sell media, but very little innovation for the buyers of media. It was then that we made a strategic decision to take a different approach and not develop a third party data model, but rather allow advertisers to deploy our platform using their data, which has proven to be a very good decision.
Goallover : Ok, so you decided to build a non-cookie based adserver model. Why was that?
Scott/TruEffect : We had been researching how best to deploy our data-agnostic technology when a large federal government agency contacted us. They were hamstrung because of policy restrictions on the use of cookies to track consumers. They couldn’t leverage the efficiencies of web-wide trafficking, campaign optimization and reporting that is made possible by using an ad server. The costs of people, errors, and media inefficiencies were simply becoming too high, so they issued an RFP for an ad serving solution that did not use cookies in any way. We won the business with our SafeServe™ Technology, which is truly cookieless.
Goallover : And then you added the “client side” cookie setting system?
Scott/TruEffect : Exactly. As I mentioned, we elected not to get into the data business. There are so many companies with data competencies, even within a single advertiser’s ecosystem, that we decided we’d let them continue to do the heavy lifting. Advertisers are smart and they’ve been investing in data solutions across their enterprise for decades. How could we beat that? The best data for targeting messages to a customer or prospect is the data the advertiser already has about them. eCRM, data warehouses, eCommerce platforms, content optimization solutions, and a host of others have rich data about the relationship a brand has with a consumer. Our position is that data will increase response to advertising better than anything else.
Goallover : From conversations I have had with many advertisers, they wish their eCRM and data warehouses capabilities were as good as you suggest! Many tell me their systems are outdated and suffer from significant legacy issues, but let’s move on. You’re telling is that your approach is one of setting “first party” cookies. Can you more clearly define what you mean by “first party”, second party, and third parties?
Scott/TruEffect : The technical denotation defined by browser protocols essentially considers a “first party” as the domain of the website the browser has loaded. If you’re on www.espn.com, ESPN is the first party and any other domain serving content into the page is a third party. Connotatively, the industry has used the term “first party” to describe the entity with whom the consumer thinks they’re interacting. Most of the time, the two approaches would provide the same answer, but our approach turns that on its head. Until TruEffect’s DirectServe™ technology was deployed, when a consumer sees an ad, say for Dell, it is almost guaranteed that banner is not being delivered from the Dell site, but rather from a third party ad network or ad server unfamiliar to the consumer. Now, when a consumer sees a Dell ad served by TruEffect, it would be seen coming from a Dell domain and all data associated with that serve would be stored discretely for Dell, the entity with whom the consumer associates the ad. Essentially, we turn a banner impression into a miniature version of a brand’s website. It is the most transparent model available.
Goallover : I am confused. Browser protocol denotations mean that the first party is the website publisher, ESPN in this example when a user is visiting ESPN.com – and presumably the second party is the visitor’s browser, is that correct?
Scott/TruEffect : Yes, although the phrase “second party” is not commonly used in place of the term “browser”.
Goallover : So does that make the advertiser, Dell in your example, the third party? Are they the third party, or is the advertiser’s adserver the third party, or the advertiser’s agency WPP? Or are they all “third parties”? Do we have a situation in which a number of third parties are muscling in on the party?
Scott/TruEffect : It seems there are enough parties to go around! You have come to see how the denotation in browser-speak can confuse the connotative reference to “first party.” Prior to our technology, one of the rare instances of an advertiser’s display ad being delivered to the browser from the advertiser’s domain on a site other than the advertiser’s primary website has been in the case of a large publisher advertising themselves - e.g. Yahoo – on another site. Advertisers, including Dell, choose not to build the infrastructure to serve ads across the web from their web domain for good reason – cost and complexity. Their display media has been delivered through the domain of the third party ad server because that’s been the only option. If Dell bought ad inventory on Yahoo.com, and TruEffect served the ad on their behalf from the Dell.com domain, Dell would indeed be a third party to the browser because Dell.com is not Yahoo.com. To the consumer, however, Dell’s display ad is coming from Dell.com, and right or wrong, the phrase “first party ad serving” was coined to reflect that. TruEffect, for its part, delivers the media from our global, redundant ad serving farms, relieving Dell of that burden. We’re not “muscling in” because we cannot repurpose the data nor aggregate it across clients - Dell’s identifiers differ from any other advertiser. In the cookie space, every domain speaks a different language.
