Dr. Gurpinder Dhillon explains why data quality in sales and marketing shapes AI trust, campaign ROI, customer confidence and business adoption.
Most companies do not have a data shortage.
They have a data trust problem.
Sales teams have customer records spread across different systems. Marketing teams run campaigns on lists they do not fully understand. AI teams build agents on top of fragmented information. Leaders ask for better decisions, but the foundation underneath those decisions is often incomplete, duplicated, or misunderstood.
During a recent conversation on The Executive Outlook, Dr. Gurpinder Dhillon explained why data quality in sales and marketing is no longer hidden inside databases. It is now a business performance issue, a customer trust issue and an AI readiness issue.
For Dr. Dhillon, the question has never been about data for its own sake. His journey began in computer science and software engineering, later moving into business, data science, and doctoral research in data quality. But one question followed him throughout his career.
Does this actually work, and how do we know it works?
Over time, that question became even more important. Why do capable organizations continue to make important business decisions using information they do not fully trust or understand?
That question sits at the center of this conversation.
For years, many organizations believed that collecting more data would lead to better decisions.
More customer data. More campaign data. More product data. More behavioral signals. More system exports. More event records.
But he believes this mindset created a new problem. Companies collected data faster than they built the discipline to clean it, connect it and understand it.
The result is visible across sales and marketing teams every day.
The same customer may exist in a CRM, an ERP system, a support tool and a marketing platform. Each system may hold a slightly different version of that customer. One record may have an old email address. Another may use a shortened company name. Another may treat the same person as a prospect, while another sees them as a loyal customer.
On the surface, this looks like an operational detail.
It can change the outcome of a campaign, damage customer relationships, and weaken trust in business decisions.
This is why data quality in sales and marketing matters so deeply. The issue is not whether a company has enough data. The issue is whether the data is accurate enough to support the decision being made.
One of the clearest examples he shares is campaign performance.
When a marketing campaign underperforms, the first instinct is often to question the creative, the messaging, the offer, or the channel. Leaders ask whether the campaign was strong enough. Teams adjust the design. Budgets move. New ideas are tested.
But sometimes, the campaign is not a real problem.
The real problem is the data underneath it.
A company may be paying to reach the same customer four or five times because the person exists as duplicate records across different systems. A loyal customer may receive a message asking them to come back, even though they never left. A prospect may receive the wrong offer because the organization does not clearly know who they are, what they need, or whether they are already part of the customer base.
That is not only inefficient.
It is damaging.
The company spends more money and creates a worse experience at the same time. The customer begins to wonder whether the brand really knows them at all.
He describes this as a data problem wearing a marketing costume. It looks like a campaign issue from the outside, but the root cause sits inside the quality and structure of customer information.
For executives, that distinction matters. If the data foundation is weak, better creative alone will not solve the problem. More automation will not solve it either. The organization must first understand whether it is speaking to the right person, with the right message, at the right time.
He does not believe every organization should begin by trying to fix everything at once.
His advice is practical. Start with the business problem.
If the issue is a specific campaign, begin by identifying the data needed for that campaign. Which systems matter? Which customer records are relevant? What information is needed to separate customers from prospects? How do we identify loyal customers, inactive customers, dissatisfied customers, or high potential buyers?
From there, the team can bring together the essential data sources and improve them for that use case.
This is important because many organizations try to solve everything at once. They begin with a massive data modernization effort before proving value in one focused area. The result is often slow progress, stakeholder fatigue and unclear business impact.
In sales and marketing, a more effective starting point may be a focused marketing data warehouse. This can connect the most important customer, prospect and campaign information into a more reliable view.
For a Chief Data Officer or data management leader, the long-term goal may be broader enterprise data quality and master data management. But for a marketing operations team, value can begin much smaller.
The question is not how much data we can fix.
The question is what decision we are trying to improve first.
Master data management often sounds expensive and complex.
Many leaders hear MDM and immediately think of large platforms, long timelines and major transformation budgets. Dr. Dhillon simplifies the idea.
Most companies have the same customer, product, supplier, or account scattered across multiple systems. Each system may hold a different version of the truth. Master data management brings those versions together so the organization can create one trusted view.
