Bite-sized ideas for little moments, so we both can stay up to date without having to wade through the sea of noise.

June 10, 2025

I once heard someone say, "There is no 'data strategy,' just a business strategy that happens to use data."

I held onto it because it succinctly communicates the fastest way to earn executive buy-in for the company's data program. By keeping the business strategy at or near the center of the data initiatives*, data teams ensure that their work is deeply tied to the chief concerns: making money or reducing costs (or both!).

Starting to develop a data strategy can be pretty simple: take an inventory of the 'forever-problems' within your business — the ones that come up again and again.

Common ones include:

• Forecasting revenue, costs, or another key financial metric [analysis-based]

• Defining you ideal client with numbers: Who are your buyers? What is their lifetime value? How do you effectively market to them? [analysis-based]

• What are value-add product or service recommendations you could make? [analysis-based]

• Profitability: What are the margins on each product or service you offer? [analysis-based]

• How do you reduce manual workflows (data entry)? [operations-based]

• How can we move data between systems and prevent data duplication? [operations-based]

• Where are opportunities to document, standardize, automate, and streamline internal workflows? [operations-based, now easier with LLMs]

• … and many more…

The technical details (APIs, data warehouses, analytics platforms, coding..) can all come later. Consolidating your thoughts around high-leverage problems, then developing a mechanism for systematic data capture around those problems, is a great place to start.

You don’t need to dive into technical details (APIs, warehouses, coding, etc.) right away. Start by identifying high-impact problems, then create a system for capturing the data you need to solve them.

**Note: R&D has a time and a place. We'll cover that later.

Use data today,

Paige

June 9, 2025

Last week, Mary Meeker and BOND's report on trends in artificial intelligence took the internet by storm. For those of us suffering from the plight of the 21st-century-short-attention span, you'd think the 340-page slide deck would be a nonstarter… if it weren't for nearly every chart being up-and-to-the-right (a notable example being the estimated 800 million weekly active ChatGPT users as of April 2025).

For this bite, let's set aside 320 of the slides, and just focus on the "so what?" of the mind-boggling capex numbers:

From 2023-2024 alone, the "Big 6" (Apple, Nvidia, Microsoft, Alphabet/Google, Amazon (AWS only), and Meta), increased yoy capex spend by 63%. For AWS, it's estimated that 2024 capex spend equated to almost 50% of the year's annual revenues.

My not-so-insightful commentary is that the big tech companies are in the thick of the AI arms race.

My [hopefully] helpful commentary addresses what that arms race means for the 99.999% of us who aren't the Big 6. State-of-the-art (SOTA) AI models are hitting the market everyday - and unlike so many advanced technologies, many models are open-source, meaning they can be wholly adapted, integrated, and leveraged by companies and individuals around the world (for free).

Building on the arms race analogy, if business is akin to a "war," the consequence of big tech's capex is that we all have access to the same tools (weapons). Where, then, are leaders to get a competitive advantage? Much like military intelligence, leaders will increasingly depend on leveraging information asymmetries to gain a competitive edge.

If we all have AI, competitive advantage will side with those able to leverage proprietary data to maximize the technology's capabilities.

When we think about it this way - we ought be less stressed over which chat to buy the enterprise license for, and more preoccupied with getting our internal data organized and optimized so that we may use any of the chat services in a way that far outpaces the competition.

Select supporting slides: