Image courtesy of Chango

Let’s back up a little bit and start by explaining one of the biggest buzzwords of 2013: big data. “Big data” simply means lots and lots of data that sits in different places, making it hard to use. Even more simply, it means more data than you have used in your marketing before.

And, why should you care? Well, more data makes you smarter, and being smarter allows you to make better decisions about who to target, when to target them, and (in this real-time bidding media world) at what price.

So, you’ve got big data, but it’s hard to use — that’s where programmatic marketing steps in. Programmatic refers to the method and tools that make all that lovely data actionable.

When we started Chango, it was with the early vision to bring programmatic to all marketers. Media exchanges were just beginning to emerge and so we had the access to target the individuals that were relevant.

But what we needed to decide was which data set we could focus on to cut our teeth. We don’t have a history of doing things the easy way, and so the equivalent of throwing ourselves into the deep end was to pick the largest data set we could find.

Search Retargeting Comes Alive

The audience of this site will know more than most that there are few data sets larger than the searches that individuals do on the major engines every day, and for an idea of scale, we have billions and billions of those events – more than Yahoo! and Bing combined every month.

Search data also have a lifetime value; these data can degrade fast, sometimes becoming quite useless within days or hours, depending on the industry you are using it for.

So. that sounded like a fun challenge. :)

When a brand wants to find new customers using search retargeting, they compile a list of keywords that they are interested in and send it over with their display ad creative.

These terms are then held in a “live” system, waiting to match against individuals who have searched on those terms. When it comes to prospecting for brand new customers, it’s hard to beat targeting just the hand-raisers that have said they want what you have!

The technology to make all this happen became known as the PMP, or Programmatic Marketing Platform, and such a system allowed rules to be written that married disparate data sets together to create an output such as relevant advertising.

Sample Executions

The majority of marketers use Search Retargeting for direct response, measuring on a ROI (Return On Investment), CPA (Cost Per Acquisition) or ROAS (Return On Ad Spend) basis. They put media dollars in and expect a certain level of return. It requires intense, hands-on optimization at the keyword level.

As an example, there is a department store that uses search retargeting as an evergreen tactic. They target keywords such as “new shoes,” “dress sale” and “clothing coupon” to find new customers that are in the market right now. They also target brand keywords that relate to the clothing labels they carry in store to capture the dollars from individuals who don’t know they are a source for those items. And lastly, they use competitor brand terms for conquesting new dollars.

For an example in travel, think about someone like a rental car company. When better to target someone than when they are searching for information about their upcoming destination, their flights or perhaps their hotel?

Beyond Direct Response

More recently, marketers are starting to recognize the value of programmatic marketing for generating “brand awareness,” a typically expensive and often vague methodology. They are understanding that such an approach can remove that wastage and even make their dollars more efficient.

Consumer Packaged Goods (CPG) brands are among the first to adopt programmatic marketing, with companies like Kellogg’s being the poster child in the industry; they are using all this data to define more accurate audiences.

Imagine a new bio-degradable diaper product. Today, the manufacturer is looking for customers based on criteria such as “females, 2 kids in the house, moderate household income.” This set of criteria is okay, but very wasteful, as not all these individuals care enough about climate change to spend the extra dollars on the product.

Tomorrow, these brands will leverage intent signals such as “cares about environment,” “expecting a baby,” and “reads about BPA in children’s toys” — suddenly, they are spending less and finding the right people both more quickly and more cheaply than before.

Programmatic Everywhere

Image courtesy of Chango

It’s the future that really excites me, though. We spend a lot of time talking about display ads, mobile and video; and with our integration with FBX (and now Facebook’s Newsfeed for search data), we even talk about social.

But what about out-of-home (OOH) advertising, TV, radio, etc.? And what about the future we see coming from augmented reality, Google Glass and who knows that else? All of these environments will be desirable to marketers, and the idea of the “right place” will become very complicated indeed.

Think about that – you want to target a mum who cares about environmental issues with a biodegradable diaper – when do you do it? On her way to work through her Glass, on her digital radio at her desk, on the billboard as she pulls into the supermarket parking lot, outside her kid’s school whilst waiting to collect them…?

Such decisions require massive data and will require programmatic systems that can find it, process it and make it actionable. Our guess at the revolution that was to come appears to have been right. For kids of the ’80s: “I love it when a plan comes together!”

Sounds simple, right? Actually, it’s very far from simple, and is actually the wrong goal to be aiming for.

Firstly, most of the barriers to building an attribution solution are extremely hard to overcome, and others are simply impossible to accomplish. Technical limitations mean we can’t tie together all of our digital touch-points, human limitations mean we can’t get everyone to agree to how the model should work, and simple life realities (such as real face-to-face communications) mean we can’t account for every day-to-day interaction.

Programmatic Attribution Modeling is a solution that moves us beyond these barriers, and focuses us back on the idea of taking action based on data, not just making sexy graphs. But first, let’s understand why this fresh approach is really needed.

Technical Limitations

Where do we begin?! To make headway with an attribution model, we must be able to account for as many touch-points as we can, with at least a reasonable sample size of the targeted population.

Some of these are simple. It is not uncommon to bring together all display buys under one cookie by using a buying platform for real-time media, and that same cookie can be served on premium/direct display buys, too. This could be your ad server, or it could be a demand-side platform (DSP), a data-management platform (DMP), or better still, a PMP (Programmatic Marketing Platform) that combines all these technologies.

