Author

Harry Morphakis

Senior Director at Artefact UK

Remember Business Objects? If you do, you might see where I'm going with this. We're heading into a period where Large Language Models (LLMs) capable of building business-level software will become increasingly commoditized. It's an impressive technological feat, sure, but it's also a potential race to the bottom.

Why? Because while everyone’s busy showing off their latest AI parlour tricks, they’re missing the real goldmine: the data itself.

Let’s break this down:

  • The Commoditization of LLMs: We’re witnessing a rapid evolution in the field of LLMs. What was cutting-edge yesterday is becoming standard today. Soon, there will be little to differentiate one LLM from another in terms of basic capabilities. The ability to generate human-like text, answer questions, or even write complex code will no longer be a distinguishing factor. This commoditization means that simply havinng access to an LLM won’t provide a significant competitive advantage.

  • The Rise of AI-Generated Software: LLMs and other generaive AI tools are becoming increasingly capable of creating a software on demand. Need a custom app for inventory management? An AI might soon be able to whip that up for you in minutes. This is revolutionary, but it’s also becoming universally accessible.

  • The Staying Power of SaaS: However, don’t expect the likes of Salesforce or SAP to disappear overnight. Much like how Red Hat has thrived in the open-source era by providing enterprise-grade support and services, large SaaS providers will likely adapt to this new landscape.They may integrate AI-generated solutions into their offerings or focus on providing the robust, scalable infrastructure needed to run these AI-created applications.

  • The Real Competitive Advantage: Here’s the kicker – the true source of competitive advantage in this new landscape will be your company’s proprietary data. Why? Because while LLMs can generate generic solutions, only your unique data can power AI models that understand the nuances of your business, your customers, and your market.

  • The Data Readiness Gap: Here’s where many companies stumble. You can have access to the most advanced AI tools on the planet, but if your data isn’t properly prepared, you won’t be able to leverage these technologies effectively. It’s like trying to fuel a Ferrari with crude oil – you shouldn’t.

Consider this scenario in the CPG world: Company A and Company B both have access to the same advanced LLM for demand forecasting. Company A has years of clean, harmonised sales data, complete with contextual information about promotions, weather conditions, and social media sentiment. Company B has scattered, inconsistent data stored across various systems. Which company do you think will be able to make more accurate predictions?

The truth is, without data that is correctly cleaned, harmonised, and modelled, companies won’t be able to take full advantage of the AI revolution. You might have a powerful AI at your fingertips, but if you’re feeding it poor quality data, you’ll only get poor quality insights out.

This is why the real race isn’t about who can implement AI the fastest – it’s about who can get their data house in order. It’s about creating a robust data infrastructure that can fuel these powerful new AI tools. Because in the end, the AI is just the engine – your data is the fuel that makes it run.

In the CPG sector, this could mean the difference between:

  • Accurately predicting and meeting seasonal demand spikes

  • Optimising your supply chain to reduce waste and improve efficiency

  • Understanding and responding to changing consumer preferences in real-time

  • Personalising marketing efforts to dramatically improve conversion rates

So while your competitors are chasing after the latest chatbot or AI writing tool, remember: the real value lies in your data. The question is, are you ready to unlock it?

The Journey to Data: The New Frontier

We’ve all heard about the “journey to the cloud.” It was the buzzword du jour for the better part of a decade. But now? We’re embarking on an even more critical expedition: the journey to data.

This isn’t just another tech trend. It’s a fundamental shift in how enterprises operate, innovate, and compete. And if you’re not on board, you might find yourself left in the dust.

What does this journey look like in practice? It’s about moving from disparate, siloed data sources to a unified, accessible data ecosystem. It’s about transforming raw data into actionable insights. It’s about creating a single source of truth that can power everything from operational decisions to advanced AI applications.

For a CPG company, this might mean integrating point-of-sale data with social media sentiment analysis, supply chain information, and customer feedback. It’s about creating a holistic view of your business that allows you to spot trends, predict demand, and innovate products faster than ever before.

The Real Value Proposition: It’s the Data, Stupid

Here’s the thing: AI is only as good as the data it’s trained on. You can have the most sophisticated LLM on the planet, but if it’s not trained on your specific business data, its value to your enterprise will be limited at best.

