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

Over the last few weeks, we've tackled particularly remarkable topics in TMAI Premium. How to analytically answer this question: Is TV dead as an advertising medium? We dived into data visualization lessons, via a climate change dataset. TMAI #279 covered why 90% of your current analytics Metrics are head fakes, and recommended Most Important KPIs. Last week, a profoundly underappreciated root-cause of analytics failure — and fixing it.

My hope with Premium is to deliver a non-normal impact on your career. If that sounds exciting, you can upgrade to TMAI Premium here
All revenue is  donated to charity. Thx.
TMAI #281: Ideal Habitat for Data, and Data-People!

It’s the culture, stupid.

No Analyst would say that decision-making in their company meets their expectations for being data-driven (or, as I prefer it, data-influenced -TMAI #270).


As always, there might be outliers. But, data-driven companies are not the norm — to the degree it is possible to be.


When the realization dawns that data efforts are falling short, the strategy instantly activated is to undertake a complicated and long project to figure out:

What is the missing data (metrics, dimensions, sources)? 

What are the missing tools (analysis, visualization, data puking)?

(On occasion) Who are the missing people (analysts, programmers, change agents)?

This instinct is understandable.

We are innately self-critical. We all live in an imperfect data world, with incomplete tools sets, and frequently don’t have enough people.


And, off we go. Down the wrong path.


We believe that if only we can get more data, more tools, and more people, the organization will become more influenced by data. It will deliver a reduction in sub-intelligent decisions.

The above becomes a five-year quest at the end of which all data goals of five-years prior will be accomplished (more data, more tools, more people), and frequently the company is still at the same level of data’s influence. Very little.


: (


If you conclude that your organization is not data-influenced — or it completely lacks the ability to answer the super hard questions (TMAI Premium #260 & #223) — then you might want to question your reality:

1. There is gold in the data, tools, and people you already have in your company. Perhaps not 24k gold, but surely 18k gold (still precious, and strong!).

Critical thinking request: Why is it that 18k gold is still not being appreciated / leveraged / impactful?

That’ll lead to a root-cause realization...

2. You don’t live in an organization where data can have any deeper influence/business impact.

BOOM!


You don’t need a five-year boondoggle. You need to create a healthy habitat that nurtures and protects the data.

How do you create this ideal, healthy habitat?


You  need to address these three non-data elements: Leaders. Process. Culture.

When those elements approach close to optimal, even 18k gold can help you build a competitive advantage in your industry. 


[Obviously, there is always room to make that competitive advantage sustainable by moving to 24k gold. But. The perfect does not have to be the enemy of the very good.]


The three contributing elements (leaders, process, culture) usually have multiple corrosive sub-dimensions that contribute to data’s lack of influence (despite loads of data/metrics/reports being everywhere in the company). 


In this newsletter, I’m sharing four patterns that occur most frequently in non-ideal habitats for data, and data-people. Across those four patterns, I'm sharing six questions to spark introspection and ideas to identify your organization's specific issues. Fix the four patterns, unlock a difficult to replicate competitive advantage!


My not-so-secret agenda:


To increase your emotional well-being by reducing the number of futile five-year analytics boondoggles in this universe.

It is important to point out that I am not clueless. I recognize that most of you might not actually have the influence in your organization to actually fix these problems. Still. Knowledge is power, as they say.


Here’s how to recognize root-causes (related to leaders, process, culture) that get in the way of data’s influence:
Solar Eclipse.
Pattern 1: Lack of Sunshine.

How much opportunity is there in your company to tell the emperor he is naked? How about the courtiers?

Does your organization have a monthly / quarterly assessment of performance against KPIs, and targets for those KPIs? Crucially, is that assessment done by a central/neutral team or a captive one?


Are there consequences for missing targets/goals for VPs (or higher)?  (Not people being fired, but consequences nonetheless.)


If these three ingredients don’t exist to some degree, data will have very little impact on the business. 


It is not uncommon for cultures to exist where:


No one will tell the CMO their idea is not very good, or 
… a campaign fell well short of expectations (or outright sucked), or 
… what worked 15 years ago is no longer relevant, or
you catch my drift… There is fear.

If truth can’t peek its head up, even the most brilliant data-insights will wither underground.