And, that cookie can also be matched against the clickers from your SEM program, with the same tools, or with on-site analytics. Tag container companies like BrightTag, TagMan, Tealium and others can help further, and some take you to a deeper level still by helping to bridge more vendors that might be on your plans.

The next step would require pure, sneaky “smarts.” Using Pinterest or Facebook in your marketing? Use your PMP to look for incoming traffic (and cookies) from those sites and add that individual’s interaction to their universal profile, knowing they have pinned or friended your brand.

At this stage, we are still missing everyone’s offline exposure to the dataset, as well as interactions with real-life influencers. In the future, perhaps the presence of RFID chips in phones (and in people ) will help overcome that; but, until it becomes widespread and measurable, we need to settle for broad, geographical trends based on the investment of advertising at the DMA level to get a partial solution for this piece.

Human Limitations

Even though there’s been progress made on the technical side, your organization must agree on what to do with that data. Typically, CMOs find this a bigger barrier than lining up the technical aspects described above.

Imagine the meeting to determine an attribution model between the CMO and the leads for Display, Analytics, SEO, PPC and Social. Display typically buys on a CPM and measures with view-through; PPC buys and measures on a CPC; SEO and Social are both producing content and monitoring their effectiveness with clicks and organic visitor growth; and, Analytics can only monitor on-site behavior and incoming click traffic, ignoring influence from exposure elsewhere. And, let’s hope no one invites the Traditional media buyer, too!

Each party’s motivations differ from the others’. Some favor a last-click situation getting credit for what brought each conversion over the line; others spend their time further up the funnel driving demand, and as such, want first touch to be the deciding factor. And, channels like Social play throughout the process and argue that organic, earned media always outweighs the value of paid media and should, therefore, be given more credit.

Get Something Done!

It’s an impossible scenario for the CMO to solve, but the worst thing to do is to do nothing. I have seen many models used over the years, and the one that seems to have the fewest limitations is Forrester’s (PDF) model, shown below.

In such a model, each touch point has value — the one further back from the time of conversion gets the least credit, the one closest gets the most. It also has the added value of being from an independent third party; and, if presented to the Impossible Group of your heads of channels, it’s hard for them to propose a much better alternative that considers each other’s programs.

There are companies that will help you with such things, too. At Chango, we have seen our clients have success with the likes of Adometry and ClearSaleing; just don’t expect them to be able to work magic if you haven’t started solving the technical problems.

But It’s Still The Wrong Point!

If an attribution model is there to help us make smarter decisions, then let’s create data that result in immediate action. The outcome of all of this hard work will tell you which channel should get more budget and which should get less, but the experience for each individual won’t change, and neither will the investment in your most valuable prospects.

Worrying about an exact division between channels shouldn’t be the focus – instead, concentrate on the now, and concentrate on the exact individuals you are trying to influence.

Programmatic Attribution Modeling is another benefit of the big data movement (“big data” simply being lots more data than you use now, and “programmatic” being the mechanism to make it actionable). In such a case, the media buy itself is programmatic, meaning that at the exact point the decision to buy is made, all the most recent data is queried and an assessment made as to whether that person needs another touch point — and, if they do, what format should it be and with what creative execution?

Programmatic Attribution Modeling only works if your buying platform is processing all of the above types of data; and if it is, then you can execute without the typical wastage, building an attribution model of sorts for every individual, not just the broad population.

Today, the executions are typically display media, with some mobile and tablet thrown in. But, PPC is around the corner and billboards, radio, TV and the like will follow suit in the future.

Go Away And Do Something

Now, it’s your turn. Go look at what you are doing today. Is every channel optimized to individuals, or to broad decisions about media planning, PPC budgets and your latest social campaign? Are there areas where you can make a more precise decision about an individual today, rather than waiting for a population-level model in the future? Cross-channel is cool; but as marketers, we are speaking to people, not spreadsheets.

More spend is better? Garbage! Smarter spending is better.

Over the last two years, I have discussed why marketers should cut their site retargeting budget, how programmatic marketing executions can change retargeting forever and, in a recent iMedia article, why retargeting is fundamentally broken.

But, these are just my opinions; and so, the team at Chango refreshed the Retargeting Barometer, asking more than 50 marketers and agencies what they really think of retargeting today.

Buying Retargeting Directly

The first surprise for me was the rate at which marketers are moving from buying retargeting through their agency to buying it directly from vendors (58%).

About 7 years ago, I was running the European arm of a major ad server, and we were just beginning to see the same type of pattern. Our business was 80%+ agencies, but the incoming leads were nearly all brands — and now the majority of ad server accounts are held by the brands.

The reason? Marketers want control of their own data, and they also want consistency in these core programs, even if they part ways with their agency.

That’s troubling news for agencies, who are already struggling to find new ways to prove their value in this rapidly changing ecosystem. But, as I often help agencies to understand, all is not lost for them — many marketers still need the execution and creative services that agencies bring. If agencies begin to realize that clients want smarter buying, not “more” buying, there are other ways for them to win, too (as outlined in the iMedia article above).

Email Retargeting: The Great, Big Missed Opportunity

At industry events, I will often ask the audience how many are using email retargeting — and I’ve yet to ever see a room where the response is more than 10%!

Email retargeting is simple: someone receives your email, is cookied and added to your retargeting pool — this allows the marketer to retarget them with display ads just as they do with site retargeting.

