Can You Get 3x ROI on Your Attribution Investment?

Berkshire Hathaway delivered a 20% annualized return from 1968-2011. Icahn Enterprises for the same timeframe delivered a 31% annualized return.
But an investment in attribution gets you a better return than either Buffet or Icahn.
After 12 months of using C3 Metrics, clients realize 15% median improvement in the ROI of their ad spend.
The cost of C3 Metrics: 1-5% of media spend.
15% median return on an investment of 1-5% [15%/(1 to 5%)] = 3x return
Nobody’s getting 3x returns these days in the financial markets. Not the Dow Jones, not your bank, not Buffet.

C3 Metrics advertising attribution measurement COO Jeff Greenfield elaborates on the four keys to the ROI improvement:
1. Purified Data (Single Source)
The purified data quality of C3 Metrics single source platform are vastly different. Meaning, C3’s platform is not what’s called a “fusion platform’ which is simply a repository of data from different (non-integrated) sources, then trying to integrate them after-the-fact.
One example is viewability. In Programmatic, 51% of display inventory is not capable of being viewed. And when 97% of all data collected for a typical campaign is display…simple math tells you that 49% of you entire data set will be wrong before it even enters the attribution equation [97% x 51% = 49% wrong].
Imagine if 49% of all data used by CFO’s to report earnings were wrong?
C3 Metrics has its own display viewability tag traveling with every one of its clients display and video ads, an seamlessly suppresses false data from entering into its attribution equation.
This, along with more quality control measures (like C3’s FraudX™ and Programmatic PCP™), prevent false positives from entering the equation resulting in the highest data quality in the industry, unique to C3 Metrics.
2. Speed
Time is money. The longer you wait to get actionable data to shift media tactics or creative…the more frustrated you will be (and longer it takes to get ROI).
Speed at C3 Metrics begins with on-boarding and integration. It’s down to a science at C3 Metrics. The record onboard for a client is two hours (very unusual). On-boarding takes on average 2-4 weeks, with initial insights delivered between weeks 12-16 (2-3 conversion cycles required).
Speed continues with Bayesian machine learning.
3. Bayesian Machine Learning Algorithmic Model – Updating with Every Conversion
One of the key aspects of C3 Metrics is its state of the art Bayesian machine learning algorithm.
Machine learning updating with every conversion…versus having to wait months for a new model to be processed, human analyzed, loaded, checked, and reloaded, etc).
Bayesian modeling is the same modeling used by Google self-driving cars, and by Netflix to determine what original content to produce, and which actors to cast.
It is the most sophisticated algorithmic model (especially when applied with machine learning) in today’s industry.
4. Business Flexibility
Every business has nuances and differences. In the education vertical, enrollments are key, not just leads & applications. For this vertical, offline enrollment conversions are imported into the platform to capture and connect the most important KPI (paying customer).
Every business has its own nuances.
In addition, C3 can incorporate conversions which happen: over the phone or at a brick & mortar location…incorporating many components of key data which would otherwise go uncaptured and disconnected from the media funnel impact.
If it’s important to your business, it becomes important to C3 Metrics.
Tell your CMO to invest those marketing dollars in the stock market? There’s no demonstrated 3x return there.
But as Warren Buffet said:
“Price is what you pay, value is what you get.”

Why Marketing Budgets Surge 30% With Attribution?

It’s budget time: when CMO’s & CFO’s dance the dance of “chicken & egg.” If the CFO increases marketing budget, how many more eggs will be produced? And has the CFO heard it all before?
On the flip side, if the CMO has already demonstrated what marketing investments produce results, and at what ROI (down to the penny)…the CFO is happy to increase marketing budget.
In CFO speak: it’s called “matching principle.”
In CMO speak: it’s called attribution.
In shareholder speak: it’s called accountability.
CFO’s get fired up about marketing accountability.
With marketing accountability, marketers get more budget (and keep their jobs longer).
How much more budget?

After one full year of using attribution, marketing budgets increase 30%.
After two years, they increase 39%.
Because CFO’s see a regular, accountable process of advanced measurement similar to what CFO’s refer to as activity-based accounting. Attribution is the CFO’s libation. Happier because a process of accountability exists (and budget increases).
But really…why?
Because the median ROI improvement of marketing spend is 15%
The math: 15% improvement which costs 1-7% of spend = about a 3x return on money.
Is the stock market delivering 3x returns?
Are banks delivering 3x returns?
For CFO’s, attribution = libation because of process and proof.
Pour a little more: 30% more.

What's the Programmatic Fraud Fooling Even You?

