Thanks for being here!
Alt Data Weekly is Powered by Vertical Knowledge
Announcement(s):
Thanks for the feedback on last week’s 3-year anniversary … the three winners picked the following books:
The Datapreneurs, Bob Muglia (with Steve Hamm)
Football Analytics with Python & R: Learning Data Science Through the Lens of Sports, Eric Eager & Richard Erickson
The Ascent of Information: Books, Bits, Genes, Machines, and Life's Unending Algorithm, Caleb Scharf
I’ll be speaking as part of a panel at the upcoming Beryl Elites Conference in NY November 6-7. Panel Topic = Alternative Data: At The Forefront of Innovations in Investment Management. Please join me!
My VK colleague Rayne Gaisford, Chief Product Officer & Head of Data Strategy, is speaking at BattleFin’s November 2 Corporate Discovery Day.
Theme that emerged in this week’s email is … product–market fit is elusive.
QUOTES
“We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” - Roy Charles Amara (as found in Andre Retterath’s article)
News Articles
Podcasts
Cool Charts
Final Thoughts (The Conference Circuit)
#1 – Neudata’s Data Monetisation - The Authoritative Guide. September 2023.
My Take: Data is becoming more valuable. Those with good data need to at a minimum evaluate the opportunity. Data will eventually be a balance sheet asset. Data monetization is not easy (whether UK or US spelling), but a worthy pursuit.
#2 – Andrea Squatrito from Data Boutique published The Hard Economics of Selling Web Data. October 2023.
My Take: This is another in a line of great pieces found on Data Boutique’s blog. Thank you for publishing.
“Data is sold not bought” (quote attributed to Dan Entrup). Anyone thinking of web-scraping in-house will quickly find it is a fool’s errand. Easy to do, but very difficult to do well … & doing it poorly adds no value, and may likely detract value. Add in the most basic compliance concerns of any mature business & collecting web data internally gets costly, quickly.
Selling is hard as the CAC (Customer Acquisition Cost) is high. But given the value our data collections are adding… we can charge enough to make direct selling a good business. Engaging the client, educating the client, understanding their needs, delivering the products consistently. It is a long, consultative sales cycle with lots of human intervention along the way.
Marketplaces for data products make sense as data products become more out-of-the box ready, one-size-fits all offerings. As the use of data & data product adoption broadens, this type of buying will happen, but in my experience most customers have unique requests and out-of-the-box isn’t always the right fit.
See some good charts from this article below in “Charts” section.
#3 – Neudata’s Web-scraping and copyright law. October 2023.
My Take: Compliance is a huge issue with data monetization. Web-scraping poses unique challenges. My experience is that this is a well-understood practice with clear guidelines needed for compliance.
The next battlefield will be AI and the training data that flows into models.
More broadly, I continue to believe popular concerns about “privacy” are changing as younger generations don’t seem to be as concerned about the “creep factor” where very targeted advertisers finds you based on some of your past online behavior.
BONUS: Rod James of the WSJ published Founders of New Private Equity Put Big Data Front and Center. October 2023. “We can ingest a lot of data points, traditional and nontraditional, and use them as a lever to create information”
What else I am reading:
Rebellion Research’s Top 10 Quants On Wall Street 2023. October 2023.
Sven Balnojan of Three Data Point Thursday published Alternative Data Inspiration List. (note: call me about web scraping data)
Sven Balnojan of Three Data Point Thursday published The Practical Alt Data Brief.
Bill Gurley’s 2012 article 10 Factors To Consider When Evaluating Digital Marketplaces.
David Tuppen’s Simplifying the ‘data product’. October 2023.
Vin Vashishta’s The Bigger Picture For Enterprise Data Infrastructure. October 2023.
Source: The Data Cloud Podcast interviews Alex Izydorczyk of CyberSyn. October 2023.
My Take: You have to keep an eye on CyberSyn if you are in the data space. CyberSyn buys & collects data. They then monetize that data via the SnowFlake marketplace.
Alternative data more accurate & less bias. Organizing the massive amount of data that are now accessible will allow decision makers a “Consumer “360 view of their domain.
It would be hard to imagine a baseball GM making decisions without access to basic data…or even without today’s advanced analytics. This just 20 years after “Moneyball” was a unique enough connect that it was made into a movie.
Outside of data, I liked Alex’s thoughts on talent stacking & how that helped him get to where he is today.
LLMS three ways:
Useful tool to clean, join, structure data
LLMS have fueled demand for properly licensed data to use as inputs for models
LLMS will democratize access to data
Non-consensus thought – the role of structured data in LLMs … structured data is not dead! LLMs allow us to take structured data and join it to things that would have been difficult to do in the past. Fuzzy joining example. Joins that are “semantically aware”
Highlights (46-minute run time):
Minute 01:00 – Interview starts; CyberSyn background
Minute 03:15 – Alex’s background & how he eventually got to CyberSyn (talent stacking)
Minute 10:00 – Alex’s thoughts on “Alternative Data”
Minute 13:00 – Data as a service (DAAS) and CyberSyn capitalizing on Snowflake’s technology
Minute 16:00 – Moneyball-ization of business
Minute 23:00 – How did Alex get introduced to Snowflake
Minute 27:30 – How does CyberSyn use the Snowflake data marketplace
Minute 33:00 – What’s new on SnowFlake’s marketplace? (you can now use SnowFlake credits to buy data)
Minute 35:00 – Alex’s thoughts on LLM’s
Minute 40:00 – Looking out 5-years or more
Minute 43:15 – Alex’s interest in video games & impact on his path
Source: Charts from the above highlights Data Boutique article: The Hard Economics of Selling Web Data.
BONUS: Subscribe to Three Data Point Thursday ... good memes and pictures (among other good info)
Source: In-person profession relationship development is still a real (and valuable) thing.
You may have seen the quote above “data is sold not bought”. Despite ongoing efforts, we are not yet to a point where data is bought at scale without a human-in-the-loop. That human is usually a salesperson.
While your personal network is a good place to start, the data conference circuit is the next best thing.
List of upcoming US data conferences (VK will have presence at all of them):
November 2 BattleFin Corporate Event
November 6-7 Beryl Elites 5th Annual Thought Leadership Conference
December 7 Neudata Winter Data Summit
January 18, 2024 Eagle Alpha Next Level Data Conference
January 24-26,2024 BattleFin Miami
When will we get to more of a self-service model? As tools get better and more data is readily available in an organized fashion, we’ll see this be more common (see what CyberSyn is doing with the SnowFlake marketplace).
But for now … expect to hear from your salesperson!
Thanks for the mention, John!