If you tag your emails correctly, you can show a custom display ad to each individual depending on what email they interacted with. Given that there is no additional cost or technology involved with this tactic, it never ceases to amaze me how infrequently it is used.

Search Retargeting Has Well & Truly Arrived

For clarity, search retargeting is the idea that an individual has performed a search on Google, Yahoo or Bing, but not yet visited your site. They are expressing the right intent; and so, we target them on behalf of clients with a display ad to win their business.

Of the Barometer’s respondents, 45% reported using search retargeting now — with 92% saying it was increasing revenue and 68% reporting greater on-site engagement. What was interesting was that a whopping 50% are also using it for brand awareness. Given that there are no restrictions on using competitor terms, 58% are sneakily pilfering directly from their competitors.

Budgets Are Getting Bigger – But That Worries Me

Not a single respondent reported that their budget for site retargeting or search retargeting would be decreasing over the next 12 months. I should be delighted; but actually, this worries me.

I worry that, with site retargeting, in particular, marketers are going to increase their budgets for all the wrong reasons — because the wool is being pulled over their eyes about what site retargeting is actually doing — rather than because they are using a proper scientific measure to look at the incremental value. We shall see where the industry shakes out on this one.

Let’s Welcome FBX To The Party

Since their launch late last year, FBX have done a great job of making noise (supported greatly by the 13 of us that were chosen as early, named PMD partners), and marketers have certainly listened.

For the first few months, it was the retargeting companies that dominated the space — given that their contracts with brands often allowed them free reign over where they placed ads and what creative ad units they used — it was a no-brainer, really.

This is at least partly responsible for all but one marketer surveyed reporting that their FBX budget would be staying the same, or increasing, over the year to come; but, I bet with the upcoming launch of newsfeed content being added to FBX, Chango’s next refresh of the Barometer will show an even higher percentage increasing their spend.

To Wrap Up…

Retargeting is big. Retargeting has driven a lot of display spending that otherwise would not have occurred. Retargeting is more than just site retargeting — it is also search retargeting and email retargeting. It is being used primarily for customer acquisition, but also for boosting branding and awareness.

Because of all of this, budgets are increasing, and at a quicker rate than before. But, have caution: don’t just spend more for the sake of spending more, spend smarter instead.

Using more data can be difficult to do, and so, the way to use more data is with an approach called ‘programmatic.’ Removing all the complexity and BS from the term, at its highest level, ‘programmatic’ just means that you are writing some logical rules that combine bits of that data together.

Simple, right?

It Means You Don’t Have To Guess Anymore

The rise of media exchanges and RTB (real-time bidding) means that as marketers we no longer need to simply shout at crowds, but instead can talk directly to individuals.

It used to be common for a marketer to brief an agency or publisher about their audience by describing them with attributes such as age, gender, household income (HHI), whether they have kids, etc., and then go find sites that felt they had a similar audience.

If you wanted males between the ages of 18 and 25, you would probably call GQ or ESPN; if your need is moms with new babies, then probably iVilliage or Parenting.com; or if you wanted financial types, then Motley Fool or Yahoo Finance. Each of these would be perfectly reasonable sites to include on a media plan, but each one inherently includes wastage that drives down the efficiency of your plan. GQ, for instance, will also have some female visitors; iVillage, some dads, and so on.

With a site buy, marketers are guessing that they will be getting their ad in front of the right person.

But then, the guessing was reduced significantly. The site became less of a focus, and the individual became what mattered. Instead of hoping to find males between the ages of 18 and 25 on your selected publishers, marketers can now cherry pick who matches the criteria by using their own data, or 3^rd-party data.

Not Getting Carried Away

This is a great shift, and RTB has changed marketing forever. I accept, though ,that marketers are on somewhat of a data high; they feel cookies and algorithms are all that matters today.

That is actually very wrong, and thankfully for the sake of us all, it turns out the context of the placement remains important, as does the creative message you choose to show (webinar recording: ‘Why The Creative Still Matters in Programmatic Marketing‘).

Real Executions With Programmatic

There are many cases where the use of big data and a programmatic approach are making a real difference already.

A.  Programmatic Site Retargeting

Know it or not, the most dollars being spent today using data are with site retargeting – in this case, the data is that of an individual visiting a site. As we also reviewed in the last article, though, most marketers don’t understand all the data they have, and are not using it in their decisioning, meaning they are not being smart.

Dedicated site retargeting companies are incentivized to make you spend more, not to ‘spend the right amount.’ They are not interested in encouraging the use of big data because it will tell you that many people shouldn’t be in your cookie pool.

As an example, if an individual is visiting a luxury car brand, but is also searching for ‘bad debt loans’ on Google, why retarget them?

If an individual is looking at the careers page of a retailer, they want a job, not your products, and so why retarget them, given they aren’t looking to buy?

If they are your fan on Facebook, why use retargeting impressions to talk to them, too – or, given they are a fan, they must be super-engaged, so perhaps double your investment in them.

This is big data, or more data, in action.

B. Search Retargeting

Another great example is search retargeting. In order to process billions of keywords every month, we construct rules that help determine which individuals are relevant and what value they have to an advertiser.

Programmatic rules might help us determine that someone searching for ‘vacation,’ ‘champagne delivery’ and ‘burberry’ are likely to spend more on a travel site than someone searching for ‘vacation’ and then browsing sites for credit cards for people with bad debt.

Hopefully you are seeing the trend here, more data makes you smarter, and means you can cut out the wastage from what you are doing.

C. Programmatic Look-A-Like Targeting

The same applies for common techniques such as look-a-like modeling. To understand the difference, think of search retargeting and programmatic site retargeting as ‘reactionary’ – someone does something, you respond – whereas look-a-like is predictive, you are using lots of data to decide who else will respond well to the message a marketer wants to show.

As an example, you could take a site like Bed Bath & Beyond. If an individual buys a lot of basics in one go and uses a ‘new mover’ voucher, then we know they just moved into a house and are a desirable target customer.

We could then look at everything else we know about that person (gender, search terms, location, age etc.) and find more people who share those attributes. Those people who overlap are very likely to respond well to the Bed Bath & Beyond messaging.

There is so much data that can be collected about individuals that a programmatic approach is being used to crunch it all and work out what makes the most sense.

And So…

Programmatic and Big Data might be new buzz terms within this industry, but many smart tactics are using them already. If you are buying display today, chances are some of the tactics you have live use this way of thinking.

But importantly, know that they highlight smarter spending, not just more ways to spend more. And so, be sure your partners are incentivized to work with you in that way.

But, there is also the more invisible footprints you leave as you go about your day swiping your credit card, driving through tolls, visiting websites, etc.. So, when we talk about Big Data, we really mean it.

Like many buzz terms, Big Data and Programmatic Marketing are actually nothing new, they are old ideas brought up to date to the modern need; and so, if you really want to succeed in digital marketing, you must understand their roots and the core promises they make.

You must also look to the future and understand the vision we have for the programmatic approach; you might be surprised what can be done.

Big Data Is Not A New Marketing Tool

Let’s start by looking at ‘big data’ in its simplest form. Big data is just lots of data! Many companies have lots of data at their disposal, and have for generations.

Think of traditional catalogue companies that have massive quantities of CRM data relating to every household in the country, or a supermarket brand with a loyalty card program that knows everything about your shopping habits, or even a credit card company that knows everything about your life and habits through your spend data — all these companies have had ‘big data’ for a long time.

So, to be a little more precise, think of big data in today’s world as being lots of data, but lots of data that is dispersed across multiple systems, and can therefore be cumbersome to make actionable in a meaningful way.

If you are a catalogue company, then dispersed data doesn’t represent too much of a problem; you can run a query against your various databases to extract the information you need, and if that takes 24 hours to process, no big deal. But, in the digital world where we are trying to make decisions in real-time, this cumbersome nature represents a real problem.

Programmatic Marketing

Therefore, the promise of programmatic marketing is to bind dispersed data together and make it actionable in a real-time, digital world. The idea is that these multiple systems are brought together through one technology, and then, rules can be written to make decisions and actions based on all that data. Think of a simple Boolean logic query that would say something like:

Show ad 1 to (active customers who have NOT bought in the last 30 days AND live in California AND are male)

Realize it or not, that is how programmatic marketing works. Now, work would have to be done on the backend to make the databases that contain that information accessible, but can you imagine the power placed into the hands of a marketer who could see 2.5 quintillion bytes of data about their audience!

Executions

Such systems are not straightforward to build, but if done right, are simple to use. The industry started with the idea of DSPs (Demand Side Platforms) for buying in real-time, and then added DMPs (Data Management Platforms) to start trying to bring some of this data together. The problem remains that such systems leak huge amounts of valuable data and force buying decisions to be made on data that is out of date.

There is a new movement of the PMP (Programmatic Marketing Platform) which we coined within the industry and relates to systems specifically designed to capitalize on the opportunities of big data and programmatic marketing in today’s market, and in the future.

This approach is already starting to change the way common marketing techniques are bought. Take site retargeting, for instance. Nearly every marketer is doing it, and while I feel strongly that it is being measured incorrectly, it is producing results.

But even with advanced site retargeting, the marketer is ignoring the true value of each individual. Sure, they might be serving them a dynamic ad with the actual product they looked at within it, but does that person really have any likelihood of converting? With big data, a site retargeting campaign can be driven by more data, and more data makes us smarter marketers.

If you are a luxury car brand, do you really want to retarget everyone who looked at your sports cars? Really? Even if at 2 a.m., they were searching for bad debt credit cards? How about saving those impressions and investing them in something else?

Or, if you are a home supply store, and someone arrives on your site from Google having searched for ‘granite counter tops’ versus the person arriving searching for ‘how to hang wallpaper.’ Which one do you think you should invest in? Big data and a programmatic framework will tell you – and then make you smarter and more efficient. Site retargeting becomes ‘Programmatic Site Retargeting.’

These types of executions are reactionary – an individual does something, and we do something in return. Big data allows us to be predictive, too, though. We can analyze the data about individuals and predict what they may want in the future. We can even analyze our existing customers and based on shared behaviors, guess who else in the population might respond well to our messages because they are connected somehow. Think of it as a ‘programmatic look-a-like’ capability.

A Programmatic Future

Looking to the future of marketing has never been more fun. The very beliefs of a PMP are that data come from many sources, and the executions that can occur from that data will be as varied. Already today, there are companies testing ideas such as digital billboards with facial recognition, or taxicabs with digital ads that change based on the geo location.

Programmatic might be the buzz term that is new within display advertising, but its implications are far more broad.

Click here to view the embedded video.