It hides beneath the surface. It can be twisted to resemble a lapse in judgment, or the equivalent of leaving your front door wide-open while you’re on vacation.
It’s called daypart fraud.
Some context:
Every day, your DSP or network sends a report containing all the normal La-De-Da things:

  • Impressions
  • Cost
  • CPM
  • Clicks
  • CPC
  • CTR
  • Perhaps viewability
  • Creative, size, etc

So what’s the problem?
All this data is sent by day…calendar day.
However, a ton of programmatic fraud starts precisely at 12 midnight.
When a brand-new day starts for networks/exchanges/publishers (at midnight): they need to make money quickly. Consequently, they open their gates to traffic which is often fraudulent.
But when you look at impressions by hour: you may discover things brands hate…

Looking at impressions by hour: 82% of this campaign are shown between 12am-3am, peaking at 3am.
12 midnight is the hour that publishers and exchanges wake up and say: “I have to make money today…starting right now!”
Botnets often come out and play between 12 midnight and 5am (when demand for inventory is low and publishers/exchanges will take anything they can get).
Very sophisticated fraudulent botnets.
Does anyone pick up on this? Not when the standard report format is by calendar day versus hour.
Because C3 Metrics deploys its own proprietary viewability tag, piggybacking with every advertiser’s impression, that impression data is summarized in an easy seen format by hour.
In the above example, simply look for abnormally shaped spikes peaking in the wee hours: and bingo, you’ve got daypart fraud.
Now, is this simply a lapse in judgment (someone forgot to deploy daypart controls when setting up the campaign)?
Or is it a combination of someone forgetting to close the front door while leaving for a very long vacation…inviting bad things to happen?
Either way, this is when fraud happens, and agencies, and advertisers typically have no idea it’s happening because the data they get is by calendar day, not by hour.
“Daypart fraud hides in plain sight, and it happens a lot more than anyone thinks,” said C3 Metrics advertising attribution measurement COO Jeff Greenfield. “if you ask for the same report you’ve been getting for years…but ask for it by hour, you’ll discover if it’s happening to you. We even have a standard email written up for advertisers to use when they send the screenshot to their network or DSP.”
Daypart fraud: it toys with the best of brands.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
All guesswork and no display makes Jack a botnet toy.
Scared now?

Is the Mother of All Direct Response Back | Mail?

Technology has changed Direct Mail. It used to be slow (months of lead time). But banks, credit cards, lawn care, and political campaigns still use direct mail with a twist.
The technology twists:

  1. Print on demand
  2. RFM database triggers
  3. LiveRamp/Acxiom cookie translation to household address
  4. Retargeting capability (essentially creating a qualified mail list)

But the two keys to direct mail at a whopping $500-$810 CPM are: your list, and your creative/offer.
Thanks to digital display and dynamic headline testing, mail creative/offer can be pre-tested in display (provided you’re measuring on viewable impressions).
But the list is still the make or break in direct mail.
Now with DMP overlays and look-alike modeling, list selection gets more refined.
New companies diving into this old medium:
Lob ($20 million funding)
PebblePost ($47 million funding)
Mail’s biggest thorn (besides high CPM) used to be reliability, but now mail data & technology companies like GrayHair software, track scans where mail is tracked down to the day before in-home (versus praying Alfred at the mail shop sent those trays this month).
A recent study conducted by Canadian firm TrueImpact showed higher recall of direct mail versus digital ads (75% versus 44%).
“At the end of the day, it’s not about CPM’s it’s about CPA’s,” said C3 Metrics advertising attribution measurement COO Jeff Greenfield. “now with cookie translation from the likes of LiveRamp, Direct Mail becomes very easy to incorporate into the customer journey funnel, and thus attribute.”
What’s old is new (with a twist of technology).
Shaken…not stirred.

Attribution #1 Item For 73% of Marketers Now | Nerds > Words

It’s the biggest growth item according to eMarketer’s survey of “Topics That Will Command Most of Marketing/Media Practitioners Attention in the US” performed by the IAB and Winterberry Group.
It’s not only #1 for 2017, but the highest growing item as well.

It doesn’t take a genius to figure out why.

  1. If you’re advertising in seven different media channels, each channel has as least of 1/7 chance (14.2%) of getting credit. To ignore the elephant in the room with last click getting 100% is…ignorant.
  2. Onboarding can take 2-3 weeks if you’re not using a onerous proxy-server attribution platform; then allow an additional 4-6 weeks for data to cook.
  3. Machine-Learning Bayesian modeling removes the need for only two models a year.
  4. Single-Source JS tagging platforms now allow data uniformity like never before.

In other words, big data (still big) is now manageable, actionable, and accurate (with the right platform). Keep in mind that garbage in, is still garbage out.
CPA now becomes ACPA. ROAS now becomes AROAS.
Nerds and Words come together.

Facebook Finally Fixes Dirty Little Secret | Accidental Clicks

This has been going on for years: accidental clicks from Facebook. Is it a product design flaw? Is it that big fingers on a small screen don’t mix? Regardless, Facebook is finally fixing it.
But what?
Data centric geeks have known about this forever:

  1. Log in to your Facebook advertising dashboard, and look at your total campaign clicks
  2. Log in to your third party research dashboard, and look at those same total campaign clicks from FB

Oftentimes, those two numbers are light years apart.

