What's a Winning Attribution Process Playbook?

Like anything in business, it’s one thing to say it…it’s another thing to do it.
One of C3 Metrics’ clients shared his process playbook for attribution buy-in and organizational adoption.
In the sixth video, Matt DiAntonio of Carbonite shares how he tackled one particularly challenging marketing channel, saving $750k each year, and then growing revenue from that channel by 2x.
But results come with process. The below videos illustrate his process which he shared at the C3 Metrics Summit, which received the highest scores and accolades we’ve ever seen.

Step 1: Start Small – Tangible Wins With Creative


“I think a lot of people ignore the upper funnel phase and jump right to the bottom part of the funnel.
But for our business, cultivation between upper funnel to conversion is really important to a comprehensive marketing measurement program.
We call this our prospect to lead space where we have inquiries that are becoming MQLs that become revenue.
All of that happens with touches.
The touches aren’t just Marketo, or Pardot, or whatever your marketing automation platform is… these are people.
We like to think of them as weird rows in the spreadsheet, but they’re people and they’re living and engaging in a world where your brand is present and every once in a while you’re going to see some piece of creative that pops with something meaningful and relevant for them to digest, and the brand becomes a trusted adviser.
In that conversation, you build some equity with stakeholders across the marketing team, and what our analytics team does is use C3 Metrics to be able to corral all of these disparate touches, then we drop them into a data Lake and we start looking for correlation.
We want to see if the creative assets (videos, touch points, etc) that are correlated to driving someone deeper into the success metrics that the CFO cares about.
When we find that correlation, we say “let’s juice the visibility” and it’s really for marketing guys to deploy that particular creative element.
If the correlation holds, you start to build more assets like it, and you start to spend more marketing dollars with evidence based metrics.”

Step 2: Discovering Trends and Exceptions (Good + Bad)


“I start every morning reviewing every single important trend in the business.
If there’s an issue, it’s better to jump on that early.
So what we’ve done is take the C3 Metrics data from FTP, create a Tableau server dashboard which takes all the things that I care about…putting them in one line of sight with dates all aligned, so I can see where there’s correlation and exception.
We have server side impressions, server side clicks, marketing spend by channel, and the same for C3 which measures consumer side actual views, consumer side clicks, and spend that’s in C3, and ultimately …..
The attributed value that comes out of it at a high level.

C3 Metrics ‘ORAC’ Consumer Journey


I’m going to talk a little bit about why we built this so what you’re seeing here is a tableau dashboard where the ORAC sections of the funnel (Originator, Roster, Assist, Convert) are split out and that red line is the Mendoza line that tracks ROAS or AVSR…did I spend a dollar did I get more than a dollar back in terms of attributed value.
One of the important things you have to keep in mind as we go through and look at either a campaign centric view or in this case we’re looking at a medium specific view this is the affiliate program mediums and tactics play different roles in the process of converting.
You have to be really conscious of not clipping your R’s (Roster touchpoints) in favor of your O’s (Originators). You want to be able to look at it and see a consistent performance across these trend lines, and if something goes jumping way up or drops way down…we can look up here at spend and say we spent less and we’re getting less of this value.
But the good news is overall ROAS or AVSR is still hovering around 2.3 to 1.4. When we talk about this, we want to talk about outcomes.
Before C3 Metrics data was available to us…
Somebody with a suit and tie would show up at your desk and says. “what’s happening with this outcome something’s up something’s down?”
Then, in the old days, everybody would run over to Google Analytics or whatever web session stuff you’re looking at will never agree with C3 Metrics, because the lasts click lies!
When people say (based on Google Analytics) something’s really kicking ass…let’s just put that aside because we need a vehicle to be able to speak intelligently to outcomes though C3 data and the process of discovering trends and exceptions, and it can only happen quickly if done proactively.
Now, with C3 Metrics data plus this process…
You begin to be able to speak intelligently to outcomes with minimal amount of effort.
Ideally you have marketing managers and directors who are in there looking at these trends who can tell that story pretty actively so when the suit shows up at my door I give them the answer.
But this can be done with any outcome. For us, it’s a trial registration but this could be video views, or MQLs generated. It could be for any of these things appropriate to your business.
 