(Video: Face-recognition comes to billboards)

Imagine as a marketer, being able to know when one of your customers is in a taxicab, and whether they opened your latest email before they got in it, whether they have bought from you recently and whether they are due to go on vacation soon. And from that data, being able to decide the value of showing them your message at that precise moment. It’s almost ‘individual marketing’!

What Does This Mean For Today’s Marketer?

We have been pushing hard as marketers to reach the magical place of always talking to the right person at the right time and with the right message, and that shouldn’t change. But big data and programmatic marketing take us into a new area – ‘right price.’

In a real-time media world, understanding how much to pay to talk to someone is as critical as knowing which person to talk to in the first place, and now you can.

What Is The Facebook Exchange (FBX)?

In some ways, FBX is another media exchange like the GDN, AdMeld, RightMedia etc., an open marketplace where display media ads are bought and sold using RTB (Real Time Bidding).

As opposed to many different domains being aggregated together in one place, the FBX offers inventory from Facebook only.

Unlike Facebook Marketplace ads that are bought on a CPC basis and utilize Facebook user profile data, FBX is bought on a dynamic CPM basis, and the buyer must bring their own data to the decision rather than use Facebook data.

Search marketers will be very familiar with the auction process as it is a second price environment, just like AdWords.

Where Is The Money Coming From Today?

Given that FBX doesn’t offer data for targeting purposes, the first money into the pot has been from the retargeting companies. They can easily use the pixels they have in place for their clients retargeting campaigns and extend that buy onto FBX. And this is happening a lot, but many other companies are following and this will dilute the retargeting dominance in the future.

Is This For Search Marketers Or Display Planners?

Well, in many ways both – think of FBX as an equal opportunity media source.

When Facebook Marketplace ads began, it was the media planners that seized the early opportunity; but, as the buying model shifted from CPM to CPC, search marketers stepped up and stole the show.

Search marketers built the better tools and had the more relevant experience and quickly put their display colleagues to shame. I was running a global display media team when this happened, and I have to reluctantly admit I saw the budget disappear from under my nose before I realized it!

The difference with FBX, though, is that you need to access the inventory through a DSP, or a company that uses DSP technology. We have discussed previously in this column about how a real-time CPM buy seems to scare off the search marketers, and so perhaps FBX is the media buyers revenge?

Search Retargeting On FBX

When we created the idea of search retargeting at Chango, it was the first time that display media and search marketing truly overlapped. Up until then, we, as an industry, were guilty of propagating the idea that somehow search + display was this magical 1+1=3 model, without really ever having more than anecdotal data.

Search retargeting, though, took that wonderful intent signal and combined it with the scale and affordability of display, and finally delivered on the promise. Now, as one of the ad exchange partners, our search data can be paired with Facebook. For us media and data geeks, that’s more than a little cool!

For the search marketer, that means that perhaps they can win on this battleground, and make the case that it is search data providing the smarts, and FBX, the reach. I certainly think that can be the case; but, given that most search retargeting is still bought as media and not search, I do wonder.

Does FBX Matter? Could I Not Just Ignore It For Now?!

Nope, sorry! Unlike the introduction of just another media exchange, FBX potentially adds 25% to the real-time media available, and results are showing that consumers are converting quicker, with fewer impressions, and at a considerably lower CPM than elsewhere. It kind of has everything going for it right now.

There are a number of ways to take advantage of it. I would advise marketers to start in two ways. First, look at your site retargeting programs and extend them on to FBX – with the media price alone, you will increase the efficiency. And secondly, add prospecting to the mix by excluding your existing site visitors from the buy, and finding new individuals using search data.

Search and social in one place? Now that’s a huge opportunity.

Image via Shutterstock

2012 has been another big year for digital. And in such a fast changing industry, how do we really expect everyone to keep up?

Being ahead of the game on the topics that matter will put you head and shoulders above the competition, so take the quiz below and see how you would fare.

Then check out my responses below to see if you agree – extra points for disagreeing and commenting with your own opinions.

1. What is the difference between Search Retargeting and Site Retargeting?
2. How/When does de-duplication occur in display advertising, and what is its impact?
3. What is the opportunity with the Facebook Exchange (FBX)?
4. Complete the sentence, “View-through attribution is…”
5. What are ‘Programmatic Marketing’ and ‘Big Data’?
6. Bonus question: Will 2013 be the year of ‘mobile’?

1.  What Is The Difference Between Search Retargeting & Site Retargeting?

In brief, search retargeting is a prospecting buy – it finds those individuals who have searched for relevant terms on Google, Yahoo! or Bing and targets them with display ads.

Site retargeting refers to targeting individuals who have visited your site and then left, often without completing the action you want them take. It talks exclusively to those individuals who have already visited, and so is a double investment in the same audience.

The two work very well together: search retargeting for prospecting, and site retargeting for increasing conversion rates.

2.  How/When Does De-Duplication Occur In Display Advertising & What Is Its Impact?

De-duplication occurs in an ad server such as DoubleClick, Atlas, Mediaplex, etc., and refers to situations when credit for conversions is taken away from a particular media placement because multiple tactics contributed to each of those conversions.

Such ad servers are trying to do good – clearly if a site generated 1,000 conversions in a month, then the ad server should not report more than those 1,000; but, given each conversion can be influenced by many placements, if you were to add up each of the conversions the vendors report, they would total more than 1,000.

De-duplication therefore, is the action of reducing the number of conversions down to the real total. And, you may say that’s a good thing – I would – but it is the way it does so that causes the problem. Ad servers utilize last-touch as their default algorithm, and that causes lots and lots of problems.

