Thanks for being here!
Announcements:
This is the 101st publication of the Alt Data Weekly. I’ve recently moved over to SubStack … please have patience with any near-term hiccups. Feedback appreciated.
You can now access 90 West’s consumer transaction data via our Insights Platform (powered by Exabel).
Theme that emerged in this week’s email is … make thoughtful decisions up-front to save yourself time in the future.
Quotes:
“The (ethical) bar here is not regulation. The bar here is setting an expectation with consumers and then meeting that expectation—and doing it in a way that’s additive to your brand.” - Anonymous senior executive quoted in McKinsey article
“…we’ve barely scratched the surface of what data can do in shaping the future of more personalized healthcare.” - Geoff Martha, Chairman & CEO, Medtronic
News Articles
Podcasts
Cool Charts
Final Thoughts (AI Advances)
Best Regards,
John Farrall
#1 – Madison Mae of Learn Analytics Engineering published It's Time to Prioritize Data Documentation. September 2022.
My Take: The past couple of weeks I have been seeing more articles about the importance of doing things the right way, upfront (theme this week!). Once you get going on your data journey, making changes gets more difficult. There is nothing better than having access to all the information you need at your fingertips. Having meta-data about your data helps you keep an eye on the big picture and makes it easy for non-data scientists & customers understand what you have and why you have it. Create a culture where documentation happens on the fly and is part of everyone’s daily process. The future you will appreciate it!
#2 – Seattle Data Guy published Onboarding For Data Teams. September 2022.
My Take: Another article themed around getting things done the right way, up-front. There is a lot packed into in this article. Of most interest to me was the idea that you want to minimize the pain of getting the new team member access. That should largely be done prior to their first day & great documentation helps.
Next is the big picture. Explain the business (often taken for granted) and run through the KPIs. This will allow for the first steps down the path of domain expertise. The goal is to build an environment where the new employee can become a confident & independent member of the team.
#3 – Alex Edquist, Liz Grennan Sian Griffiths, and Kayvaun Rowshankish of McKinsey authored Data ethics: What it means and what it takes. September 2022.
My Take: Another article about making good decision up front in your data journey. McKinsey defines data ethics as “data-related practices that seek to preserve the trust of users, patients, consumers, clients, employees, and partners”. The idea of intentionally creating a data culture upfront & with C-level leadership is important. Many will get so focused on their specific individual goals and assume someone else in the organization has vetted the data, data usage that falls outside your company culture may occur.
BONUS 1: Ed Lavery published in Global Banking & Finance Review Alternative data is an investor’s greatest ally against volatile markets. September 2022.
BONUS 2: Dan Joldzic of Alexandria Technology launched a new Crypto tracking service using NLP crowd sentiment … an innovative use of alternative data sources.
#1 – S&P shared a webinar titled Overcoming Data & Analytics Challenges: How to Maximize Your Data ROI. September 2022.
My Take: S&P makes the effort to speed the time between finding data and understanding if there is value. They discuss process for trialing new datasets. This takes time, effort, and expertise. S&P is attempting to make this process as easy as possible.
S&P is addressing the right problem. Finding and evaluating new datasets is difficult. I am not sure how this is made easier. If it is easy, everyone would be doing it. We have a long way to go.
There is no doubt firms are becoming more data driven. S&P platform seems easier & more straightforward than most, but I still think the real value requires the time, effort, and expertise that many firms still do not have in-house, even utilizing these platforms.
Highlights (54-minute run time):
Minute 03:00 – webinar starts, and agenda shared
Minute 05:00 – Hannah Sayson, product manager for S&P marketplace presents how to best use the S&P marketplace to find data
Minute 14:00 – demo of marketplace
Minute 17:30 – Shantunu Dobhal discusses process for trialing new datasets (“workbench” tool)
Minute 20:00 – demo of “workbench” system to allow the testing of new datasets
Minute 33:00 – John Schirripa shares an overview of data buying journey and ultimate delivery
Minute 38:00 – explanation of S&P’s efforts to make this (incorporating data into your current process) as easy as possible
Source: Selects slides from S&P’s Webinar highlighted above:
Most interesting to me: the center box below…tools are getting much better … that many jobs will not be required.
Most interesting to me: data sourcing is still a major problem, there is too much data!
Most interesting to me: data prep take too much time … tools are getting better to help with this! S&P is helping with this by prepping data for consumption.
Source: Speaking of tools getting better (see above) … A.I. (artificial intelligence) continues to advance. I had thought this would be one of those technologies that people talk about forever with no hard use-cases for the average person. I am being proven wrong.
The closer technology operates like magic, the closer it is to broad adoption.
I had written about this a few weeks ago on my personal SubStack and continue to mess around with these tools.
Some samples:
This is a cool one… you simply enter a key phrase or question and the article is written for you.
I requested bullet point answers & this is what I plugged in the interface:
"What are some advantages of being a Cleveland Browns fan?"
This is what the application generated for me:
Having a team to support during the NFL season
Feeling like you're part of a community
Feeling like you're part of something bigger than yourself
The feeling of pride when your team does well
The feeling of satisfaction when your team wins
Second example:
I asked for a Pros & Cons List:
“Should I be a Cleveland Browns fan this year?”
This is what it generated for me:
Pros:
You can't get much lower than the Browns, so you can only go up as a fan.
There's a sense of camaraderie among Browns fans because they can all commiserate with each other over the team's struggles.
You'll be able to say you were a fan of the Browns before they were good again (assuming they eventually turn things around).
Cleveland is a great city, so being a fan of the local team is a way to show your support for the community.
NFL football is enjoyable regardless of who your favorite team is, so you might as well root for the home team!
Cons:
They have a history of losing: The Browns have had a long history of losing, dating back to their founding in 1946. In their 74 years of existence, they've only made the playoffs 20 times and have never won a Super Bowl.
They're not currently good: The Browns are coming off a 3-13 season and haven't had a winning record since 2007.
They're not particularly fun to watch: Even when the Browns are winning, they're not necessarily an exciting team to watch because they tend to rely on defense and running the ball rather than throwing it around the field.
Enter natural language … out comes a picture.
Input: “impressionist oil painting flowers yellow with green stems”:
Output:
There are some cool demo videos. Code is created using natural language.
This was easy to use. You enter a couple data points and the system generates A LOT of logos (some good…some not so good).