XDBS founder Kartik Anand argues that when competitors can buy the same tools, the advantage depends on how a business interprets information and on developing people who can act on it.
An accurate customer database can still leave a sales team unsure whom to call. A campaign can hit every delivery target without producing a useful commercial conversation.
As AI reduces the time and cost of routine execution, parts of the work a B2B intelligence provider does become easier for others to reproduce. That puts pressure on providers to demonstrate what their expertise contributes beyond the output.
Anand founded XDBS in 2012 and has grown to more than 2,500 people, with operations across the Middle East, India, the United States, South Africa, and Costa Rica. At that scale, connecting delivery with customer outcomes requires information to travel back through the business, not simply out to the client.
“Clients want a clearer connection between the work they buy and the commercial result,” Anand says.
Meeting that demand means following the work past delivery. Accurate information is a starting point. Its usefulness depends on whether it helps a customer recognize an opportunity and decide how to act.
“The hardest change is getting closer to what happens after delivery,” he says. “That requires better feedback from the client’s sales organization and a willingness to examine how clients use the intelligence we provide.”
Better feedback can reveal whether targeting was wrong or apparent interest was weak. But a provider cannot control a client’s pricing or ensure its sales team follows up effectively. Accountability needs to reflect those boundaries without becoming an excuse to ignore disappointing results.
“We should own the quality of what we deliver, while being honest about the factors in conversion that we don’t control,” Anand says.
That requires a more demanding conversation than whether a campaign met its delivery commitment. Both sides need enough visibility to establish what happened next.
WHAT SURVIVES WHEN EVERYONE HAS THE TOOLS
AI investment is an incomplete answer to competitive pressure. A business can get faster without getting harder to replace.
“Access to AI alone won’t be much of a distinction,” Anand says. “A competitor can buy many of the same tools.”
He puts greater weight on the information those tools use and the provider’s understanding of the customer. Knowing someone’s role and employer does not establish whether they have a relevant problem or the authority to address it.
“A verified contact is useful, but it doesn’t tell you whether that person has a relevant need or whether the timing is right,” he says.
The same test applies to proprietary technology. Owning a platform means little unless it improves the work in a way the customer notices: identifying a more relevant audience, for example, or making delivery more reliable. The advantage has to survive scrutiny beyond a demonstration.
“The test is whether clients would find the business difficult to replace after its competitors gained access to similar AI,” Anand says.
For an intelligence provider, learning from results is commercially important. When a campaign underperforms, the team needs to investigate the failure and use what it discovers. Producing more activity without understanding the problem simply repeats the expense.
Also Read: AITech365 Interview with Raj Koneru, CEO/founder of Kore.ai
COUNT THE CHECKING BEFORE CLAIMING THE SAVING
The argument raises an obvious question: What happens when AI contributes to the judgment itself? If a system can help assess an account or recommend whom to approach, human interpretation cannot simply be assumed to remain the advantage.
The practical test is whether the technology improves performance once you count checking and corrections.
Inside a business, faster execution can obscure the work that remains. A report generated in minutes still needs its sources assessed. An automated recommendation may require investigation before someone can act on it.
“If a tool produces something in minutes but creates hours of checking, the saving is smaller than it appears,” Anand says.
“I want the business problem explained before the technology.”
His starting point is the existing operation: where effort goes, which failures are expensive, and who is accountable for fixing them. A pilot needs to be measured against that baseline under normal working conditions, including difficult cases. Implementation costs count, as does the effort required to get people using the system consistently.
The handover into daily operations makes ownership critical. Someone must maintain the improvement after the implementation team moves on.
“There should be a named operating owner who can explain the result and take responsibility for sustaining it,” he says.
Success is measured by what happens to the whole assignment, including the work required to make the output usable.
GROWTH EXPOSES THE COST OF UNCLEAR OWNERSHIP
As the business expands, preserving the context behind decisions becomes a management challenge.
More customers and employees mean more handovers, often between people who were not involved in the original decision. A new team inherits a commitment without necessarily understanding why it was made.
“Early in a business, people carry a great deal of context in their heads,” Anand says. “They know why a client was promised something or why an exception was made.”
That context has to become accessible. Otherwise, employees spend their time reconstructing agreements or correcting misunderstandings that could have been avoided.
“More hiring doesn’t solve unclear ownership,” he says. “It can create more handovers and more opportunities for something to be misunderstood.”
For a business operating across markets, the implications extend beyond documenting a process. Teams need to understand who can resolve an exception and when a decision requires escalation. Adding capacity without that clarity leaves more people dependent on the same small group for answers.
FASTER WORK CHANGES HOW PEOPLE LEARN
The commercial argument has a direct consequence for employee development. If software absorbs preparation work previously assigned to junior staff, companies need to reconsider how those employees acquire the experience that the assignments provided.
“Removing routine work doesn’t automatically prepare someone for more complex work,” Anand says. “We have to design that learning deliberately.”
Take an employee assembling a client report. AI can help draft it. The employee still needs to understand where the figures came from and recognize when an explanation does not fit the evidence.
A manager who reviews only the finished document may miss how the employee reached a conclusion. Discussing the reasoning creates an opportunity to examine an assumption or show how to investigate a discrepancy. Exposure to customer conversations helps employees understand what their work is being used to decide.
“That takes management time,” Anand says. “We shouldn’t assume technology removes the need to teach.”
For a business selling intelligence, that investment affects the quality of what reaches the client. Employees need to explain findings, acknowledge uncertainty, and recognize when further investigation is necessary.
“I’d measure development through the responsibility someone can handle: whether they can identify a problem, explain their recommendation, and recognize when they need help,” Anand says.
That returns the question to the sales team, which holds an accurate database but is unsure whom to call. The data has value. The provider strengthens its position by helping the client decide where to focus and explaining the reasoning behind that choice.
As AI takes a greater role in those decisions, the business still needs people who can assess its recommendations and take responsibility for acting on them.
– Republished with permission from Fast Company Middle East