Q1. Can you tell the emperor that he is naked?


(Emperor is anyone +2 levels up from you.)

As companies grow, it is very normal for there to be politics and in managing up and managing laterally.  To avoid this situation, high-performance companies have independent, often central, assessment of performance. This results in a consistent, non-manipulated, audited cleanly, assessment of reality.

This is crucial for the culture, and frees individual team / division / group analysts from grading their own homework, or being fearful of their own leaders if they speak the truth.


When a culture has a common assessment of reality, you can bet your bottom data will have influence. Else, everyone knows the portrayed reality is fake… Data-influence withers on the vine.


Q2. Has your company invested in an independent, consistent, cross-team/division assessment of reality?


They say sunshine is the best disinfectant. I’ve come to identify that that is only true if there is a closed loop that ends with accountability. 

Purpose > KPI > Target > Execution > Assessment > Accountability vs. Target/KPI > Purpose.


If this closed loop exists and (independently assessed) success is identified and publicly rewarded… data will have immense influence (even 18k quality data). 


If this closed loop exists and (independently assessed) failure is identified and publicly communicated with changes being made… data will have immense influence (even 18k quality data).


If quarter after quarter, success after success (or failure after failure) there is no public accountability, even 96k gold quality data (vs. 24k) will have the equivalence of mud.


Q3. Does the closed, purpose > accountability loop exist in your company? 


Without sunshine being propagated, data, no matter how fantastic, will rarely earn influence that can deliver material business impact.


Answer the three questions above, then fund more data projects (if the answers are f’ yes!) OR fix the root causes because more data projects will only mean more frustration.
Missing Strategic Ideas.
Pattern 2: Great Measurement, Shoddy Strategy.

I had a funny moment with a CFO in London a few years ago


We’d worked very hard on a super sophisticated, difficult, project to assess the offline impact of digital investments. It had taken us six months. We had to build innovative measurements that truly were cutting edge. I was so proud to present both the methodology and the results.


The CFO was sincerely impressed with the first part. 


Then they said: “But Avinash, the results are so poor!


And I said: “Yes. I’m afraid, great measurement can’t make up for poor business strategy.


They smiled.


The focus shifted from me (data) to the VP (strategy). Hard questions followed.


It was perfect. It was precisely what should happen.

First lesson to glean: It is extremely rare to get this kind of exposure for the work we do to the very top of the company. Seek it. It is a priceless influence for data.


Second lesson to glean: You have to have a great business strategy, a great marketing strategy, a great product strategy, etc., etc., etc., for data to show anything of value. If your strategy sucks — as it often does — there is only a little that data can do to power success.


In fact, if strategy is consistently missing, or crap, great data analysis easily gets painted as the enemy for constantly being a downer and telling us nothing is working, why don’t you try harder to find what worked, why do you need to be so difficult, you are not a team player!!!


In small and medium-sized companies, this is less of a problem. Failure is public, the cost is huge. Sunshine is everywhere. You change or you die. 


In larger companies, in multinational companies, a lack of strategy is a surprisingly common problem (a lack of closed loops does not help). In that environment, more data, smarter analysis, brilliant recommendations, will have almost no impact. 

Q4. Do you really have a data/tools problem, or you have a missing biz strategy problem?


Great analysis cannot make up for a lack of business strategy.


Get strategy, before you get more data/tools/people.
Fragmentation.
Pattern 3: Analytics Resource Fragmentation.

No matter how many people your company has dedicated to Analytics, they are likely not enough.


There never are!


: )


Hence, it is crucial that Analysts — data teams — that you do have are organized for impact. 


In my experience, an optimal org structure is 10x more important than constantly chasing more data collection, more BI tools, and so on and so forth. 


When Analysts are not organized optimally, here are the resulting ills: 

A. Invent once, scaled infinitely won’t happen. 
B. The sum of parts, 1 + 1 + 1 + 1, will be 2.5, less than the whole. 
C.  The number of methodologies, visions, execution strategies will sprout like mushrooms on a log in the darkest part of the forest. 
D. There will be unclear career paths for Analysts, which will likely lead to attrition.
E. (Obviously) Data’s influence with VP+ will be as abundant as rain in the Sahara.

Pause. Reflect.


