How to sell: Data readiness for AI

Why does the AI story start with data?

Some customers will be keen to embrace AI, grabbing what they can and deploying it without doing the planning around data and security. That comes with risks – risks that can be largely avoided by taking a measured approach.

“Before they can integrate AI effectively, organisations must address how they collect, store and manage their unstructured data, particularly at the edge. It is vast, dispersed across multiple silos, and difficult to access and manage efficiently,” says Nick Burling, chief product officer at Nasuni.

“Unstructured data – the 80-90% of enterprise data consisting of documents and emails, to images, videos and design files – is emerging as the backbone of AI innovation, redefining how enterprises harness intelligence across their organisations. This vast, often underutilised data holds the key to deeper insights, smarter automation, and more contextually aware AI systems. The potential value from unlocking it has never been greater,” adds Burling.

Nicole Reineke, AI strategist at N-Able, agrees that the foundations have to be solid if the results are to be meaningful.

“Simply put, AI is math. So, while outputs can be exactly what we we’re looking for, they are still just the result of models trained on data. Which brings us to one of the oldest, truest sayings in computing: garbage in, garbage out,” she says.

“If the training data is biased, incomplete, outdated, or just wrong, like inconsistent ticket notes, vague status fields and knowledge base articles written in 2014, then the AI’s output will reflect that. The bad information is handed back to us, just as a more polished version. If the data you input is wrong, the AI story will be too – the data shapes the story AI tells,” adds Reineke.

What needs to be done to get data ready for an AI deployment?

One of the initial challenges is deciphering the customer claims from the reality of their position.

“Our research shows that  enthusiasm outpaces readiness when it comes to deploying AI. Some 94% of companies are increasing spending on products and services to support data readiness for AI, but only 21% have fully embedded AI into their operations,” says David Zember, senior vice-president of worldwide channels and alliances at Qlik.

Partners might well have to push the data preparation message to underline the importance of starting off in the right way.

“Data readiness is one of the most overlooked stages of any AI deployment, despite it being a crucial deciding factor in whether the deployment will succeed. To truly become ‘AI ready’ organisations must focus on the visibility, governance and mobility of their data,” says Sonya Matthieu, senior director partner for the UK and Ireland at NetApp.

Data readiness is one of the most overlooked stages of any AI deployment, despite it being a crucial deciding factor in whether the deployment will succeed

“In practice, this requires consolidating any silos and ensuring data is stored in the right place for the right workload. This means whether that’s on-prem, in the public cloud, or even at the edge,” she adds.

Hubert Składanowski, machine learning manager at Datactics, has some advice for those looking at the steps needed to get customers prepared.

“Typical checklists include data being complete so it’s free of gaps, being correct with consistent data types, high volume with equally split classes to allow good representation but avoid bias and, crucially, it should represent the use case and the domain,” he says.

“If the production data is often unclear, your training data should not be cleansed by default, unless the same cleansing is applied before the input hits the AI in production,” adds Składanowski.

How can you support it and build on that process with further services?

The ideal situation is that once a partner has worked with a customer to prepare the ground for AI, they will then be the one to reap the rewards of providing additional services. There are some suggestions of what those opportunities might look like.

“Once the groundwork is in place, partners can layer additional value by introducing data management and AI-ready services. Observability, cost optimisation and sustainable storage that reduces the environmental footprint of growing data estates are just a few examples of what this could look like,” says NetApp’s Matthieu.

“In fact, this is also where partners can establish long-term relationships. Data environments are not a once-and-done exercise, but something that is continuously evolving. As a result, customers will continually need support monitoring, securing and optimising their AI workloads,” she adds.

There is also going to be a need for partners to continue getting more of a customer’s infrastructure connected if insights from across the business are to be unlocked.

“Many legacy systems were never designed to detect biased sources, outdated inputs or unclear data origins. This is where partners can add value, by offering services that track these AI-specific risks, helping customers continuously validate their data, so they know it’s still fit for use,” says Zember.

Partners can add value by offering services that track AI-specific risks, helping customers continuously validate their data, so they know it’s still fit for use

“From there, partners can build additional services around automated data quality checks, strong governance frameworks, model monitoring and bias detection. These layers help maintain reliability as AI scales. By combining strong data foundations with ongoing checks and safeguards, partners can help customers keep risks under control and ensure AI outputs remain accurate, trustworthy and actionable over time,” he adds.

Amy Illingworth, head of channel marketing at Exclaimer, also has some ideas about where the channel can go after the initial data readiness process.

“Once a reseller has helped a customer get data ready, there are a few routes they can take to upsell ongoing support. The natural next step is supporting with ongoing data maintenance – customers rarely have the time or resources to keep data clean, so there is clear recurring revenue potential in monitoring and remediation,” she says.

“Advisory around governance is another growth area, especially as AI regulation evolves. Channel partners are in a strong position to guide small and medium-sized businesses through the complexities if they have limited expertise or understanding. Then there are integration services, such as connecting cleaned, governed data into AI platforms, customer relationship management [CRM], collaboration tools, analytics and line-of-business apps,” adds Illingworth.

“The most advanced partners will optimise efforts for their customers’ verticals to ensure meaningful support, tuning data pipelines and models for particular industry use cases that create real impact. In essence, strong data readiness creates an upsell path into automation, insight services and ongoing AI tuning and optimisation,” she adds.

What are the prospects for those that develop data readying skills for 2026 and beyond?

The realisation by more customers around the importance of getting their data into a good position should drive the market forward.

“Data preparation isn’t glamorous, and it may not be a headline – but it is the foundation of every successful AI initiative. Treat your data with the same care you treat your strategy. If the foundation is weak, it’s a failed mission from the get-go. Once you have developed the data readying skills, you will be in a better position to approach AI innovation and apply AI tools, techniques and frameworks to explore solutions,” says N-Able’s Reineke.

Data preparation isn’t glamorous, and it may not be a headline – but it is the foundation of every successful AI initiative

Exclaimer’s Illingworth stresses that this is not a one-time sale, and relationships between partners and customers should develop over the next year and beyond.

“As the technology moves from the shiny new tool to business as usual, the channel opportunity will follow suit by moving from selling tools to managing the ecosystem, including data quality, ongoing model performance, integration into workflows and compliance. These are all sustainable recurring revenue streams and can foster stickier customer relationships,” she says.

“By developing these skills now, MSPs [managed service providers] and resellers can position themselves as strategic partners rather than commodity suppliers. This is not a one-off project category – it will grow and evolve as organisations scale their AI usage, making data readiness one of the most durable capabilities a partner can invest in,” adds Illingworth.

The last words go to Dan McAllister, senior vice-president of global alliances and channels at Boomi, who expects the opportunities for skilled partners to continue into 2026.

“As AI advances, partners that can help organisations to build strong data foundations will be crucial, and will determine which companies get real value from their deployments,” he says. “But more than this, the channel partners that build strong data management offerings now will be ready for whatever innovation falls after agentic AI. This readiness will put them in a prime position to lead the next wave of innovation.”