The real competitive advantage lies in unlocking the value of your proprietary data. It’s about creating a robust data platform that can:

  • Aggregate data from disparate sources across your organisation

  • Clean and standardise that data for use in analytics and AI applications

  • Make that data accessible to the right people at the right time

  • Ensure data quality, security, and compliance

This is where the rubber meets the road. And let me tell you, it’s not easy.

Let’s break this down:

Aggregation: In a typical CPG company, data might be spread across ERP systems, CRM platforms, supply chain management tools, and countless Excel spreadsheets. Bringing all this together is a herculean task, but it’s essential for getting a complete picture of your business.

Cleaning and Standardization: Raw data is messy. Different systems might use different formats or naming conventions. Cleaning this data and standardising it is crucial for any meaningful analysis or AI application.

Accessibility: Data is only valuable if it’s in the hands of people who can use it. This means creating interfaces and tools that allow business users to access and analyse data without needing a PhD in computer science.

Quality, Security, and Compliance: With great data comes great responsibility. Ensuring the accuracy of your data, protecting it from breaches, and complying with regulations like GDPR is non-negotiable.

The Challenge: More Than Just Tech

Here’s where many enterprises stumble. They think this is purely a technological challenge. But it’s not. It’s an organisational one.

Enter the concept of data mesh. As Zhamak Dehghani, the originator of the data mesh concept, puts it: “Data mesh is a decentralised sociotechnical approach to remove the dichotomy of analytical data and business operation.”[^1]

In other words, it’s about breaking down silos, fostering a data-driven culture, and empowering domain experts to become data producers and consumers.

This requires buy-in across the enterprise. It necessitates upskilling your workforce. And yes, it demands significant technological investment.

But what does this look like in practice? Imagine a CPG company where:

  • The marketing team can access real-time sales data to adjust campaigns on the fly

  • The product development team can instantly analyse customer feedback to inform new product features

  • The supply chain team can predict and prevent disruptions before they happen

  • The C-suite can get a real-time, holistic view of the entire business at the click of a button

This is the promise of a well-implemented data mesh. But getting there requires more than just new tech. It requires a fundamental shift in how we think about and organise around data.

The Path Forward: Platform Engineering and Model Building

So, where do we go from here? For the next few years, the focus needs to be squarely on data platform engineering and model building.

This means:

  • Investing in robust data infrastructure

  • Developing data governance frameworks

  • Building and training models on your proprietary data

  • Fostering a culture of data literacy across your organisation

Let’s explore each of these in more detail:

Investing in data infrastructure: This goes beyond just buying the latest tech. It’s about creating a scalable, flexible architecture that can grow with your business. For a CPG company, this might mean implementing IoT sensors in your supply chain, setting up real-time data pipelines from retailers, or creating a central data lake for all your historical data.

Developing data governance: With great data comes great responsibility. You need clear policies on data usage, quality standards, and security protocols. This is especially crucial in the CPG sector, where you’re dealing with sensitive customer data and potentially valuable trade secrets.

Building and training models: This is where the magic happens. By training AI models on your unique business data, you can create predictive tools that give you a real edge. Imagine an AI that can predict product trends before they happen, or optimise your pricing strategy in real-time based on a multitude of factors.

Fostering data literacy: This is perhaps the most challenging but also the most rewarding aspect. It’s about creating a culture where everyone, from the CEO to the frontline staff, understands the value of data and knows how to use it in their daily work.

It’s a tall order, but the alternative is being left behind in a world where data is the new oil, and AI is the engine that runs on it.

The Bottom Line

The AI revolution isn’t coming – it’s here. The winners won’t be determined by who has the flashiest AI demos. The real victors will be those who can effectively harness their data to drive real business value.

In the CPG world, this could mean the difference between a product launch that fizzles and one that takes the market by storm. It could be the key to optimising your supply chain to weather the next global disruption. It could be what allows you to personalise your customer experiences in ways your competitors can only dream of.

So, ask yourself: Are you ready for the journey to data? Because ready or not, it’s already begun. The platforms are being built, the models are being trained, and the race is on. Will you be leading the pack, or playing catch-up?