The Marketing < > Analytics Intersect, by Avinash Kaushik
This edition of TMAI Premium is free for Standard Edition subscribers. 

We've had a banger start to 2022 in Premiumlandia. We tackled the application of brilliant analytics to B2B companies, a 7-step approach to unsucking your advertising, learned the power to simplicity with three immediately actionable examples that drive extraordinary change, and, last week, my fav, how to approach to decision making when every tool/methodology you use gives you a different answer!

To earn a giant job promotion/compensation bump, upgrade to TMAI Premium here 
- all Premium revenue is donated to charity.
TMAI #297: CPIS | A Revolutionary Efficiency Metric.

Today, a short story about Marketing’s incremental value, but first, a succinct definition to get us on the same page:

Incrementality: Business results from marketing tactics that would not have occurred otherwise.

When you report data out of Adobe Analytics, or run a report in your CRM system to assess how effective the coupons were that you mailed to your customers… You have no sense of the incremental success or failure of your efforts.

Ex: The coupon redemption shows that 82,000 phone sales were driven by the coupon. CELEBRATION TIME!! Wait. Wait. Wait. Some number of those people would have bought the coupon anyway. Right? So not all 82,000 sales were incremental.

Ex: Paid Search Ads report shows that 13,498 conversions were driven by Paid Search ads for brand terms. CELEBRATION TIME!! Wait. Wait. Wait! Given that you also rank #1 in Organic results for those brand keywords, it is fair to assume not all the 13,498 sales were incremental. Some (/many) of you would have gotten those sales anyway. Right?

A subtle change of posture, a dramatically different view of reality. Inception. :)
A Short Primer on Measuring Incrementality.

In my work across different companies and clients, I’ve used numerous strategies to measure Marketing’s incrementality. Some of my favorites include…

1. Running Matched Market Tests (MMTs) in a whole bunch of cases (incrementality of TV advertising, effectiveness of coupons, digital’s incremental impact on store sales, etc.). We have a synthetic algorithm that helps us pick matched markets, that we split between test and control, and we measure incrementality.

2. Building Media-Mix Models using sophisticated machine-learning algorithms. These algorithms allow us to understand the incrementality of the entire cluster of marketing spend (online, offline, brand, performance, email, whatever we spend money on). Oh, and we can then split by channel, and get incremental Sales, and CPS, for Google, TV, Facebook, Print, and everything else.

3. Using tools like Conversions Lift Experiments, which Facebook and Google and others provide freely. In the context of that channel, for conversions that happen online, we can measure incrementality.

4. Occasionally, we inject variations in our spend patterns over time (and maybe channels) and then use advanced statistics to identify incrementality across those variations. I freely admit it is not easy, and you have to have scale to pull it off.

There are other strategies we are experimenting with all the time, as I’ve shared in past Premium editions of this newsletter (subscribers, see: #233 Solving Incrementality; #260 Analytical Strategies with Data Privacy Guardrails; #271 The Future of Analytics). 

Broadly, the use of super-advanced modeling and sophisticated experiments covers almost all applications of incrementality measurement.
An Implications of Incrementality Story.

Being able to do all of the above at scale means, among other things, you can be super cool and introduce super sexy new metrics for our Marketing spend.

We are used to reporting Cost Per Sale (CPS) for Facebook or Print or Radio.

If you have an advanced implementation of incrementality measurement, you can instead use a super sexy new metric: Cost Per Incremental Sale. CPIS.

True story…

For an online channel X, the number of sales claimed after proper application of Data-Driven Attribution (not Last-Click!) were 10,000 units at a CPS of $90.

Being fully aware that measurement without Targets is a horrendous waste of time, there was a Target to hit for CPS: $120.

$90 < $120.


Wait. Wait. Wait.

By now, you know what’s coming.

The client had implemented incrementality measurement for this channel. Nothing sexy. We simply used the Conversion Lift Experiments that they provide right out of the box. Easy peasy lemon squeezy.

It helped identify that 8.7% of all claimed conversions were incremental.

Translation: 870 of the 10,000 claimed conversions would not have happened without Paid Ads on channel X.

Three seconds of computation in your head later… The Cost Per Incremental Sale is $1,034.

Not $90.

$1,034 > $120.

: (

It does hurt when you first get the CPIS number. Not gonna lie.

But, from a fiscal responsibility perspective, it is good to know what the reality is. You can't improve what you don't know.

Obviously, your CFO is delighted at the sophistication of analysis that can allow her/him to have this depth of clarity about Marketing’s effectiveness. (Your CFO's love - attention - is worth its weight in gold.)

Your life objective, if you consider yourself anything close to an Analysis Ninja, should be to measure all three types of incrementality. And remember, Attribution is NOT incrementality.
Your CPIS Sucks, Time for a Big Mistake.

Allow me to be clear on what your Cost Per Incremental Sale metrics is saying.

I’m walking into the Macy’s store already. A Macy’s employee stops me at the door and gives me (on average) $90 in dollar bills before I walk in!

I was going in any way. I was going to buy a shirt anyway. They still gave me cash from their Marketing budget to walk into the door.

[Yes. Imperfect metaphor. Macy’s is not giving me, the customer, the cash, it is often giving it to the advertising vendor and only giving it to me directly in the form of coupons, promotions, rebates.]

Obviously, handing cash to people as they are about to do business with you anyway is profoundly wrong, and heartbreaking.

When the Finance team, Strategy Team, or the CMO understand this utter lack of incrementality (and CPIS), they make a big mistake.