Goallover : OK, that’s clearer. So, is TruEffect perhaps then a fourth party, metaphorically a “corner man” for the advertiser – empowering the advertiser to dominate the third party arena
Scott/TruEffect : We’re certainly in the advertiser’s corner, but we’re not a “fourth party.” We’re not adding another player to the mix which would be how the industry has characterized it with phrases such as “fourth party re-directs” which are forbidden on large web portals. Publishers however, have viewed this as a means to help optimize the media performance for their advertiser clients, so I wouldn’t take the boxing analogy too far.
Goallover: And if this is the case, are you saying that TruEffect can basically ensure that Dell (for example) can ensure that Dell does not enable other third parties access to their data.
Scott/TruEffect : Yes. One of the strongest motivators for advertisers has been the ability to control the data resulting from their media activity. This data has traditionally leaked out of the organization and enhanced the data set of the third parties and media companies. The media consumption, response and conversion data of their customers and prospects created by serving ads is extremely valuable to marketers, and they dislike allowing others to benefit gratis, particularly when it can be used to the benefit of a competitor.
Goallover : Yes, that is totally clear, and I can well understand why Dell may not want their information assisting IBM, which I can see happening in the event that both Dell and IBM deploy a third party product which aggregates campaign knowledge, though conversely of course Dell may benefit from the knowledge gleaned by the third party from IBM.
From an earlier conversation we had, i understand that your depositions to the FTC were arguably responsible for their realisation that different cookie-setting models were actually viable? How do you think that has shaped their position on behavioural targeting to date?
Scott/TruEffect : We submitted our comments to the FTC pursuant to their call before and after the eHavioral Advertising Town Hall in November, 2007. We also presented at that event in Washington, D.C., declaring the fall of the third party data model. The FTC has been very interested in the novel approach of a first party data model and used the information we submitted as background for their Staff Report on Self-Regulatory Principles for Online Behavioral Advertising.
Goallover : So are you saying that, in your opinion, TruEffect is the champion of the client and consumer, while other adservers are really just champions of their own volition?
Scott/TruEffect : Yes, with one caveat. The third party model has been a force behind the explosive growth of the internet by enabling ad revenue to fund the fantastic content and cool services we have come to love. That said, we’ve brought to market an alternative that defines the next chapter in this still very young medium. Consumers have relationships, and often deep ones, with merchants they know and trust. Not only do they exchange information with them, they expect the merchant to remember them and treat them differently because of that relationship. It’s no different online, but the advertisers relied upon some very smart and agile entrepreneurs who saw an opening and built an entire industry by developing a proprietary data model. We’re proving that the data defining the relationship between a brand and its customers/prospects is the single best data to use when communicating to them anywhere on the web.
Goallover : In simpler terms, you are from the “silo” school of adserving then, where each advertiser has control over their own data, and it is not aggregated by a third party that is hidden from the consumer?
Scott/TruEffect : Correct. Until now advertisers thought they’d have to build their own ad serving infrastructure to use their data to drive web-wide ad campaigns, which is so daunting that even the largest of them elected not to do that. We’ve eliminated that barrier elegantly. Let me share an example of consumer expectations with regard to the data they share.
Offline, consumers gladly share information with a clothing store about their tastes, measurements, even about the reason for the new clothing purchase such as an important event in their lives. They fully expect the clothier to remember that information so their next visit can be personalized. They do not expect the salesperson they thought worked for the clothier, but was actually a third party, to run down the street to the shoe store and share this information so they can be prepared with coordinated footwear when you walk in. We call it the “cool-to-creepy” factor.
Goallover : But then again, if the advertiser wants only to buy adverts that are shown to prospects who have subsequently visited their site, they must share this information with their suppliers, or they cannot make such a re-targeted buy. On that note I have two questions. Firstly, can TruEffect empower an advertiser with a “container pixel” which can safely enable the dropping of a third party supplier or aggregator pixel?