Step 3: Data Integrity = Rowing In Same Direction


“We put together a data integrity process to make sure we’re proactively looking at this stuff when you think about C3 as a tool for us.
We have Commission Junction, we have Marin doing our search, and we have an ad server for display; we look at impressions, clicks and spend.
Let me show you what that looks like and take you through this project we call Ion Cannon because everything in our world has a cheesy Star Wars name if anybody’s familiar with what ion cannon is extra credit to you.
What you’re seeing in the green bars are the spend in commission junction and in the blue bars is the spend in C3 Metrics.
That gray bar down at the bottom is the variance between the two of them, and it’s really important to note that we’re not looking for constant zero percent variance…we’re looking for a consistent variance.
For example, think about impressions. C3 Metrics measures impressions very differently than the way an ad server measures an impression (serve side versus consumer side).
We’re not trying to get Tigers to change stripes, we just want to see a consistent variance and I’ll show you what that looks like.
Here’s when the alignment and variance looks correct, but here’s where it looks off…so what’s going here?
This story is pretty interesting: it’s a rogue (fraudulent) vendor in Commission Junction who started botting their own site driving false conversions which looked glorious to Commission Junction (and they were paying them) until we showed them this data…saying, take a look at our run rate and take a look at this anomalous jump. This is BS
This is all fraud, and we showed Commission Junction this exact report: that green line coming back down is Commission Junction getting them out of the by cutting them completely.
They agreed that the rogue vendor was botting and they cut them. Without C3 Metrics, a typical marketing manager or director would look at the siloed, last click, not fraud-corrected silo (not C3), and say it’s converting well, but because we control the data in our tableau data, we scrubbed out the fraud that C3 caught, so we don’t have to constantly have and our trend goes back to a nice flat normal expected line and everybody has confidence in the data.
Now here’s a different situation which is much more common where we have this steady 0% variance on the left-hand side; green and blue bars hug and agree doing all sorts of good stuff. But all of a sudden this variance starts to pop up, and within 12 hours we had somebody on the operational side saying did you release something that didn’t have tagging?
What happened here right is a really good story for the business, and it shows how we have to talk about things as a tea. What occurred was that the B2B side of our business found good success for “click-to-call.” So think about how that would align in the data: they started turning on “click-to-call” which increases spend from Marin search, and increases clicks because click-to-call counts as a search click but there’s no outcome that we can tie that to in our world.
That side of our business is smiling and dialing, but before you jump to conclusions and say the ROAS or AVSR is are down and spend is up, you need to color this differently and everybody comes to the table prepared, and see this is good for the health of the business, and keep on going”
 

Step 4: Running After Walking (Very Advanced)


We built a tool called Viceroy that we’re really close to open sourcing for anybody who wants to be able to use it in conjunction with C3, and it’s a tool that anchors on a standard UTM value that assigns a pid into the tactic right we assign one individual ID to the tactic that creates the link that can get either pushed through an API to an ad server or it can get sourced by marketers so they can move unfettered without being slowed down.
I see this all the time: someone has to jump through three different offices, and it takes ten days to get a link.
I think it took fourteen days at the start of last year in Carbonite.
We put all that stuff in and we have people agree with it. It’s mutable, it’s not immutable…so we can go in and change it and then we push a publish button and everything that was published goes ouinto the data lake, and as these things start moving and shaking in the world around you can start to show reporting for a campaign just by following the process the right way.
You want how me all the pages that came from you know these PIDs that were under a campaign pivoted on its head. Then show me all the activity we’re seeing which had lead-generation intent and it starts to pivot the cube and works amazingly well wrapping the right context around your tool.
 

Success: The Challenging Marketing Channel: $750,000 Savings + 2x Growth


When I got to Carbonite, there was a golden goose…a sacred cow, and that was Affiliate.
This marketing channel hid behind last touch attribution for so damn long, they were impossible to move away from the notion that failure could ever be touched. They had this group of people singing “cumbaya” with flawed last click metrics.
We all raised an eyebrow and said, “We’re data people, we’re not going to come in with emotional assertions of their ridiculous nature, we’re going to come in with some data that is going to prove to the team that we can do something better.”
About two and a half months into the process with C3 Metrics, and absolutely mind-bendingly terrible results of last click were exposed.
This golden goose that for so long had been the sacred cow was returning pennies on every dollar spent and we said, “Listen guys, I know this is going to be hard. We’re going to take the blanket away and make some really tough decisions.”
The initial response, “we can’t do that, because they’ve been loyal for so long!”
And we said, “Sorry, we’re cutting their commissions and we’re going to cut them drastically.”
Everybody came in and we see this really strong jump upward in terms of the ROAS or AVSR.
I don’t know how many people are familiar with what that number means, but it’s an Attributed Value to Spend Ratio.
For every dollar we spent, we’re getting back an accretive business value.
They declared victory, and tried to plant the flag and on to the next issue.
But I said, “No, no. This is an evolution.” We need to come back and say, “What’s the next lever to pull?”
We see total volume increasing and we said, “Let’s take that extra money and go make some bets.”
Here’s the big success story in the slide. It isn’t the big jump in ROAS or AVSR, it’s the fact that as a group we were willing to take some risks.
We had a really solid level of confidence in our listening device and worked through finding a new lever that was able to drive business value and ultimately if you were to look at this, this is the first couple of quarters of C3’s existence, the affiliate program now drives quite a bit more volume.
Now…it’s about 2x total volume.
The exact same level of pay in terms of the cash out and we have this up to about 2-2.5 at any point in time.
Really, really cool success story.
 