In fact, it’s not uncommon to see differences like the above (real scenario, real ad campaign).
But Why?
Facebook measures the click from their server side, and third party research practice is to measure the click after it fully loads in the browser page.
A huge amount of those clicks never fully load (as seen in the above data) for a variety of reasons:

  • The page load speed of the site for mobile is extremely slow
  • The click was perhaps accidental (60% of mobile display clicks are accidental in the industry)
  • Wireless connection speed for the user is slow
  • Consumer then exits, going back to Facebook

Is it Facebook’s problem exclusively? No.
Does it raise a ton of questions as to why Google Analytics or other third party vendors see one thing (using standard research protocol) and why Facebook’s server side click numbers are not just a few percentage points off within reason…but miles off? Yes, it does raise a ton of questions.
But now, according to yesterday’s Adexchanger article, Facebook is going to remove clicks and click cost for Advertisers if a user bounces in under two seconds after clicking on an ad.
A secret no longer. A discrepancy no longer.

Alive + Kicking – TV Upfronts Pull in $19.7 Billion

The upfronts just placed $19.7 billion worth of prime-time commercial spots on cable and broadcast TV networks, an increase of 5.9% over last year, according to the New Jersey research firm Media Dynamics.
Cable TV networks took in $10.6 billion of advertising commitments, up 7.6% from 2016.
The pendulum is beginning to swing back to TV (on a national basis) for two reasons:

  1. TV has 11.7x the reach of digital
  2. TV is also free of viewability and fraud issues still plaguing digital

At this year’s Advertising Research Foundation conference, Chief Research Officer of CBS David Poltrack presented findings from a study of 863 cross-media TV + digital campaigns.
For the top 60% of campaign performers: TV delivered 70% reach, but Digital delivered 6% reach.

Local TV ad revenue is another story (not so pretty).
In Comcast’s Q1 2017 earnings call, Steve Burke said, “national and local markets are quite different and exhibit different characteristics…there’s more weakness in local.”
Local’s Bright Spot: HH Addressable (For Cable)
For Cable and Satellite providers in the U.S., household addressable TV advertising is becoming a reality: over 50 million households are now HH addressable. Spectrum (Time Warner) rolled out addressable targeting in 11 million households. Comcast’s addressable HH count is only limited by the speed at which their set top box manufacturer in South Korea can make the boxes.
Local broadcast, however, may experience flat to negative growth on a same station basis as its largest vertical (Auto) shifts approximately 30% of ad dollars away from TV into digital because TV still has not provided the measurement, data, and attribution.
TV = Typewriter?
“Until very recently, TV has given its advertisers the same delivery report that it created on a typewriter,” said Jeff Greenfield, COO of advanced advertising attribution measurement firm C3 Metrics “here’s your GRP’s, here’s your unit delivery, here’s your cost per GRP…it’s essentially the same report from 30 years ago. But now, things are changing with the likes of C3’s DriveShaft. Cable and Linear TV providers simply have to catch up to the attribution metrics compared to the mountain of metrics in digital. Only then, will local begin to stem the revenue loss.”
National TV, however: alive and kicking.

The Robert Downey Effect | Attribution

What can marketers learn from Robert Downey Jr?

  1. Failure to admit there’s an obvious problem = bad things.
  2. Believing there’s a better way + deciding to change = great things.

From 1985-1990, the movies he stared in grossed an inflation-adjusted $529 million. But as his drug and alcohol problem persisted, movies he starred in grossed less and less.

In 2003 he hit rock bottom; finally admitting there was a better way…and decided to change.
Then his box office gross rose 11.5x
The same is true for modern day marketers. For those who admit there’s a better way (cross-channel attribution versus last click/single channel attribution), great things happen.
12 months after using C3 Metrics for attribution , ROI of media spend increases by a median of 15%.
“There’s a saying we have in attribution…you’ve got to go down to go up,” said Mark Hughes, CEO of advanced advertising attribution measurement firm C3 Metrics “once you see everything in a data cleansed, x-device joined, and cross-channel funnel, you can’t ignore unprofitable tactics which have to be removed. But then, you begin to rebuild and scale the profitable and healthy tactics.”

In the second year of using C3 Metrics attribution, marketers spend more. 30% more. In the third year of using C3 Metrics, marketing spend goes up 39%.
“We call this the Robert Downey effect,” said Jeff Greenfield, COO of advanced advertising attribution measurement firm C3 Metrics “after you admit you have a problem …last click, viewability, X-Device, rampant click fraud, single channel siloed measurement…and decide to change, things get much better.”
Just like Robert Downey’s career, the results of hard change are obvious.
But what’s obvious isn’t always apparent (before you decide to change).