At Chango, we are commonly asked to find brand new prospects for our clients, (and who wouldn’t want new prospects!), and then, the client will have site retargeting in place to help those individuals convert.

By design, search retargeting should be bringing the new people, and site retargeting helps to push them over the line. But when de-duplication occurs, all the credit is given to the tactic that last touched that person, which of course is site retargeting…

De-duplication is really an attribution problem, and one that causes smart marketers to make really dumb decisions.

3.  What Is The Opportunity With The Facebook Exchange (FBX)?

The Facebook Exchange is the latest advertising offering from Facebook. You are likely aware of their Marketplace ads that are sold on a CPC basis, and use Facebook member data to decide whom to target. In contrast, the Facebook Exchange (FBX) operates on a dynamic CPM basis, uses the advertisers’ 1^st party data, not Facebook data.

Currently, the majority of dollars flowing through FBX are rumored to be retargeting dollars – this makes most sense as a site visit is the easiest 1^st party data for an advertiser to work with.

The opportunity is broader, though. There is no limit to the data that can be used, and as an example, Chango became the latest integrated partner for FBX, and uses a massive amount of search data to prospect for new customers based on their searches on Google, Yahoo! and Bing – the intent of search married with social.

4.  Complete The Sentence, “View-Thru Attribution Is…”

View-thru (VT) attribution is a way to measure the impact of digital advertising. A view-thru event occurs when an individual is exposed to an ad, goes on to make a conversion or a visit, but has never clicked on that ad.

VT is one of the most controversial topics in digital today because marketers have been led to believe everything in digital should be 100% accountable down to the last cent or pixel. Advertising is advertising though, and given the many different possible touch points there are for an individual today, getting a 100% answer is impossible.

I like to use a Winston Churchill quote to respond to marketers who don’t believe in VT (see How Winston Churchill Solved the View-Thru Attribution Argument).

Anyway, what Winston Churchill did was offer a woman a large amount of money to have sex with him, to which she agreed. He then asked if she would have sex with him for $20. She responded, “What do you think I am, a prostitute?” Churchill replied, “Ma’am, we’ve already established that you are a prostitute. Now we are just negotiating the price.”

So, when we talk about VT, the argument is we know advertising influences consumers, otherwise we wouldn’t run it; therefore, we have already established what sort of metric VT is, now we are just negotiating how much we should use (see: View-Thru Attribution For The Search Marketer).

5.  What Are ‘Programmatic Marketing’ & ‘Big Data’?

Don’t we make things difficult for ourselves?! We have already adopted these terms, declared them as the next big thing, and then completely screwed up the definitions! You would be forgiven for getting this one wrong.

Big data, by definition, is a set of data so big that it’s cumbersome to use. Oh great, that one is helpful! Programmatic is a rules-based method for using data in your advertising program. So, a simple answer is that programmatic advertising is a way to design complex targeting rules, and it solves the problem of big data by making it usable (see: Can Search Marketers Own ‘Programmatic Marketing’ & ‘Big Data’?).

6. Bonus Question: Is This The Year Of The Mobile?

No.

How Did You Do?

Did you get all 6 about right? If you did, then well done – but if you didn’t, I suggest you use the holidays to catch up a little. None of these topics are going away in 2013, and in fact will only grow in complexity and scale. Right now is the time to ensure you understand those things that have moved from being advanced topics to being the basics.

Happy holidays everyone!

The reasoning behind such an argument, in part, comes down to simple efficiencies and management fees, but also because search marketers can contribute a unique type of thinking to media because of their background in auction-based buying.

I am yet to see it really be proven true.

The one case where search marketers have ‘stolen’ real revenue from display teams has been within Facebook advertising (not the new FBX exchange) – when launched Facebook advertising was obviously the domain of media buyers, but when they switched to CPC buying, the search marketers leaped on it and claimed it for themselves.

At the time, I was running a global agency media team, and I saw a big portion of my dollars disappear across the office with little say in the matter!

As an industry, we then said search marketers should really own the new world of RTB (Real-Time Bidding) and exchange inventory because it was less about the creative and context, and more about the strategy for getting the right price in the auction.  (See Why Search Marketers Are The Future Media Planners.)

But this largely has not happened – search marketers can not get past the barrier of needing display creative units, and rarely will tackle or own the debate around view-thru attribution, which is absolutely required for display. (See Why Search Marketers Are Losing Out With Search Retargeting.)

Even with something like Chango’s Search Retargeting, it is still 75% bought by display planners, and only 25% by search marketers – even though it is keyword-focused!

Then Along Came ‘Big Data’ & ‘Programmatic Marketing’

When I read about big data and programmatic, I have two feelings – the first is one of excitement, because this is what real-time marketing has always been about, marketers just didn’t know it.

But, I also sigh because these buzz terms already have many different meanings, and confusion abounds throughout. So lets start by defining what these things are:

Big Data: Really big sets of data, cumbersome to collect and manage, and difficult to extract meaningful actions in a realistic time frame. Basically, its lots of data!

Programmatic Marketing: It has come to mean the way in which RTB media is bought, using data to make marketing decisions in real-time. Accurately, it refers more to the actual action of how media is bought in real-time – an approach that allows rules to be written that have varying outcomes given the circumstances.