Scott/TruEffect : Yes – our TruTag is a sophisticated form of container pixel that does offer control to the advertiser over the myriad number of entities that ask to track consumer visits to their site.
Goallover : Ok, so you have the portfolio of advertiser-side pixels which can define visitor, prospect, suspect et cetera. And your solution can drop the relevant tracking pixels on behalf of the advertiser’s chosen tracking partners.
So secondly, if the goal is that an advertiser does not drop third party pixels on customers to enable retargeting, how does TruEffect advise its clients to transact when they want only to target their prospects, and de-target suspects?
Scott/TruEffect : We are firmly in the camp of working with site publishers and ad networks to improve performance of our clients’ media buys. They need to make a decision based on the data they have, which can be very rich, about the best impression to offer to the advertiser to fulfil their media goals. Pixeling the client site is often necessary to achieve the best result. That said, when we receive an ad call for a client, our servers can read the client cookie and recognize instantly whether this is a known customer, a prospect or even a suspect. If the client has asked to serve only suspects and prospects, we have a couple of options. 1.) Ask the publisher if they would like the requests for known customers re-directed back to them for a different advertiser who may be willing to pay a higher eCPM for the impression, or 2.) Deliver an ad for a different brand within the family or even a marketing partner. Brands commonly work closely with adjacent products or affinity partners. For the first time, we can enable a brand to dynamically repurpose ad inventory based on the existence or absence of a relationship with the consumer.
Goallover : So are you saying that, at the time of serving an impression for the advertiser and commanding the placement of the advertiser’s cookie, it is not possible to simultaneously command the placement of a cookie for another party – allow me to explain the rationale for my question:
I am thinking of the scenario in which I am an advertiser buying ads from Specific Media, MSN and Yahoo, and I wish to show an advert for a maximum of 15 impressions per unique user across my buy. Thereafter, I don’t want to show any more ads to that user for 30 days.
The question is this – if TruEffect can command the placement of a cookie for an advertiser, can it not simultaneously command a “no more please for 30 days” cookie drop for the Specific Media, MSN and Yahoo ad servers? If it can, then my three suppliers’ all know that I don’t want a unique user to see another of my ads for a period of 30 days?
Forgive the length of this question, but I ask the question because I believe that the biggest issue facing the online ad serving world is one of waste - de-targeting - waste management. While everyone has been chasing the gold, nobody has considered the waste – the inconvenient truth of the wasteful, greenhouse gases of the banner world so to speak! Can the advertiser using TruEffect drop a cookie on behalf of a publisher when we want to say “Don’t give me that user again please”!
Scott/TruEffect : It is important to keep in mind that every ad impression is comprised of two serial decisions: i.) media allocation - of the eligible ad campaigns running on a publisher network, what advertiser would most value this user? And ii.) message allocation - once the media allocation decision has been made, which of the available advertiser messages will resonate most with this user? When you talk of waste using your scenario above, the media allocation decision would be the most appropriate place to “frequency cap” this user, which would reside with the publisher network. TruEffect, on behalf of the advertiser, could redirect the browser back to the publisher ad server if we manage the frequency capping.
Goallover : I’ll take that as a “no” then - I understand that each publisher can be contracted to frequency control their own delivery, though they do not always deliver in full when frequency capping comes into hard focus - but let’s keep this dialogue going and see where it takes us. The buy side retains buy side insights and the sell sides retailing sell side insights, correct?.
Scott/TruEffect : We believe the optimum result for our clients will be a combination of our advertiser-centric approach together with sell-side targeting when there is little or no relationship. The beauty of the solution is that it allows the media seller to optimize the media selection decision in series with our selection decision based on the client data.
Goallover : So do we end up in a situation whereby TruEffect clients may still need to collaborate with the third party advertisers, or network aggregators, to deliver optimal effectiveness?
Scott/TruEffect : The best results will come from a coordinated effort between the direct seller of the media impression and TruEffect on behalf of the advertiser.
Goallover : To put the question another way, are we saying that although TruEffect enables the advertiser to gain greater ownership of the conversation – dialogue - engagement with the online user, the advertiser may need to cede back some control back to the publisher, or the ad network, in order to maximise effective deployment of inventory?
