|
|
This edition of TMAI Premium is free for Standard Edition subscribers.Â
Over the last few weeks, TMAI Premium editions have tackled deep into the world of applicable artificial intelligence, sharing specific recommendations for activation in six major clusters that will drive profits. We've shared how to apply AI to humble efforts like picking KPIs, or leveraging brand new ones never possible before. Last week we focused on leadership, and how to craft one's user manual and put it to productive use.
You can upgrade to TMAI Premium here. All Premium revenue is donated to charity. | | TMAI #363: The Modern Analytics Maturity Model -P1
| This is a common pattern across companies, agencies, non-profits:
1. Tools exist to collect and provide data. Free. Paid. Numerous.
2. Multiple methodologies are deployed to process data. Free. Paid. Numerous.
3. Teams exist, often fractured, often scattered across the org to do data. Paid. Numerous.
4. People exist with varied skills, not just in the “data team,” to report. Paid (well, hopefully!). Numerous.
5. Agencies are hired to outsource cheap data work, and/or drive data's innovative learning agenda. Paid. Numerous.
If it is a small company, it might only have 1, 2, 3, 4, and fewer of each. If it is a large size company, it will have the entire spectrum of 1, 2, 3, 4, 5, as outlined above. In the top 200 companies on the planet, there will be multiple copies of 1, 2, 3, 4, 5 spread across divisions, sub-brands, regions, countries, and more.
In 2023, it is exceedingly rare for a company to be data-poor.Another common pattern is that it is unlikely companies are using insights and actions derived from data to consistently solve for a local maxima. It is extremely rare for a company to consistently use data to solve for a global maxima.
Yes, some have a bit more, some have a bit fewer data/tools/resources.
What binds you and me and everyone else is that you have more data than you realize (or honestly, even wanted in the first place). You are probably spending far more money on the data ecosystem than you realize (small or large company).
: ( | The Challenge in Simplifying Complexity.
Deep in my soul, I believe that a radically honest assessment of real reality can be a game changer - for those who want to change the game, leave their mark on the work world.
So, a couple of months ago, on an 11-hour flight, I set about trying to build something that would empower the game changers to change the game – smartly, faster.
The challenges to solve:
Understand of the full spectrum of possibilities today via analytics – from beginner to the bleeding edge.
Look across each dimension of the data ecosystem to ensure a human can assess both the tactical and the strategic. (A hard problem to solve.)
The goal is to boil the ocean, but to not make it feel like boiling the ocean – because that feels scary.
Expand minds. You can only grow/go, if you know that that is a possibility.
Every industry/team/company is a special snowflake. Build something that is extremely specific to everyone, while ensuring it does not feel generic (because that will impede adoption).
Fill it with hope. That better is possible. That you can make step-wise progress.
This is the most difficult challenge because the reality is that the world of analytics is insanely complicated: Make things as simple as possible, but no simpler.
I call my magical solution: Modern Analytics Maturity Model (MAMM).
It is a framework through which you can understand the full spectrum of possibilities in the data ecosystem, assess your present reality (as honestly as you permit yourself to be), and know exactly what you need to do to move up and to the right (/towards the global maxima).
Like my other frameworks, (DC-DR-DA, ABO, IAbI, Impact Matrix, Incrementality, Reach Build Engage for YouTube, Global-Regional-Local org structure, See-Think-Do-Care for an expansive marketing strategy, Smart Clusters for a muscular data existence, AI Activation framework), my focus is on simplifying the think holding full confidence that your brilliance can easily figure out the ink.
I’ll introduce a chunk of the maturity model this week. We’ll hold hands, and complete the enriched learning in next week’s Premium edition. I’ll also share exactly how I am using it with clients to for a radical assessment of real reality, and extract 10x value from an investment in the data ecosystem. [Not a Premium member? Sign up here. Spend $100 to add a min of $1,000 to your annual compensation.]