Three Final Secrets


There are three things I want to tell everybody in this space whenever I have the opportunity. The first one is, adopt an executive strategy. Take ownership of the conversation and be concise. What does that mean?
I’ve sat in enough board meetings and I can help you all through this. Don’t come to the table with a ream of paper. If I had a nickel for every time I saw the marketing team show up with every click-through and, stop. Just stop.
There’s three questions that you should answer at every one of these meetings. One is more important than the other.
What’s working? What’s not? Don’t run away from what’s not. Executives and leaders, they don’t want cheer leaders. They don’t want doom sayers, your job isn’t to nit pick every possible thing that sucks.
Your job is to say what’s working, what’s not and what are we doing about it?
That last point is the absolute most important piece.
When you start talking about what you’re doing about it and you keep an active log of it and say, “Listen, we had this conversation. As a team we’re driving change in the organization. Is it going to work? I don’t know, but we’re doing something that shows that we have ownership and accountability in this company.”
All of a sudden what you’re going to see is the entire C-Suite start showing up at these meetings. We saw it.
Literally, we have a Thursday meeting and our entire C-Suite sits in the back of the room because they want to hear what’s going to happen next. They want to see who’s playing ball and who’s owning it and that’s the most important part. If you leave this out, the value that you’re bringing to the organization starts to go.
I hear a lot a lot of people say “we have this awesome system there’s all sorts of good stuff coming in here but we don’t see any change happening.”
When was the last time you actually helped someone else own some of the chains so they look good. If I’m the one to ring the bell, other people are going to say, “f*ck you, I’m here to do a good job.”
Make the other people in your company successful, but don’t lose the paper trail and what we’re doing about the paper trail, who you’re empowering, and how it’s changing the company.
Connect the dots for everyone and show them the business value. That’s the piece that’s always missing.
Celebrate success and don’t take all the credit.
Holds quarterly and annual reviews. It may feels a little bit like you’re opening up your eye and putting a needle in there. It’s a lot of work and the reality in today’s world to be an impactful leader or analyst is to know your numbers inside out.
QBR’s quarterly business reviews don’t need to be a twenty page diatribe, just: what’s working, what’s not, and what are we doing about it?
Get that momentum started and then you become the Exec Team’s best friend because they need all the information you have: to say what’s working, what’s not, and what are we doing about it.
If you’ve taken every meeting that popped up onto your calendar, you’re not making a difference in your organization. You become a consensus member of these people who hold meetings because they need to have everybody there giving a thumbs-up for them to drive change. It’s the wrong way to do it.
I’m here to tell you it’s okay to say “not right now.” Don’t just say no, just not right now.
These are my priorities, here’s what we’re doing, and I need time to be able to do this in a way that’s meaningful right…and “not right now.” Your priorities aren’t mine: the company’s priorities should be what we’re looking at, and if there’s something that needs to change let’s do that, but make time to stop and think.
Schedule think time. Put think time on your calendar for these C3 tools whether you’re using customized dashboards like ours or the standard C3 dashboards.
If you put the work into building the right mindset which takes effort, and then you don’t do anything about it… you become one of the problems (which is I’m running around doing so much I can’t figure out what I should be doing and that’s the biggest problem).
If you can’t say what it is you should be doing and making a little bit of an impact on that on a daily basis you’re not helping your company.
We don’t need more people eating more snacks in the middle of meetings, we need people to make an impact.
 

All about Matt DiAntonio (Belichick of the Playbook)


Spent 10 years at Silicon Valley’s Trend Micro which is a security vendor that vendor competed with Norton and Kaspersky and McAfee, working with a really bright CMO and a really stingy CFO.
That CFO and I had an accord where I would sit down and take him through the numbers and be data first, and then he would give me the money I wanted.
My CMO told me I was really good at analytics: grabbing all the data, telling the story, and being concise.
I thought the CMO was politely telling me I was a sh*tty marketer and was trying to get rid of me…but he asked me to build an analytics team for Trend Micro and I did that for four years.
At some point I had a bigger vision for what I wanted to do, which was to bring all of the data together to sit at the vortex of the company and tell cross-functional stories.
I love marketing analytics: but it doesn’t mean a heck of a lot if you can’t impact the top + bottom line…driving positive, accretive behavior.
 

