History Repeats: What Spreadsheets, Accountants, and AI Tell Us About the Future

Reader's note: Our last ChangeWave newsletter focused on "From Numbers to Knowledge: Crafting an Analytics Strategy That Works." In the age of AI, an analytics strategy is a must. Today, we dive deeper into our future and what we should focus on.

Why Clean Data Matters More Than Ever in the Age of AI

Remember when everyone said spreadsheets would replace accountants? Spoiler alert: They didn't. Instead, accountants who embraced spreadsheets became more valuable than ever. Today, we hear similar predictions about Large Language Models (LLMs) replacing knowledge workers. But here's the truth: just like spreadsheets, these AI tools will enhance rather than replace – and they'll only be as good as the data we feed them.

What Are Large Language Models, Anyway?

Think of an LLM as a super-smart student who has read millions of books. This student (let's call them "AI Allie") can write essays, answer questions, and even crack jokes. But here's the catch – "AI Allie" can only work with what it has learned. If those millions of books contained mistakes or outdated information, guess what? Those errors also show up in AI Allie's work.

Here's a real-world example: Imagine using an LLM to analyze customer feedback. If your customer data is messy – with duplicate entries, misspelled names, or incorrect categories – the LLM might tell you that "John Smith" and "J. Smith" are different customers with different preferences. Not very helpful, right?

The Data Quality Imperative

Remember the old computer programming saying, "garbage in, garbage out"? With LLMs, it's more like "garbage in, garbage out – but now at the speed of light!" Clean, accurate data isn't just lovely; it's essential. Think of it like cooking: even the best chef can't make a delicious meal with spoiled ingredients.

Three Quick Ways to Improve Your Data Quality

1. Start a Data Clean-Up Routine. Schedule weekly data "health checks" like you'd schedule regular car maintenance. Create standard naming conventions (is it "Customer ID" or "Client Number"?)Run basic quality checks (Are there phone numbers with letters? Addresses without zip codes?)
2. Make Data Quality Everyone's Job. Yes, there has to be executive buy-in. You also need to train your team on proper data entry. Web interfaces developed by your organization must enforce data quality. Create clear guidelines for data handling. Celebrate good data practices (Maybe "Data Champion of the Month"?)
3. Set Up Quality Gates. Install basic validation rules (No more birthdays in the future!)Create a simple review process for any new data. Schedule regular data cleaning sessions (Think of it as spring cleaning but for your database)

Learning from History

Remember Y2K? Or when people said email would create a paperless office? Technology often doesn't replace things entirely—it transforms them. Spreadsheets didn't eliminate accountants; they made them more efficient and freed them to focus on higher-value work like analysis and strategy.

The same pattern emerges with every major technological advance. ATMs didn't replace bank tellers, and online shopping didn't kill retail. Instead, these technologies changed how we work and created new opportunities.

Where Do We Go From Here?

LLMs represent an incredible technological leap forward, but they're not magic. They're tools – powerful tools, but tools nonetheless. Just as a Ferrari won't run well on low-quality fuel, LLMs won't perform well with poor-quality data.

Remember the dot-com bubble? The social media gold rush? The cryptocurrency craze? Each wave brought both opportunities and overblown predictions. While some companies capitalized on these waves to achieve remarkable success, others faced devastating downfalls. The difference often came down to fundamentals – including data quality.

The Future Belongs to the Prepared

As we enter this new era of AI and LLMs, the winners won't be those who unquestioningly adopt new technology. The winners will be those who build a strong foundation of clean, reliable data. They'll be the ones who understand that AI isn't about replacing humans – it's about augmenting human capabilities with better tools and insights.

Want to get started? Begin with small steps: Clean up one dataset, standardize one process, and train one team. Remember, even the longest journey begins with a single step—preferably one based on accurate data.

While Large Language Models (LLMs) represent a revolutionary advancement in technology, their effectiveness depends entirely on the data quality they process, much like how spreadsheets enhanced rather than replaced accountants. Organizations must prioritize data quality through systematic cleaning, standardization, and validation processes to harness the true potential of LLMs. The historical pattern of technological advancement shows that tools augment rather than replace human capabilities, making it crucial for businesses to build robust data foundations. Success in the AI era will belong to organizations that maintain high-quality data practices while thoughtfully integrating new technologies.

Now, here's a quick word from our sponsor!

Versalytix is your ideal partner in this data quality journey because we combine deep analytics expertise with practical, real-world experience in data governance and AI implementation. Our team has successfully guided numerous organizations through data transformation initiatives, offering battle-tested methodologies that deliver measurable results. Drawing from our extensive experience working with premier organizations and our research partnerships with institutions like Stanford University, we understand the technical and human aspects of maintaining data quality in the age of AI. Our proven track record in analytics consulting, combined with our understanding of emerging technologies like LLMs, positions us uniquely to help your organization build the robust data foundation necessary for AI success. Learn more at www.versalytix.com.

Inspirational people for this newsletter (could be a conversation, presentation, blog, interaction, book, video that I watched of yours, podcast, or anecdote that was shared and I remembered while writing this post):

Barry Chaiken Bruno Aziza Nancy Morgan Cindi Howson Amit Prakash Papa Diaw Phil Walton Asha Saxena Heath Glass Mia Umanos Ania Rodriguez Phil Santoni Israel Ayoola Andy Dé Cat Brunson Sarah Pallett

Midjourney Prompt: A split scene showing an accountant from the 1980s with a paper ledger transforming into a modern data analyst working with holographic AI displays, professional office setting, photorealistic, cinematic lighting, depth of field, high detail. Mood: Forward-thinking, optimistic but grounded