Thanks for being here!
Announcements:
Our transaction data is now available at more granular “row level”.
Our Gift Card data now available for single merchants.
Theme that emerged in this week’s email is … the importance of getting those early wins & driving real business impact.
Quotes:
“Data Ecosystems are networks of organizations and individuals that collaborate to share data in a secure, ethical, and beneficial way for everyone involved.”
“Data initiatives are too focused on democratizing access to information and not focused enough on driving business impact.“
News Articles
Podcasts
Cool Charts
Final Thoughts (Dinner Parties & DWYSYWD)
#1 – Matthew Bernath published Data Ecosystems Enable Secure, Privacy-Preserving Data Sharing for the Benefit of all Individuals. March 2023.
My Take: There is a ton of value to be created by organizing & sharing data. From traffic management in cities, to improved healthcare, to better consumer experiences … all these day-to-day activities can be better if data was applied to how the services are delivered. And, similar to the message from last week, privacy is certainly an issue as people don’t want criminals stealing or to be “doxed” to the world, but I really think privacy is less of a concern, especially among younger people.
#2 – Emilie Schario published The Limitations of Data and Analytics. March 2023.
My Take: Focus on clear outcomes and start there. Pretty charts are one thing, but how are you improving the business? Answer the question ‘why?’ something is happening, not just that it is happening. Our margins are up … great … why? That marketing campaign provided a poor return … great … why?
This article calls to mind two themes that have come up repeatedly, the importance of early wins and domain knowledge. Understand the business well enough that, as a data person, you can come in, deliver something that will make people understand the value of the data investment.
#3 – Mohit Lad from Cisco ThousandEyes published Predictive analytics could be the future, but we must solve the data problem first. March 2023.
My Take: The holy grail of a data driven enterprise is the ability to seamlessly predict outages & system failures using big data analytics. Your system has to work, or you lose customer attention very quickly. Data quality is an issue. Scale is an issue. Progress is being made. The big shift is troubleshooting moving from reactive to proactive. It is all based on timely, high quality data.
“Predictive intelligence should be thought of as a guiding hand that helps businesses see and measure performance across all networks that impact the user experience, forecasts issues based on historical data and influences decision-making.”
BONUS: Neudata published Foot traffic data: The authoritative guide for data buyers and sellers. February 2023. Followed by Clickstream data: The authoritative guide for data buyers and sellers. March 2023.
BONUS 2: Seek.ai’s Sarah Nagy spoke at Data Driven NYC. March 2023.
BONUS 3: Michael Recce is now writing on SubStack. Sign up!
#1 – Invest Like The Best podcast with Patrick O’Shaughnessy interviews Auren Hoffman - A Deep Dive on Data. March 2023. Alternative link here.
My Take: Auren is a must follow if you are in the data business. He has built successful companies in the past (LiveRamp), and is now in the process of building a data business in SafeGraph. He is accessible and publicly shares much of what he already knows & what he is in the process of learning.
This interview helps define the different types of data businesses and is a helpful framework to use when discussing the broader business of data.
At the end of the day, all data is rows & columns … and you want to cover as many as possible. The more frequent the data changes, the more valuable it is.
Auren started SafeGraph as an answer to the trend of massive amount of data being siloed in a few very large organizations. Broadening access to important data types is the opportunity.
Types of data (all data is focused on one of these):
People
Places (SafeGraph)
Organization
Products
Phases of data:
Ingest (more sources = better)
Reducing to just the good stuff
Delivering to customers (documentation, etc)
Highlights (69-minute run time):
Minute 03:00 – interview starts; discussion of 2x2 matrix (see below)
Minute 05:00 – “religion data” companies (ie FICO)
Minute 08:30 – great businesses = “could fire the top 100 people in the company and the company would still be great.”
Minute 09:00 – great businesses = “decreasing CAC over time.”
Minute 11:00 – “religion application” company (ie Verisk) & data co-ops
Minute 13:30 – “truth application” company (ie Bloomberg)
Minute 15:00 – “truth data” company (ie SafeGraph)
Minute 17:00 – 4-types of data; selling data is hard; ingredients analogy
Minute 19:00 – how to think about the end market
Minute 21:30 – features of a good data set (must be true, transparent, easy for customer)
Minute 25:30 – feature of a data biz: slow growth initially, but long-term are very attractive
Minute 27:00 – SAAS biz vs Data Biz
Minute 30:00 – struggles of the early data business (long sale cycle, blind of use cases)
Minute 32:30 – foundation of SafeGraph and finding big problems to solve
Minute 35:00 – privacy technology improvements
Minute 38:30 – SafeGraph data
Minute 40:00 – primary use case and core customers
Minute 43:00 – phases and data ingest and building a data product
Minute 47:00 – data business founders (weird archivists, comfortable being on the side, humble)
Minute 50:00 – study Bloomberg as this is a really good company
Minute 51:45 – importance of being transparent, especially in the world of data
Minute 53:00 – most common way to fail … to try to be in more than one quadrant at once
Minute 55:30 – importance of join keys (hard to join products)
Minute 57:30 – conversation moves from data towards other interesting topics: Auren’s dinner parties (see Final Thoughts below)
Source: Truth Vs. Religion: What Kind Of Data Company Are You?
Source: Dan Entrup’s Its Pronounced Data. Dan updated his alternative data market map.
This one is from me. An attempt to visualize the temporal value of data (as discussed on the podcast summarized above).
Kelvin Mu’s market map of generative AI companies, sorted by segment, &total fundraised amount.
Source: DWYSYWD
Do What You Said You Would Do
In theory, this is easy…but it’s not.
In theory, doing this wouldn’t make you a stand out…but it does.
When you tell someone you will call them, call them.
When you tell someone you will deliver something, deliver.
When you tell someone you will make your bed, make your bed.
“Show up. Be consistent. Do what you say. Give more than you take.” – Unknown
There are two common excuses:
1- You forgot. This is not a good excuse. Figure out how to not forget. Use the tools that are available to you. Your phone is in your hand all the time, use it to help you not forget.
2- Things don't go as planned and you can’t deliver what was expected, when it was expected.
Perhaps you said you would do something on a Monday and circumstances outside your control cause a delay. As early as possible, be sure to tell whomever is relying on you that there is a delay.
Perhaps you said something in passing and weren’t even sure the other person is expecting anything from you. Deliver anyway.
What if there deliverable is unclear and open to interpretation. Do your best, and deliver something. At a minimum, that will demonstrate you care.
What if you hoped to deliver good news and it is now bad news. Perhaps no one would likely ever call you on it. Don’t ignore it. This is dangerous. Delivering bad news in direct fashion will build trust.
Follow through. Be reliable. Communicate delays & bad news quickly.
BONUS (this was tail end of Auren’s Podcast summarized above):
Great dinner parties
Appeal to both introverts & extroverts
How to appeal to introverts, where to sit, structure, one conversation, let them know who else is coming, next day follow up and share contact information the next day, everything is off-the-record
Have one conversation, not multiple (good acoustics)
Need a moderator, 8-14 people is ideal
Food is least important
Let people know what the conversation will be about
End early, leave them wanting more