[Bonus Reflection: Taking everything in your head, all your experience, all your failures, the seven challenges above, and converting it into something codified, simple, is a uniquely fun hard exercise. I recommend it as a way of stretching your unique brain, realizing how much you know/don’t know. Build a framework. Teach it. It brings such joy.] | | The Modern Analytics Maturity Model.
Evolution rules the earth.
We are all at some stage of the evolutionary process – human or data. : )
In my strategic consulting practice, I’ve defined six stages of analytics maturity.
0. Stand Alone: Early. No cohesive strategy or people investment. Employees make the best of whatever’s already there. A report from finance or if there’s an analytics tool – however implemented – or someone with a monthly subscription to SEMrush. Some decisions. Usually, un-integrated.
1. Descriptive: Central, often IT or Eng team owned, open a ticket, and we will make you a report existence. The focus: WHAT. Success is having some sense of some key metrics in the business. Fewer silos. Understanding data, figuring out what to do is left to receivers of the, often, data pukes.
2. Diagnostic: The beginning of data being able to have an impact, because the purpose is WHY. Why do we see what we see in the data? The existence might not all be centralized, though it really helps. In addition to the numerous tables and charts, the dashboards have some words that share the insights. The minimum viable existence for an analytics practice is Level 2.
3. Strategic: As they say in America: You are cooking with gas. Silos break down (strategy, divisions, data, teams, tactics, decisions, politics). The WHAT + WHY are not answering hey, how am I doing with Paid Search. At the minimum it is hey, how is my unified Search strategy impacting profit (expansion of the altitude, and a smarter KPI!). Crushing Level 3 is unifying Brand + Performance analytics!
4. Predictive: Looking back and understanding outcomes is super cool, true glory is the ability to predict the future. It is scary because you will be wrong, it is simply a matter of how wrong. My objective for our teams: Be right 70% of the time, about hard decisions. It is an incredible competitive advantage because your competitor is either making zero hard decisions or not making them upfront. Level 4 makes you special, truly special, when you can do this not as a one-off, but do it repeatedly AND at scale.
5. Deep Learning: Over the last five years, machine learning algorithms have unleashed a glorious set of possibilities for anyone with lots of data. You’ve always gotten challenges from your CxO, and you don’t even know what to ask the data to tell you! Unknown Unknowns are, often, a solvable problem now. You can use a classification decision tree or random forest in the spirit of: Can you please figure out what’s going on in three years of data that explains our lameness (/also politely called ground truth). For Analysts & Companies, Level 5 is orgasmically magnificent – and immensely profitable.
What level reflects the real reality of your analytics practice?
Some of you can answer that question instantly.
Some of you are reflecting back to me: Well, it is complicated. One team has done one experiment. For Paid Search, we are using Google’s Machine Learning thing. We don’t call it WHY, but we have something a little like that. Does it count if IT owns analytics (no!)?
They are all good questions, and thank you for taking my question so seriously.
A good framework has more detail, more layers for you to work through, a bit more detail to help you assess real reality. And, I have all that (bunch of it today, below, and bunch more next week).
Greatness comes from high standards. With that high bar in mind, allow me to share this advice upfront: A chain is only as strong as its weakest link.
When assessing your real reality level, grade your weakest link. If one team has used ML once, it does not count. Assess your level by reflecting on the most common practice inside your company/team. What’s the most frequent occurrence? If you’ve once done a large-scale matched market experiment, but most data people work on producing reports based on tickets opened by the various teams… You are on Level 1. And, Level 1 is better than Level 0.
Here's your initial view of the beautiful beluga whale that is my MAMM… |
|
[Ok, so maybe it could use some whiz bang design. I invite your help. Also, some of you are experts in branding. If you have a snazzier name for the model, I welcome your suggestion!]
Let me first draw your attention to the, literal, bottom line, the last row. Business impact.
You would rather not be happy with the level you are at because a few additional smart (sometimes hard) choices will result in a radically higher impact. The difference: From: Send me the numbers. To: Why don’t you come present your recommendations to the board of directors?