Think of it this way: Programmatic Marketing is the process by which Big Data can be used within marketing (but it’s ironical, given the real definition of ‘big data,’ that the data is too large to be used meaningfully; solving the problem makes ‘big data’ go away. :)

As an example of what Programmatic Marketing might look like, we reinvented the concept of simple site retargeting and made it Programmatic Site Retargeting.

For a home improvement client, instead of just retargeting everyone who visited the site with an ad for the last product they looked at, we combine a programmatic approach with ‘big data’ to extract more value.

So, for an individual who arrives on a Friday afternoon looking for a new kitchen, we bid high and at a high frequency cap because they are likely to make a big purchase decision that weekend.

But an individual who arrives on a Sunday morning looking for instructions on how to hang wallpaper has very little value, and so the rule automatically digests this and bids very low.

The promise of Programmatic Marketing, therefore, is more efficiency.

Doesn’t That Sound Like A Role For The Search Marketer?

And, we come full circle. A search marketer (in general) tends to be more analytical than a display media planner; they are more comfortable in a quantitative world and have much better Excel skills(!). Display planners still think in isolation from tactic to tactic, and search marketers tend to consider their spend as a whole. We know the story.

So what’s different this time?

Well, Programmatic means more than simply buying in an auction environment. It is inherently more technical and rules-based. And, the question should not actually be ‘can search marketers own programmatic marketing,’ the question the search marketer should be asking themselves is ‘can they evolve faster than display planners into Programmatic Marketers?’

It seems somewhat of a fair race today. The search marketers appear to have more relevant skills, but the media planners own the all-important budget. The display planners also have access to the right tools (like Chango Madison which is the first PMP Programmatic Marketing Platform), but search marketers know that they should be targeting individuals, not shouting at crowds, the whole point of the RTB revolution.

What Happens Next?

Unfortunately, there is likely to be a lot more confusion to come, but with confusion comes opportunity. If I were a media planner or search marketer today, I would be learning everything I could about programmatic marketing, knowing that when my company or agency comes looking for someone to fill that role, I would be best placed to own it and make it successful.

But driven in part by imitators confusing buyers, the questions have become much more sophisticated, and as we continue to move along the adoption curve, this will likely continue.

Beyond the basics, there are 7 questions that are most commonly asked of us:

• Where do you get the search data?
• What effect does Google SLL have?
• How does this differ from site retargeting / remarketing?
• Who buys this – the search team or the display team?
• Isn’t Search Retargeting just for direct response?
• Can it scale?
• What type of keywords should I use?

Where Do You Get The Search Data?

This is the one area that people unfamiliar have the most confusion with. As a general rule, vendors work with a number of different sources, dependent on what they find to have been effective.

At Chango, we prefer the use of primary event data – this is otherwise thought of as ‘referrer’ data and can be captured when an individual has performed a search on Google, Yahoo or Bing and then landed on another site. The receiving site can see the referral URL and therefore the search term that was just used.

There are partnerships in place that financially compensate those site owners for firing code and helping that data be captured, and the practice is compliant with the standards of the NAI.

Other sources vary – there is toolbar data that captures searches as they occur through a co-branded toolbar, there are ‘tier 3’ search engines more than willing to monetize their inventory and other forms of software.

Not all data is created equal, and so the trick is to analyze what works in each situation and use that source.

What Effect Does Google SSL Have?

Some, but nothing dramatic. I am sure anyone in the Search Retargeting business would prefer they didn’t make this decision, but the impact on data has been relatively negligible, and nothing that impacts the scale of campaigns.

How Does This Differ From Site Retargeting / Remarketing?

It surprises me that this still gets raised so much, but as a micro-industry, we also have to take responsibility for such a terrible product name

In short, site retargeting (or site remarketing) is about showing display ads to those individuals who have visited your site and then left. They are your existing visitors, and maybe even your existing customers. Site retargeting has a lot of value, but let’s be clear, that value is in generating additional revenue from an audience you have already paid to generate.

Conversely, Search Retargeting is a prospecting tool (when done right). Its goal should be about findingnew individuals who have not yet visited your site.

If that is understood then it should be clear that site and search retargeting are not competing with each other at all, they are in fact complimentary and can / should be run together. Search Retargeting should bring you the brand new people, and site retargeting is there to ensure they convert if they wander off from the purchase funnel.

Useful to help understand these distinctions further is the ‘7 types of retargeting‘ inforgraphic, as well as the Retargeting Exposed white paper.

Who Buys This – The Search Team Or The Display Team?

Great question, and one that we got a bit wrong in the early days! The thinking was Search Retargeting would be ideal for the SEM folks; it is based on keywords, it uses new future of RTB (real time bidding) that many of us think search marketers could dominate, and was performance driven.

The reality is that Search Retargeting is fundamentally a display buy, and there remain mental barriers in the minds of search marketers that seem a sticking point.

Approximately 80% of our business comes from display buyers (split between agencies and direct clients), and 20% from SEM teams. Both types of buyers behave very similar in each group, but if anything, the display buyers tend to have a shorter buying cycle and bigger budgets because they already own budget for ‘banners’.

Isn’t Search Retargeting Just For Direct Response?

No, no and no… Again, as an industry we have to take responsibility for the naming – putting ‘retargeting’, or even ‘search’ in anything immediately makes marketers think of direct response.

But think about how Search Retargeting actually works – like any form of behavioral media buy, it is about defining appropriate audiences by their attributes.