Ad Fraud | DOJ Fails To Win Felony Convictions (Docket# 16-CR-441)

The first ever trial for advertising fraud failed in the Federal Court of New York. Over 20 brands were cheated out of their digital ad dollars including: Nike, Disney, Procter & Gamble.

Despite great lengths of extradition from Amsterdam of a European citizen to be jailed and tried in the 2nd District Court of New York…the DOJ failed to obtain conviction of any of its felony counts surrounding Fabio Gasperini’s ad fraud ‘botnet’ case.

In a seven day long jury trial, a dozen+ witnesses from computer experts to ad industry veterans, to government experts failed to convince a jury of the felony charges (full details here) which would have resulted in 70 years of prison sentencing.

Instead, after two years of case preparation the only sentence given was a simple one year misdemeanor. The defendant churned up more wasted court time with an appeal (appellate court decision 17-2479 2d Cir. 2018 here)

“Ad fraud is not unusual at all…40% of programmatic display is fraudulent, and 51% of it is non-viewable, yet advertisers are changed for it,” said C3 Metrics advertising attribution measurement Chief Attribution Officer Jeff Greenfield. “In attribution measurement, a platform has to purify all that fraud before it’s measured and modeled. It doesn’t mean that fraud goes away, but fraud reveals itself in poor ROI plus other fraud controls.”

Ad fraud has been around for ages, but Docket# 16-CR-441 was the first time it went to trial. It may be the last time it goes to trial, because it’s a highly complex scheme. Ad fraud isn’t a criminal trial, which is far easier for the average American jury to understand.

It raises an interesting legal question: do the ad exchanges share in any culpability? If generally known ad fraud practices are happening, but not enough is done because of pressure to make revenue, are publishers legally culpable too?

The IAB SafeFrame initiative to roll-out viewability code (C3 Metrics open-sourced its viewability code to the initiative) seems be dead in the water. What’s the incentive for a publisher to place a piece of code on their site if their revenue drops?

U.S. law is often determined by legal precedent, and this is the first-ever advertising fraud case to be tried in the U.S. It paves the way for whether DOJ decides to extradite European, Russian, or Chinese ad fraud botnet citizens (if they can find them) with evidence beyond a reasonable doubt, or if DOJ will pursue easier to prove, criminal cases with US citizens.

Because of Sarbanes Oxley, CFO’s can now be sentenced to jail. Will CEO’s and their advertising supply chains be sentenced to jail?

Enter the newest issue for the auditing profession to tackle.

"5.4" Invoking Marketing Dept Audits by PwC, EY, Deloitte?

According to Salesforce’s 2020 Report, 5.4 = the average number of marketing channels used in 2017 by advertisers. In 2018, it will climb to 6.2 different channels.

Yet according to the DMA Report from 2017, only 13% of advertisers are using advanced attribution. Meaning, 87% of advertisers still give credit to one channel (also known as last click measurement). Imagine: if all 100 yard dashes were 87% shorter, only 13 yards long.

Now enter a real-life example of measurement shortfall. The U.S. Army’s squander of $220 million ad dollars discovered via Adweek’s Oct. 5, 2017 audit document titled “The Army’s Marketing and Advertising Program, Return on Investment.”

The audit revealed, “only 3 of the 23 marketing programs generated a positive impact during the year…For [fiscal years 2018-2023], the U.S. Army would continue to use $220 million for the same ineffective marketing programs.”

$220 million of waste discovered: paid for by every American tax payer.

The Army, like 87% of all advertisers, didn’t use modern advertising attribution, and the Army used > 5.4 marketing channels.

In 2018, however, the Army chose C3 Metrics to be the unbiased platform measuring cross-channel performance of every ad dollar spent.

5.4 ≠ 1 (Junior High Math)

5.4 channels does not equal one channel.

Statistically, any marketer not using modern attribution who’s spending ad dollars in 5.4 channels will be mistaken in its measurement by 5.4x.

Without the “Matching Principle” Nothing Matches

Every CFO and auditor understands the “matching principle” also referred to as activity based costing. Unit level revenue matched to unit level cost.

Between digital fraud, lack of online display viewability, and inaccurate ‘last click’ measurement used by the US Army: measurement failed to incorporate basic accounting standards which inadvertently allocated credit to the lower funnel and the last marketing expense…versus a full funnel, activity-based costing approach.

This, is why the Army missed 20 of 23 KPI’s (and cost U.S. taxpayers $220 million), and why C3 Metrics is now in place for the US Army and US Taxpayer going forward.