Yummy, right?
Anchor your thinking in business impact.
Stepping back up.
In nearly every TMAI Premium, you’ve noticed my obsession with purpose. I find it is so incredibly clarifying for all manner of think and i​nk (ex: in my role as the Chief Strategy Office at Croud). Adding, who is our primary client, the individual we spend time with when we do this work, simplifies all manner of complexity.
The first assessment you’ll make on my behalf is: What purpose is served by a majority (60%) of our analytics investments in tools, people, processes? You’ll take that answer, and add who is our primary client to achieve deep insight?Level 0: Unfocussed purpose, mostly data in random places, relying on the goodwill of whoever has the motivation to find some data, so something with it.
Level 1: The purpose is to share WHAT is happening in the business, where possible. The results, when published systematically and in a timely manner, are used primarily by front-line employees.
Level 2: The primary client are Directors in a company. That is the result of data’s purpose being to explain WHY something is happening. You are personally spending time explaining the why, vs. simply shooting off pre-scheduled reports (which will have 90% WHAT and 10% WHY – when you try hard).
Level 3: An outstanding purpose: big business decisions, at a VP level, often impacting the entire division, product area, resulting in some changes to marketing strategy!
[Note: It is not that reporting is dead in Level 3. There is some, it is still powering $10 of impact. It is just not the primary purpose of the data ecosystem’s existence.]
Level 4: At least 30% of your data ecosystem investments (tools, people, process, agencies) are solving for known unknowns. “We keep failing at x, we just don’t know why!” You are having pre-scheduled monthly meetings with the CMO of your company, and their direct reports (who are possibly Very Scared that you might call them out, so they have spent a lot of time giving you “context” and less than gentle nudges to not only point out 90% failure but spend a lot of time on the 10% of their budget that delivered some success). Life can be difficult for the Analyst at L4. I have cuts on my heart to prove that. You want to be here because… Well… Just look at that number in the Business Impact cell!
Level 5: The scary CFO is your client! I love Level 5. The purpose is powering business altering decisions. You are playing with the most bleeding edge in data tech. And, your client is just as analytical (if not more so) than you are, just as logical, and, very often, just as less interested in politics and in your quarterly meetings frequently says: Can we just listen to the numbers, please? There is no upper limit on business impact when you are on Level 5. Your client will bring high standards, massive influence, and ask scary questions. It is such fun – if you are a let’s change the game type.Phew!Slow down, catch your breath.Reflect.What's your assessment of your company's level of analytics maturity? | Next Week.
There are nine more dimensions for us to lean, each sparking a new round of reflections, giving you an in-depth view of the maturity of your data ecosystem.
For example, #6 is the type of business decision you are empowering today.
Obviously, for each level, one of the dimensions is which KPIs (#9) you are using – that tells your maturity instantly! : )
[Need a Premium subscription? Here.] | | Bottom line.
There is a lot more that your data ecosystem can deliver when it comes to business profits.
Step one is to understand the chaos of uncoordinated activity happening today, and what it reflects about your Sophistication, Purpose, Client, (and additional insight-sparking dimensions).
Step two, taking steps to change your real reality and thus delivering on the true promise of data.
Step three, making plans to go see real beluga whales in the Arctic because you now have both the time and the money to do it – because you live on Level 5!
Carpe Diem.
-Avinash.
PS: For a higher resolution version of the MAMM: Right-mouse click, open image in a new tab/download. If you are a Premium subscriber, email me after Part 2 for an Excel version, which will also have the accordion slices. | Committed to investing in your professional growth?
Upgrade to TMAI Premium here - it is published 50x / year. | |
|
|
[ Subscribe] Â |Â Â [ Web Version]Â Â |Â [ Unsubscribe]
                            ©2022 ZQ Insights  |  PO Box 10193, San Jose, CA, 95157, United States of America |
|
|
|