Some call this the golden record.
But he is clear that MDM does not always need to begin with a large platform. A smaller company may begin with an Excel file or a simple database. The first step is to recognize where the same customer appears in different ways and begin resolving those identities.
As the organization matures, the process can mature as well.
This framing is useful for leaders because it removes the fear that data quality work must always begin with a major investment. What matters first is discipline. The platform can come later when the scale and complexity demand it.
The deeper lesson is that every organization working with customer data needs some version of master data thinking. Without it, sales, marketing, finance and support will continue operating with different versions of the same truth.
The conversation becomes even more urgent when Dr. Dhillon talks about AI.
He does not believe AI usually fails because the technology itself is broken. In many cases, AI is doing exactly what it was built to do.
The problem is that it is working with incomplete, fragmented, or low-quality data.
This is what makes AI dangerous in a business context. It does not simply give wrong answers. It can give wrong answers with confidence.
He shares the example of a bank that deployed an AI agent to handle customer queries. The model itself was not an issue. The problem was the customer data available to it. The same customer existed across multiple systems, but the AI agent could only access one part of that fragmented identity.
As a result, it responded using incomplete information.
Customer satisfaction has dropped. Business confidence in the AI initiative weakened. The funding team began questioning whether the project should continue.
The lesson is clear for executives leading AI programmes. Do not blame AI when the foundation underneath AI is broken.
AI trust begins with data trust.
If an organization wants AI to support sales, marketing, customer service, or decision making, it must first make sure the data behind that AI is connected, reliable and fit for purpose.
He is not negative about AI. In fact, he sees sales and marketing as one of the most practical areas where AI can create value.
AI can improve customer experience by guiding users inside a product. It can support customers through intelligent chat or voice assistance. It can help teams understand product usage, identify customer issues and respond faster.
It can also support campaign planning.
Agentic AI can help marketers gather the right data, understand what information is needed for a campaign, identify customer and prospect segments and reduce manual preparation work. This can support prospecting, demand generation and account-based marketing.
But the condition remains the same.
AI works best when the data is ready for it.
If the customer's identity is fragmented, if duplicates are everywhere, if segment definitions are unclear, or if data sources are not trusted, AI will only accelerate confusion. It may help teams move faster, but in the wrong direction.
For leaders, the opportunity is not simply to add AI into sales and marketing. The opportunity is to combine AI with strong data quality, clear process design and business context.
That is where real value begins.
The hardest lesson he shares is that technology is almost never the whole answer.
Early in his career, he believed that if something was built well enough, accurate enough and smart enough, people would naturally use it. But the experience taught him otherwise.
Technically strong projects can fail quietly if people do not trust them. Modest projects can succeed if teams believe in them and use them consistently.
This is one of the most important leadership lessons in the conversation.
Technology may be half the job. The other half is human. It is a culture. It is behavior. It is helping people let go of familiar spreadsheets, manual processes and old habits. It shows them why a new way of working is better and gives them enough confidence to use it.
This is especially important in data and AI programmes. A model that no one trusts will not change the business. A dashboard that no one uses will not improve performance. An AI agent that customers reject will not create value.
The goal is not to launch technology.
The goal is to create trusted adoption.
His message is timely for every executive investing in data, AI, sales growth and marketing performance.
Before asking which platform to buy, leaders should ask what business decision they are trying to improve. Before blaming a campaign, they should examine the data behind it. Before scaling AI, they should ask whether the information feeding it is complete, connected and trusted.
Data quality in sales and marketing is not only a technical foundation. It is the foundation of customer experience, campaign efficiency, AI trust and business confidence.
The organizations that understand this will not simply collect more data. They will build discipline to make their data usable.
Because in the end, the value of data is not how much of it a company owns.
The value is in what trustworthy data allows people to do.
Dr. Gurpinder Dhillon joined The Executive Outlook to share his perspective on data quality, master data management, AI, sales and marketing performance and what it really takes to build trust in business decisions.
Listen in for business strategies and interviews with business experts to help you build a business and a life that you love.
About Us