Nielsen’s former Chairman & CEO John Dimling, who joined C3 Metrics in 2012, said he was a champion of Nielsen and Arbitron’s ambitious but ill-fated ‘Project Apollo’ to create a single-source measurement system.

Project Apollo failed to succeed with old technology, but modern, complete attribution platforms with massive computing power prove to be today’s answer.

Over-counting is the most common error in marketing accountability.

When adding up revenue from siloed tracking systems, the industry over-counts conversions/KPIs by 5.4x. Each marketing channel and vendor tries to claim 100% of credit versus the activity-based method deployed every day in accounting.

Without the matching principle, nothing matches.

Enter the Auditors

Section 302 of Sarbanes-Oxley requires that quarterly and annual documents “contain no untrue statements.”

CEO’s and CFO’s who’ve made statements about marketing/advertising effectiveness based on single channel marketing measurement versus modern attribution measurement, may be putting themselves at risk under Section 302 of SOX.

It’s analogous to being out of code for electrical. Knob-and-tube electrical can still be used, but it’s out of code. Ungrounded electrical can still be used, but it’s out of code.

As attribution is generally well-known by CMO’s at publicly traded companies (but not being utilized) statements made by CEO’s and CFO’s regarding their advertising effectiveness may be considered to be untrue because the reporting mechanism of last click is untrue.

Here’s where it gets murky: if the CEO or CFO knew about the difference between last click marketing measurement and advanced attribution measurement (or their auditors knew)…what did they know, and when did they know it?

But, is “I simply didn’t know,” a defensible argument for violation of SOX Section 302 (which is punishable to a CEO and CFO by jail term and fines).

If a certain portion of the U.S. Tax Code changes, and the CFO simply didn’t know…Section 302 doesn’t care because auditors (PwC, EY, Deloitte, KPMG, etc) are specifically there to impart their knowledge of untrue statements and or practices.

But it begs the question: as the big four accounting firms enter the space of marketing and advertising, they now know about both the matching principle and the difference between last click versus advanced attribution.

Once auditors know, their publicly traded clients then have to know so they’re not in violation of Section 302.

Six, Lies, and TV Grapes

Abraham Lincoln would love Fox + AMC’s six second TV spot format. Lincoln said, “if I had more time, I would have written a shorter letter. “Shortening of TV spots, according to Nielsen, has been happening since 2014:
15 Sec TV Ad Mix

  • 2014 = 29%
  • 2015 = 30%
  • 2016 = 34%
  • 2017 = 36%

In 2016, Twitter pioneered six second video spots with Facebook adopting the six second format in 2017, suggesting that in the cross-screen world, 15’s and 30’s don’t fit with our fast, Smartphone thumbswipes.

Creatives and the TV ad industry cried travesty upon Fox’s adoption of the six second spot in linear TV, but the times they are a changing.

Consumers have shorter attention spans:

  • Consuming headlines online by scanning them
  • Tuning out linear TV commercials as head & hands fall to their Smartphones

Today, advertising by committee takes over where stakeholders want to jam every possible feature/benefit into a TV spot versus a single, powerful, unique selling proposition of a brand.

As Abraham Lincoln knows, writing a shorter spot takes longer. Longer to analyze what your unique selling proposition really is, and doing it in six seconds.

Fox is selling those six second spots, breaking the same ground when MTV launched the 15 second spot in 1985. Even Microsoft, with a hefty TV budget, bought six second spots on AMC.

Shorter spot length will not only add to cross-screen versatility between Facebook, YouTube, and linear TV, it will actually force creatives (and those who approve TV creative) to distill a clearer brand message.

Branding will skew towards six second spots. Direct Response will play in the longer formats of 30, 60, 120 seconds (as well as 28 minutes).

Lies: GRP’s and Ratings
A high ranking executive at one of the big four accounting firms privately revealed that disparity between actual set top box data viewing and Nielsen GRP’s was not off by acceptable levels, but by large magnitudes.

This executive called it “the lie in TV that everyone has accepted.”

During TV Week in New York, a veteran ABC TV sales rep said, “I hate GRP’s.” 20 years of their ad sales confirmed the GRP being a flawed metric.

Perhaps the GRP is a flawed metric:

  1. For data-centric reasons. Moving towards a much larger ratings sample given that we no longer have most of America watching three networks, but 2/3 of America watching many little networks in many small numbers.
  2. For creative reasons. According to Chief Research Officer David Poltrack at CBS Television 70% of the impact from TV advertising is due to creative. If your creative is 4x as effective, then you can get the same result with 1/4 of your GRP’s.

Do GRP’s matter?