Scott/TruEffect : I wouldn’t characterize it as “ceding control” as the media seller does in fact own and have claim to the ad inventory to monetize anyway she sees fit. Within the context of a single transaction, cooperation using both technology platforms to optimize decisioning is ideal. Within the broader context of buying strategy and resource allocation, however, TruEffect provides exceptional insights into the behaviour of an advertiser’s customers and prospects across sites and networks using metrics they’ve never had before. How does this network perform with respect to penetration of specific product SKUs across the buy? How does that site perform with respect to penetration of my customers by shopping segment such as “New to Store” or “Seasonal”?
Goallover : Ok, let’s park that dialogue for a while and discuss the difference between the TruEffect approach and the WPP approach to data. It was recently reported that WPP has implemented new terms and conditions which stipulated that all data would be the sole property of WPP. To that end, TruEffect appears to be violently disagreeing with the WPP approach. Can you expand on this issue for us – and please be candid, we’re all grown-ups here.
Scott/TruEffect : Data is a powerful asset and every company strives to gather and control data to be competitive. There is value to consolidation that agencies leverage every day for their clients. WPP has a rich client roster that deserves to benefit from the scale, reach and perspective WPP offers – seriously. But, if they were to develop a third-party WPP-centric data model, it will not deliver as well as an orchestrated brand-centric model that could be managed by WPP on their clients’ behalf.
Goallover : So are you suggesting that WPP is simply trying to contract its way into the position of data aggregator, or third party, and how do you think this impacts on the FTC guidelines?
Scott/TruEffect : I do not know what WPP is trying to do. I do not have any information about how they would deploy technically which would be important to interpret adherence with the FTC’s Principles.
Goallover: Methinks you’re being too diplomatic Mr Scott! I think WPP’s acquisition of 247RealMedia was the first step towards a new model in which WPP cookies start being dropped across the web. Perhaps someone finally told WPP that, (for example) Atlas cookies were perhaps serving Atlas clients rather better than they were serving WPP clients, to the detriment of WPP’s proprietary knowledge. In the UK for example, Atlas empowered two companies side by side, one selling technology to WPP, and another one brokering inventory to WPP and other agencies. Rather dumb of the agency to fall for that one, but hey – apparently you can fool some of the people some of the time and WPP didn’t have sufficient expertise at the advertising technology level.
Leaving WPP aside, with respect to the bigger picture, and the entire industry debate, who do you think we are not hearing from and what do you think they have to offer the discussion?
Scott/TruEffect : Advertisers. The brands themselves are conspicuously absent from the dialog. The advertiser community has funded the growth of the internet and has an enormous stake in its success. Over the next decade, failure to embrace the digitally-based engagement model with consumers will be fatal. Consumers know and understand the brands that provide them goods and services. The brands are the best stewards of that relationship, the data surrounding it, the standards, protocols and policies defining it, and the regulations protecting it. Honestly, would adware or spam have ever existed if the brands were engaged in those decisions?
Goallover : Yes, they have been conspicuous in their absence, a very good point, and perhaps a good point on which to close this conversation. Scott, it is really interesting to talk with you. I certainly understand TruEffect’s position and proposition more fully than I did before our conversation, and I am grateful for your time and explanations. I hope our readers learn as much as I did. Thank you very much for speaking with Goallover and telling us more about TruEffect. I wish you every success with your business and look forward to revisiting some of these issues with you as they unfold over the coming years.
Tags: BEHAVIOURAL, cookies, Dell, Excite@Home, first party, FTC, IBM, MatchLogic, Relationship Advertising, Ron Hill, SafeServe, Scott Nelson, third party, TruEffect, WPP Discussion
2 comments for “TRUEFFECT COO - SCOTT NELSON”
All of this first, second and third-party cookie talk reminds me of Groucho’s contract negotiation with Chico in the Marx Brothers classic, A Night at the Opera (http://www.youtube.com/watch?v=C-zR2pM_S5U)…enjoy!
Great insight into the product. We’ve worked with Trueffect’s ad serving solutions, and they’re great - so the cookie-less product will probably be the next generation system.