When buying from a premium network this is often demographic or psychographic, when buying from a direct site it is about who they are for being on that site; with Search Retargeting, it is about the interests and needs they have based on what they are actually telling you what that is.

Brand buyers have woken up to this in a big way – why invest $250k on a takeover with 40% wastage, or a premium placement with an uncertain outcome, when you can find the right people to influence at a quarter of the cost! The early adopters are winning big, the rest will come in time.

Can It Scale?

Yes thankfully, otherwise, I am packing up and going back to England! In the early days scale was limited by both access to data and access to media. Both of these problems were overcome a long while back in the evolution, and there are several retailers signing off more than $1.5m a quarter.

From a media perspective being tied into all the media exchanges is a big advantage, and grants access to over 90% of the population. And RTB is only at 20% of digital media buying today, set to rise to 50% over the next 3 years, and so the opportunity only grows.

The media quality has changed too – it is no longer the remnant no one wants, it is all media.

What Types Of Keywords Should I Use?

This very much depends on your goals. The primary categories we see are:

• Current PPC keyword list (which will usually cover all product and category terms)
• Head terms that have been optimized out of PPC because of cost
• SEO wish list
• Competitor brand and product terms

What If I Have Another Question Not Answered Here?

Feel free to ask it in the comments section below, it is great to have open dialogue right here, or you can contact me directly with anything specific or confidential.

In its simplest form, Search Retargeting allows a retailer to find those individuals who have searched for a relevant term on Google, Yahoo! and Bing, but have not yet visited their site. This latter point is critical – Search Retargeting is unlike a typical retargeting or remarketing program in that it focuses on net new customers, not those who have abandoned your pages.

Retailer Case Study

In a recent campaign for a major traditional grocery retailer (with a growing online business), Chango was used to deliver against a set of diverse, and sometimes surprising, set of challenges.

A. Challenges

Our grocery client is in a competitive market, one that is fundamentally evolving because of changing buying habits, especially the use of online grocery ordering. They must compete with behemoths that can outspend them 100 to 1 and are household names, not just brand names. They have dollars to invest, but like many challenger brands, they must be smarter than their competitors, and squeeze out every last drop of benefit.

They struggle to rank for important home delivery terms in SEO because the market is too crowded, and their budget doesn’t stretch to head terms in PPC like [supermarket] because in some markets it is simply cost-prohibitive.

Like many such retailers, their product portfolio has diversified, and they want their audience to know they are a one-stop shop for flowers, electronics and wine. And as a high-end retailer brand, they are extremely cautious about the quality of their ads, as well as the placement

Most importantly, they want net new customers, as these will bring the maximum value over their lifetime.

B. Methodology

Search Retargeting is about defining audiences based on their intent, and that intent is defined by the keywords that the individual is searching on. The audiences defined included:

Curious about online grocery: [online grocery], [online supermarket], [supermarket home delivery], etc.

Competitor conquesting: brand terms of competing supermarkets and retailers, and the brand names of their major in-house product lines

Flower buyers: [order flowers], [mother’s day flowers], [flower gifts], etc.

Wine buyers: [order wine], [wine cases], [wine delivery] and brand names of wine information sites and guides

Each individual was shown an ad that was relevant to the category they were searching for. All ad placements were screened with a three-layer brand safety methodology to ensure suitable placement.

C. Results

• Against an ROI target of 6:1, Chango delivered results peaking at 11:1
• 42% of all orders came from net new customers
• One third of all sales were from individuals who had searched for a competitor’s brand term
• 20% of sales came from new customers buying wine
• 26% of sales came from new customers buying flowers
• The gap created by the SEO and PPC environment was filled, providing a presence for consumers on strategic terms

D. Driving more efficiency

Smart buyers know that a good campaign should be optimized positively, and negatively, against simple ROI or ROAS (return on ad spend); yet, this often was the hardest point to teach media planners when I was on the agency-side.

What we mean by this is that a campaign should be measured against all its goals, not just the single goal of ROI. You sacrifice some ROI in exchange for achieving a secondary goal that has strategic value.

With this particular campaign, the client was very happy with the ROI we were achieving, a peak of 11:1 return instead of the 6:1 goal.  The campaign was analyzed, and it was seen that we could invest more in wine and flower terms, even though they were low performers for ROI – this lowered the ROI initially, but generated thousands of new customers to strategic product ranges.

In addition, the retailer deployed the Exclusion Pixel site-wide, which tells the system not to target any individual who has already visited the site, therefore driving up net new customer percentage, and removing the overlap with an existing site retargeting buy.

Each week, we would supply a list of order IDs associated with the Search Retargeting program, and the client would match them against their CRM system, giving us access to data that helped the optimization further for net new customers.

The Q4 Gold Rush

In a recent eMarketer / Chango study, eMarketer reports that this retailer is typical of others who are turning to Search Retargeting. Nearly all respondents (93.9%) reported that they were using it to acquire new customers, and two thirds (63.6%) were using it to conquest from their direct competitors.

A reason for the latter is that unlike PPC, the targeting of competitor terms is hidden from view; you may see an ad being served for your competitor, but you don’t know which keyword was the trigger. This prevents the bidding war that often breaks out in search.

When coupled with high returns and consistent performance, retailers are realizing that they may have found an early Christmas gift. In the white paper, Retargeting for Retailers Exposed, seven real marketers from brands including Resolution Media and iCrossing speak out on the roles retargeting can play, and also the pitfalls to be wary of.

Happy Q4 conquesting!