When execs from big accounting firms and ABC say they hate GRP’s, something’s going to shift.

It will likely be a shift to TV attribution, and measurement of creative down to the penny.

TV’s Grapes: HH Addressable

TV is in an amazing renaissance now.

TV has lagged behind digital in reporting metrics, even though the mountain of digital metrics are unfortunately: last-click, siloed, not viewability corrected, and not fraud corrected.

TV’s catching up to digital with HH addressable TV.

In fact, HH addressable TV is at least 6% cheaper than programmatic display.

Imagine TV retargeting on a 46 inch TV: and now the same math makes HH level TV retargeting which is 41% cheaper than programmatic display.


Fraud-free; HH level TV retargeting = 41% cheaper than programmatic display.

Water to Wine

TV isn’t quite turning water into wine…but the new grapes are ready to gather.

Ad Fraud "SportBot" On NFL + ESPN Sites Rips $250MM

“SportBot” served 340 million fraudulent programmatic video ads triggered by visits to some of the most premium publisher websites and sports brands including NFL team domains, ESPN and CBS Sports.
Unlike most bots which focus on no-name sites, the “SportBot,” discovered by Forensiq, focused on premium sports sites (most of which have not deployed the IAB’s ads.txt transparency code)

Here’s How “SportBot” Worked
Someone downloads a program with malware, which allows the botnet to access an internet browser.
That person visits premium name websites which calls a database of domains from the “SportBot.” Upon a domain name match (say ESPN.com) the “SportBot” trigger browsers to open in the background, causing video ads to be served.

“The sports site is not losing any revenue from the bot, but their name is being taken advantage of,” said Amit Joshi, director of data science at Forensiq. “They look like normal, residential IPs where the bot is opening invisible browser windows and loading the ads legitimately on the sites, then poses as a publisher and selling them into a network.”

How Much Did They Take?
Forensiq estimates the ad fraud to be annualized at a $250 million take.

“Botnet fraud lives in two places primarily…mobile and video,” said said Jeff Greenfield, Co-Founder of advertising attribution data cloud C3 Metrics. “Video because the CPM’s are the highest in digital. Mobile because accidental clicks (with a high CTR) can hide click fraud in a big number. Fraudsters have the same mentality as the mob…they skim where there’s big numbers.”

P&G’s CMO, delivered a tongue-lashing to the digital media industry, citing: “75 cents of every dollar P&G spent in digital media never reached the consumer.”

“Programmatic display is like a knife-fight every day,” said Co-Founder of C3 Metrics advertising attribution data cloud Jeff Greenfield. “Because our C3 viewability tag runs with every impression…it has to fight a technological knife fight in a field laden with fraud. But you also have to algorithmically remove fraud in order to false positive correct outcome signals. Fraud won’t go away, but the C3 attribution data cloud can deliver fraud and viewability corrected numbers.”

Jerry Jones, meet Jesse James. Botnets have replaced rifles.

Why Viewability Built-In vs Bolt-On Solves 49% of Attribution

When solving for user-level attribution, 97% of the data collected are not clicks, not TV spots, they’re programmatic display views.
Add clicks, impressions, tv spots, etc together…create a pie chart, and you’ll see 97% plus or minus a few points are display views.
Viewability for viewability’s sake is good. But connecting user-level viewability to outcomes is really the key.
White washing viewability at an aggregate level versus a user-level: will keep you stuck in the old days.
Connecting bolt-on user-level viewability to attribution is nearly impossible. Many have tried.
But user-level attribution with built-in viewability is seamless, frictionless, and automatic to attribution.
Meaning, any display ad that isn’t viewable does not even enter the attribution credit equation.

Because attribution is like a balloon: if the falsehood of viewability or the lie of fraud takes up space in the balloon…it robs credit from the actual tactics creating and accelerating demand.
If you allow fraud and viewability to stay in the balloon of attribution, it’s like a massive tumor taking up space.
Viewability is complicated. C3 has been performing viewability since 2009, and this is perhaps the best explanation of viewability unpacked.
P&G’s Marc Pritchard said that 75 cents of every dollar P&G spent in digital media never reached the consumer. His math is pretty close.
Here’s why 49% of user-level attribution is solved with built-in viewability:
97% of all campaign data = programmatic display impressions
51% of those programmatic display views are not viewable (source: C3 Metrics).
Here’s The Math
(97% x 51%) = 49%
49% of your campaign data is wrong before you even begin.
If you have viewability built-in to user-level attribution, 66% of attribution is solved.
Want data-driven ROI?
Built-in viewability is a huge piece of the pie.