The instinct is to look at this extraordinary behavior by the Marketing team and cut their budget.


If you have no faith in your team, if you feel your Agency has a causal relationship with your suckage… Then by all means the least you can do is stop the bleeding.

Cut. The. Budget.

But, realizing you get the team/Agency your incentives create, and hence the problem could be your leaders who might be fueling the suckage with the incentives they’ve created… cutting the Marketing budget is an action akin to cutting off your legs to run faster.

When you see a CPIS, you should take a different approach.
Your CPIS Sucks, Time for a Big (Profitable) Fix.

Assuming you believe in the skills your team has, assuming you believe your Agency has your best interests at heart (double check both)...

If you find your CPIS to be out of whack, your instinct, the Finance team’s instinct, the CMO’s instinct, should be to *incentivize better Marketing.*

When you have terrible CPIS numbers, as above, it is simply an indication that your Marketing team is executing a strategy that is either lazy, unimaginative, or constrained.
A lazy strategy is targeting an audience on TV who are already your customers, or still believing that big events (like Oscars or Super Bowl) are all you need to buy to have effective reach, or your Search strategy is basically top five brand keywords for your company, or dedicating 25% of your Marketing budget/team to Organic Social, etc. All. Lazy. Activity.

IF this explains your terrible incrementality results: 

Your Marketing team is set in their ways, the last new idea they had was in 2011. They need a metaphorical kick to work hard.

Moving on from the lameness of Organic Social is hard work (because your CMO still believes in it – with no data).

Creating a profitable non-brand keyword strategy is hard work.

Leveraging all the new machine learning based auto-optimizations (from link text to bids to conversions) is hard work.

Having an always on spike and sustain TV strategy is hard work.

Identifying where your incremental customers live/express intent takes a lot of hard work.

Create incentives to, well, incentivize hard work.
An unimaginative strategy is one that has not figured out that you don’t win TikTok by running pimpy ads – you make TikToks, not ads, or still relying primarily on Print and Radio or Trade Shows and industry publications to drive new business, or chasing just “link building” to drive SEO results, or running TV ads as YouTube ads, or, OMG, still managing your AdWords campaigns by hand because you distrust machine learning, etc.

IF this explains your terrible incrementality results:

Your Marketing team requires a massive and urgent infusion of new objectives, targets and, most crucially, new blood. They’ve flat lined and need a metaphorical defibrillation.


Hire new people (with a lot of experience and with none at all).

Hire leaders with skills beyond effectiveness in managing up the chain.

Shut down an entire team’s effort for six months (remember, it is not incremental in any way).

Reward people who’ve taken risks and spectacularly fallen on their face a couple times because at least they are trying.

Destroy a culture that sets low, easily achievable targets.

Any VP that has been in the same role for five years in Marketing needs to go lead a different team (reboot the hunger).

Fire one of your six agencies and hire a new (hungry) one (or bring that work in-house).

If you want incrementality in an unimaginative marketing strategy, deliver defibrillation.
A constrained strategy is exclusively a problem created by company leadership. Your CEO might have an incomplete view of Marketing’s purpose (ex: they think Brand Marketing is a bunch of hooey), constraining their ask of Marketing. Your CSO believes the only purpose of Marketing is to funnel leads and hold one large event each year. Your CMO might still be entrenched in Marketing strategy and tactics that made him initially successful 15 years ago (or worse, jump from fad to fad every six months!). 

It is very hard to operate under the stewardship of such leaders, and drive truly incremental impact from Marketing… Because you are operating under a constrained set of possibilities.

IF this explains your terrible incrementality results:

Your leadership needs an incrementality obsessed CMO with fresh eyes, an open mind, and a willingness to create a culture of truth-speaking. Or, you require a CEO educated in, and appreciative of, the long-term brand value of exceptional Marketing and short-term profit value of the same. Or, you need an influential CFO who will deliver a shock to unconstrain Marketing.

I’ve given you very specific recommendations for a lazy strategy, an unimaginative strategy, but I really don’t know how 75% of you reading this newsletter will fix the constrained strategy problem. (25% of you who are CMOs know exactly what to do.)

Perhaps two things: 

1. I urge you to recognize when you are in a constrained strategy situation, and know that your Marketing needs a complete reboot.
(And if that is true, your career is going down the tubes without a change in strategy. Gently, quietly and quickly find some place else to be.)

2. Gatekeepers – Sr. Directors, VPs – will stop you from speaking your CPIS truth to your CxOs. BUT if you get past the gatekeepers, use data and storytelling to help your CxOs appreciate the current heartbreaking reality. Speak to the inner child in them. 
Bottom line.

Did you think the tiny flip that created the humble CPIS metric could question the very foundation of your Marketing’s existence?

Cool, right?

You don’t have to chase insane complexity, revolutions can be powered by simplicity.

(See three more ideas in TMAI Premium #295.)

A closing tip. As always, you want the BFF pair of Effectiveness AND Efficiency. CPIS is an Efficiency metric. Pair it up with an Effectiveness metric like Sales Volume or Revenue or MAUs etc.

Your quarterly target is 1.7 mil Incremental Sales Volume AND CPIS of $45. Or, your goal for March 2023 is to deliver 800k Incremental MAUs at a CPIS of $865 each. 

Go get ‘em Tiger!

Committed to investing in your professional growth?

Upgrade to TMAI Premium here 
- it is published 50x / year.
[Subscribe]  |  [Web Version]   [Unsubscribe]

                                                        ©2022 ZQ Insights  
|  PO Box 10193, San Jose, CA, 95157, United States of America