Deriv’s Chief Growth Officer explains how behavioural data, AI segmentation and 90-day predictive models help forex and CFD brokers identify high-value clients, reduce churn and improve customer lifetime value.
13 July 2026: Deriv Executive Explains How AI Helps Brokers Predict Client Value and Churn
In retail trading, customer acquisition performance is commonly measured by scale: the number of leads generated, registrations completed, first-time deposits secured and accounts opened through media spending. For forex and contracts for difference (CFD) brokers, however, the more important question is no longer how many clients enter the sales funnel, but what types of clients they are. In an interview with Finance Feeds, Deriv Chief Growth Officer Prakash Bhudia said brokers are increasingly focusing on client behaviour after registration rather than judging acquisition quality solely by lead volumes or first-time deposits. (Source: Finance Feeds interview with Deriv Chief Growth Officer Prakash Bhudia; published: 13 July 2026.)
Rising acquisition costs, tighter regulatory expectations, declining tolerance for aggressive client acquisition practices and intensifying competition across paid search, affiliate marketing, social media and introducing broker channels are forcing brokers to reconsider what constitutes high-quality acquisition. Large volumes of low-intent, poorly informed or unsuitable clients can create operational pressure, weaken retention, increase chargeback and complaint risks, and reduce customer lifetime value (CLV). By contrast, a smaller number of higher-quality clients can improve conversion rates, extend trading lifecycles, increase customer support efficiency and strengthen compliance outcomes. This shift is changing how brokers design their sales funnels, moving the focus away from pursuing traffic alone towards intent signals, suitability assessments, education-led onboarding and more intelligent segmentation.
Customer Acquisition Is Only the First Step
Bhudia believes the real test begins after registration. He emphasised that most companies in the industry no longer treat customer acquisition cost and the first deposit as their principal performance measures. These figures are still tracked, but they do not by themselves explain the substance of a client relationship. Deriv also no longer analyses clients as a single broad category, instead applying a tiered approach:
Clients are categorised by status, including active, high-risk, dormant or churned, with different responses applied to each group rather than a one-size-fits-all approach.
Customer lifetime value is treated as an explicit priority in growth planning rather than a figure reviewed hurriedly after the event.
Withdrawals are treated as a critical trust-building stage. Bhudia said that, as at June 2026, Deriv had automated 97.4% of client withdrawals, while the successful completion of a new client’s first withdrawal represents one of the most important moments in establishing trust.
“In the past, user acquisition was the key priority. Now, it is only the first step. What happens after a user registers is what truly determines product and marketing decisions.”
The First Few Days Distinguish Intent from Curiosity
Bhudia said the first few days after registration usually reveal whether a client is serious or merely exploring, making it possible to distinguish intent from curiosity with a reasonable degree of clarity. According to him, the strongest early behavioural signals, broadly ranked by importance, are as follows:
Speed and size of the first deposit: clients who deposit soon after registration and commit a meaningful amount in a single transaction are more likely to become high-value clients.
Demo account usage: clients who practise and explore through a demo account before opening a live account have significantly higher retention rates than those who do not.
Completion of the first trade: clients who place at least one trade are more likely to develop long-term trading habits than those who make a deposit but never trade.
Bhudia said Deriv operates a model based on a 90-day window that uses these signals to identify 68% of future high-value clients. The company is now adding richer behavioural data, including in-app activity, feature usage and time spent on the platform, to improve the model. He also cautioned that, although speed is the most visible indicator, firms should be sceptical of any approach claiming that a single figure can provide the complete answer.
Deposit Size Alone Does Not Demonstrate Client Quality
Bhudia said market segmentation has moved beyond basic geography or account size, describing such approaches as little more than “demographics disguised as segmentation”. Deriv now uses behavioural analysis, focusing on engagement patterns, how clients respond to educational content and whether their behaviour is linked to promotional campaigns. He said the company’s AI nurturing engine and AI persona agents were built specifically to bridge this gap, treating clients according to real-time behavioural profiles rather than static tiers based on deposit size. He also warned against rejecting bonus-driven clients too early, noting that some of them become the company’s best traders after 18 months.
“The mistake is assuming that deposit size reflects intent. Deposit size only shows how much someone is capable of depositing, not what they will actually do. A large initial deposit tells you what a person can do, but not what they will do in future, and people often confuse the two.”
AI Requires Processes to Be Rebuilt Around It
Bhudia acknowledged that artificial intelligence (AI) is helping some brokers improve the quality of their sales funnels, but much of the industry still relies on static segmentation, categorising clients by location, deposit tier or acquisition channel and leaving those classifications unchanged. In his view, leading brokers distinguish themselves by using AI not merely to label clients, but to respond to their behaviour in real time. The relevant comparative data are shown below:
| Metric / Item | Value / Details | Date | Category |
|---|---|---|---|
| Proportion of client withdrawals automated | 97.4% | As at June 2026 | Trust building |
| Future high-value clients identified by the 90-day model | Approximately 68% | Current | Predictive modelling |
| Performance of AI-personalised emails | Two to 2.5 times that of standard emails | Current | Client engagement |
| Potential conversion of bonus-driven clients | Some become top traders after 18 months | Long term | Segmentation insight |
Bhudia cited Deriv’s customer service agent, Amy, as an example, saying it now handles the majority of client interactions worldwide. He attributed this result to “rebuilding the workflow from the ground up rather than automating an old script”. He acknowledged that implementing AI had not been straightforward and required substantial effort, as methods that appeared effective in theory sometimes failed in practice. In his view, the challenge is no longer whether AI tools exist, as the technology is clearly available, but whether companies are willing to redesign processes around what AI can actually do rather than forcing it into systems created before AI emerged.
Questions About AI and Client Quality for Forex Brokers
Why are brokers no longer using acquisition costs and first deposits as their principal measures?
According to Bhudia, these figures are still tracked, but they do not by themselves explain the substance of a client relationship. As acquisition costs rise and regulation tightens, brokers are paying greater attention to client behaviour after registration, including whether clients return proactively and trade sustainably rather than disappearing after making a single deposit.
Which early signals after registration are most useful for identifying client intent?
Bhudia considers the most visible signal to be the speed and size of the first deposit, followed by whether a client practises through a demo account before opening a live account and whether the client completes a first trade. Clients who place at least one trade are more likely to develop long-term habits than those who deposit funds but never trade.
Why does deposit size not necessarily indicate client quality?
Bhudia noted that deposit size only indicates how much a client is capable of depositing, rather than what the client will actually do in future. He also cautioned against rejecting bonus-driven clients prematurely, as some may develop into the company’s strongest traders after approximately 18 months.
How does Deriv use AI for client segmentation?
Deriv applies behavioural analysis, focusing on engagement patterns and how clients respond to educational content. Its AI nurturing engine and persona agents treat clients according to real-time behavioural profiles rather than static deposit tiers. The company says its 90-day model can identify approximately 68% of future high-value clients.
What must brokers do for AI to be effective?
Bhudia believes leading brokers no longer use AI merely to describe clients, but to serve them in real time. The key issue is not whether AI tools exist, but whether companies are willing to rebuild processes around AI’s actual capabilities rather than forcing the technology into legacy systems.