How to Explain Attribution to Your…

Explaining attribution to your CEO, even your Mom…it’s complex. Journalists, analysts, and even marketing veterans don’t fully grasp it.
So in five minutes, we created a video that makes a complex subject simple (and in the last 47 seconds opens the door to peek in further).

What happens when you’re using attribution?
One year after deploying attribution, marketing budgets surge by 30%.
Why?
Because ROI on media spend increases 15-44%.
A 3x return on investment in attribution…which is a better return than Warren Buffet.
Pessimist? Discover how attribution’s biggest pessimist saved $750k/year.
Data driven marketers…don’t get left behind.
Students studying marketing…don’t get left behind.
Grandma…you’re on board.

Uber Suing Dentsu + Future of Fraud

Uber filed suit against its agency Fetch (a division of Dentsu) seeking $40 million in damages for fraudulent ads stewarded by Fetch. Uber spent $82.5 million in online advertising with Fetch over two years.
Marc Pritchard, CMO of P&G gave a tongue-lashing to the online advertising industry indicating 75% of all digital media dollars spent never actually made it to consumers…titling his presentation a wake-up call.
Uber’s lawsuit sends a shock wave across the ad industry.
But legally, they may not have much to stand on. The first ever click fraud trial in the Eastern District Court of New York (USA v. Gasperini, Docket# 1:16-cr-00441) was lost by the Feds.
Great pains were taken in USA v. Gasperini by the Feds to track down the click fraud operator, obtain extradition, capture the fraudster in Europe, and send him back to Brooklyn for what advertisers hoped to be the first of many click fraud convictions.
But Gasperini was acquitted on all felony charges.
This case, however, is a civil case, and Judges/venues play heavily in the courts.
“It’s a seismic shift in online advertising” said C3 Metrics advertising attribution measurement CEO Mark Hughes. “Everyone knows that there’s fraud…40% of mobile display, and video is fraudulent, and 68% of programmatic display is not viewable.”
“But just as global warming is not going to change soon,” continued Hughes “fraud won’t go away anytime soon. But what can change are specialized systems of attribution which automatically remove fraudulent touchpoints, remove non-viewable touchpoints, and measure them against their cost.”
“If this was done this from the beginning,” Hughes elaborated “Uber would have discovered the ROI on those tactics were unprofitable a long time ago, and saved at least 20% of their loss, or $16 million.”
Make no mistake, this is a seismic ground-shift.
Marc Pritchard is no longer the only one saying “I’m mad as hell, and I’m not going to take it anymore.”
Uber, in an unprecedented move seen in online advertising, is saying that and more.
It’s not a wake-up call…it’s a wake up blast.

Where Are You in the 4 Stages to Attribution Zen?

President Bill Clinton said, “In today’s knowledge-based economy, what you earn depends on what you learn.”
But change…is hard. After defaulting to last click and siloed measurement every week for years, undoing that weekly repetition (harmful to the ROI of your ad buy) is difficult.
Tolstoy said change is so difficult that, “Everyone thinks of changing the world, but no one thinks of changing himself.”
Change comes in stages, and change towards multi-touch attribution comes in four stages.

Stage #1 – Last Click
If you’re spending ad dollars in multiple channels…say Google non-brand, pre-roll video, programmatic display, social display, SEO, paid content, and TV…that’s seven channels.
To think there’s only one “converter”, one channel that converts, one row or cell occupying the outdated label of converter…and none of the six other channels you spend hard cash on don’t contribute at all?
That’s logical insanity.
That’s last click. That’s playing baseball with a football.
 
 

Stage #2 – I Want to Believe
The biggest indicators that you’re in Stage 2 (I Want to Believe) are the following:

  • When you add up the conversions from each reporting silo (AdWords, RocketFuel, etc), you end up with 4x the number of conversions you actually have.
  • When programmatic display vendors claim view-through credit resulting in 4,000% ROI, but revenue is growing single digit…something makes you feel funny.
  • After compiling data for the monthly report, conversion rates from seven different channels don’t mathematically make sense…and you worry when Finance will figure this out.

You want to believe there’s got to be a better way.
Not worrying about programmatic’s crisis of 68% unviewability.
Removing click-fraud from taking conversion credit.
Unifying cross-device.
Having the opportunity to integrate offline media and TV into the user-level journey.
Normalizing and fractionalizing credit amongst a team of touches versus only one player in a team.
Doing so with cutting edge, sophisticated modeling: preferably Bayesian and machine-learning Bayesian.
Desire to believe.
Yes Virginia, there is a funnel.