Q5. How optimally are your analysts organized?

If there is no hope/initiative to fix this (because of company politics, lack of leadership courage, whatever), the chase for ever more data and reports will add marginal value.


If there is no hope/initiative to fix this (because of company politics, lack of leadership courage, whatever), the chase for ever more data and reports will add marginal value.

[Bonus tip: If the only way you can scale analytics in your company is to keep adding more and more headcount, something is wrong with the org structure.]


This answer is so important, I’d dedicated the entire TMAI Premium last week to this issue and the approach to optimization. (See #280.)


Let me summarize.


There is an optimal org structure, at each stage of a company’s lifecycle. 


There are three basic analytics org structures:


1. Centralized (small companies, young analytics practice, early in sophisticated thinning)

2. Decentralized (rapidly growing company, divisional free for all, massively empowered local teams)

3. Centralized Decentralization (mature companies, invent once and scale infinitely mindset, strong central core pushing the envelope, local pushing contextual relevance and agility)

Each option above is right at a certain time (see last week’s newsletter). It is typical to go from 1 to 2 to 3 as you grow (from hundreds of thousands in profit to millions to tens of millions).

Once you are above a certain size, you want to make sure you settle at 3. Have you?


[In the spirit of leadership, Premium members, please, please, revisit #264 for the least desirable leader for Analytics teams and paradoxes for successful leaders!]
Looking Backwards, Moving Forward.
Pattern 4: Data Only Looks Back. 

Our reports typically look backwards. Almost all of yours as well. :)

 
One everyday example: How did my campaign perform?


This is normal. It can do a world of good. Especially if combined with a snap by Thanos that will make the above root causes disappear.


But. There’s one more signal you are looking for before investing in large data/tools/people projects: You start getting asked forward-looking questions.


Some examples:


How much budget should I ask for our department for next year?
What is the likelihood that my campaign will hit its pre-set target?
Where is the next tranche of our high LTV customers?
Is this creative likely to deliver above-normal results when we launch it on TV?
What happens if I cut our Facebook spend by $10 mil? 
What happens if I increase spend on TV it by $25 mil?

All forward-looking questions. Unpredictable. Sometimes ambitious. Almost always requiring sophisticated analysis. Even advanced algorithms.


Data will rarely achieve higher influence than being asked forward-looking questions because the norm is to make all these decisions based on gut or tradition or just because.


If you are being asked forward-looking questions, you are being invited to meetings by the Finance team, the Strategy team, and others who are not really your normal data customers. Oh, and you are being invited to meetings where you and I might be the most junior person around the table.


You’ve arrived.


Data has influence.


Data has trust.

Go invest in the craziest data project you want to, buy the most expensive tools mankind will make, hire analysis ninjas as double the salary they will get anywhere else. It will all have super-high ROI.


But. First.


Q6. Are you being routinely asked forward-looking questions? Are the answers actioned?


The killer question.
Ideas.
Bottom line.

It might seem odd that the author of two best-selling books on data — 14 language translations! — a super popular blog, and a 50,000+ subscribers newsletter on analytics is telling you: Hey, don’t do massive data/tools/people projects.


It is the result of age/experience.


Data is a lot less transformative merely by being available. Like a seed planted, it needs optimal soil, it needs nourishment, it needs tending and care, it needs sunshine (!), and other elements that allow it to grow and thrive.


Simply using better seeds without the above elements  is producing waste. You need to both build the habitat on arable soil, and plant the seeds. 


Before you go off on yet another data capture/processing/puking effort… Stop. Ask the above six questions. Reflect on the answers. Fix the root-causes. Now plant the seeds.


It is not easy. But. There is no other choice.


Genuine personal joy will follow from genuine impact on the business.


-Avinash.

PS: Yesterday, I bumped into a non-analytics story illustrating the devastating impact of not building an ideal habitat — in this case for innovation:

How IBM lost the cloud. 
Insiders say that marketing missteps and duplicated development processes meant IBM Cloud was doomed from the start, and eight years after it attempted to launch its own public cloud the future of its effort is in dire straits.
Committed to investing in your professional growth, and transform your impact/comp?

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

©2021 ZQ Insights 
|  PO Box 10193, San Jose, CA, 95157, United States of America