Getting your data ready as the AI race heats up

On Mar 20, 2025

Just months into 2025, and the global race to innovate around AI has already delivered a slate of major developments. Most recently, the launch of (DeepSeek in China) set off discussions about the downward trend on the costs of AI for businesses, and that was just weeks after the UK government laid out its plans to make the country an AI superpower.

We’re seeing a collision of AI policy and business response, and the UK government’s promised infrastructure improvements will create a newfound urgency to create or implement solutions that can take advantage.

But are businesses ready? While organizations may already be using AI in different capacities to boost productivity and make better decisions, for those gains to improve in line with the technology’s rapid development they need to consider one key factor: the caliber of their data.

Eilon Reshef

Cofounder and CPO of Revenue AI platform Gong.

The three elements of ‘good’ data

There are three pillars that underpin strong data foundations – quantity, quality and context – which leaders across business functions will need to understand to reap the benefits AI promises for their teams. As the pace of innovation quickens, the sooner these principles can be applied, the earlier companies can start extracting value from their AI adoption journeys.

1. Quantity

Many data stores rely heavily on manually entered data, which opens the door to human error and inconsistencies. These existing data gaps cannot be ignored any longer or organizations risk letting them continue to widen, putting them further and further behind the pack as others adopt AI.

For example, a tech firm that solely uses CRM data to drive sales for a new product might overlook signs of buyer hesitation that were never entered into its platform. Scale this across processes in a world where AI is integrated throughout organizations, and soon you’re left with a lot of misguided decisions. However, an AI solution that automatically captures data from multiple sources like calls and emails – wherever prospects engage with them – could get deeper insights into how they can better communicate the product’s benefits.

Businesses need to be bridging the data gap today, adopting tools and processes that enable automated data capture across various touchpoints. Even within functions that haven’t yet integrated AI, effectively collecting data now is the first step on the roadmap to eventually doing so.

2. Quality

Having lots of data means little if it isn’t objective and trustworthy, and this is where human error or unintentional bias can prove a pitfall. If the dataset powering a model is incomplete, biased or outdated, even the most powerful AI’s outputs will be flawed and impact any decisions made that are based on them.

Organizations will only be able to take advantage of the rapid pace of AI innovation if they adopt automated data capture to minimize manual inputs. Otherwise, they’ll continue running into the same issues over and over: outputs that reflect the assumptions and biases of whoever first entered the data, rather than reality.

3. Context

AI only becomes truly powerful when it can merge lots of high-quality data with the specific context it is being used for. A work landscape where processes are AI-enhanced across functions, ultimately the vision many have for the technology, cannot be achieved if organizations don’t map their discrete data to the relevant business context.

Say a customer wants to churn because they are unhappy with the service they’re getting. An AI model that doesn’t understand the business or customer won’t generate a meaningful response to their concerns. Arming it with the right contextualized data, however, would allow the AI to account for past interactions, the length of the relationship and available products or services in order to generate a tailored response that will actually help the customer.

Why data matters – now and in the future

No amount of government support will change the fundamental principle that in order to deliver accurate and actionable outputs, enterprise-grade AI solutions need to have the right data feeding into them. Without a strategic approach to gathering data and applying the right context to it, even the most technically advanced AI platform will come up short, even with the infrastructure gains countries are racing to put in place.

Limited or inaccurate data will lead an AI application to produce unreliable or superficial results, like advice for engaging with prospects based on incomplete CRM data that, as a result, could be biased. On the other hand, an AI strategy built on strong data foundations lets organizations tap into deeper, more relevant insights with unmatched speed in the short-term, while setting the stage to take advantage of new advancements as they emerge from different parts of the world.

There’s no shortcut to success with AI. Anything built on shaky foundations is at risk of underdelivering from the start, and that’s where we are with AI. There is still so much potential to unlock but many organizations still aren’t putting themselves in a position to fully realize it.

Organizations that get it right will be able to apply AI to existing operations more effectively and give themselves a competitive edge in capitalizing on the global race to build out the necessary infrastructure.

We’ve featured the best AI website builder.

This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

Source
Las Vegas News Magazine

Ai alibaba Amazon android apple Cyber Security data breach DDoS