Stage #3 – Instinct Takes Over
Instinct takes over and you act.
You act to do throw away last click’s 6 inch yardstick.
You deploy attribution because all your instincts (and maybe Finance) point to the fact that last click and siloed measurement doesn’t pass the common sense test.
You become the change you want to see happen in the world.
The biggest hurdle in your company, however, is education.
You believe. You have acted.
Like Jim Morrison of The Doors, you have to sing: “Break on through to the other side,” and get everyone listening and singing in unison.
Education is the hardest part of all, because not only are you changing minds, but also hearts.
Hearts that have been invested in making their media channels succeed.
Hearts that could be afraid of the unknown.
These are the pessimists.
But this path of pessimism has been travelled before.
Now…
Now reinforcements have arrived in the form of AttributionCertification.com
In 45 minutes, ten years of attribution research is bestowed upon any marketer or non-marketer.
When anyone asks a question you’ve answered 17 times, you simply say:
“Have you been Level One Attribution Certified?” and park that question until they’ve been Level One certified.
As Yoda said, “You must unlearn what you have learned.”

Stage #4 – The Force Is With You
You’ve deployed attribution many months ago.
You’ve successfully flown through the asteroid field.
Your team has unlearned.
Now you know great accomplishments aren’t done alone: it took the combination of Luke, Leia, Han, and Chewie to find success amidst challenge.
You too, are experiencing success.
After one year of attribution, marketing budgets surge by 30%.
Why?
Because ROI on media spend increases 15-44%.
A 3x return on investment in attribution…which is a better return than Warren Buffet.
You have become the change you want to see happen in the world.
You have learned.
You have earned.
You no longer work for marketing, because marketing works for you.
The force…is with you.
 

Why CTR & VTR Are Dead Metrics in Programmatic Display

Marc Pritchard now knows it. Any serious marketer (P &G) wouldn’t be caught dead using click-through-rate or view-through-rate in programmatic display. CTR and VTR are rubbish.
Here’s why:
CTR
A simple formula of clicks divided by views
Click: The Numerator
Clicks are massively overstated at least 50% because of botnet click fraud.
Thus, if the numerator (clicks) is wrong, it throws off CTR wildly.
100 clicks/10,000 views = 0.01
50 clicks/10,000 views = 0.005
>> By the error that botnet click fraud creates, the difference of real CTR is 2x.
That’s just the numerator…
Impressions: The Denominator
“Impressions” are not actually viewable…and in order to have a valid click-through-rate from a human, impressions need to be viewable (in other words, how can a human click on an ad if she can’t see it).
Impressions, are simply auction wins (this is how the IAB defines an impression: when the auction is actually won and the vendor got the right to serve the display ad).
In programmatic display, however, 68% of impressions are not viewable (C3 Metrics data since 2013; comScore cites 69% in programmatic).
Because impressions have to be viewable to have a valid click-through rate, the denominator (views) is wildly thrown off.
100 clicks/10,000 views = 0.01
100 clicks/3,200 views = 0.03125
>> By the error of viewability versus impressions, the difference in the real CTR is 3x.
That’s the error in the denominator.
The fraction is wrong in both clicks and impressions.
The problem is, you don’t know if your clicks are wrong, or your views are wrong…so the data calculating the most basic metric in display: is now rubbish.
VTR
View-through-rate is the metric of last decade. Just because a retargeting tag is present, is only an indication that the person went to the website, and a vendor slapped a retargeting cookie on them.
VTR is based on the “impression.”
But an impression has nothing to do with viewability. Again, impressions, are simply auction wins (this is how the IAB defines an impression: when the auction is actually won and the vendor bought the right to serve the display ad).
VTR does, however, require a cookie be dropped.
Because cookie drops have nothing to do with viewability, it’s like driving a car without your gas tank
“full” indicator never working.
Simply search “nested iframes cookie bombing” to get an in-depth explanation of how cookies are dropped all the time without a display ad ever being displayed.
As one CMO said, “if I’m getting a 4,000% ROI by way of view-through conversions, I’d be swimming in hundred dollar bills and driving gold Rolls Royces.”
View-through-conversions are the metrics for amateur marketers.
The Answer?
Yes, it’s self-serving.
But attribution at a user level which has built-in viewability, fraud controls, false positive controls, which also fractionalizes all media in an AWACS-like platform is the answer.
One referee.
One referee that’s not distracted by garbage data or garbage claims.
One referee that can model accurately and fast.
One referee that delivers a 3x ROI for the investment in that referee.
One referee that allows CMO’s to keep their job longer and have their budgets raised 30%.
A referee that’s not owned by a media company.
A referee that has instant replay.
A referee that cares about making the right call.
A referee trusted by CFOs.
A referee relied upon by CMO’s.
Let’s play ball.