When good data goes bad: Fighting fraud and silos in Southeast Asia's largest economy

In today’s fast-paced digital economy, the customer journey is paramount. However, businesses are quickly learning this journey is only as smooth as the data that underpins it. Customers no longer judge brands on their promises or marketing messages, but on the seamlessness of their digital experiences.

“That trust is only as good as the data that you have,” explained Soren Beaulieu, regional content and strategy director at MARKETING-INTERACTIVE, who moderated a recent roundtable luncheon.

Hosted in Jakarta by MARKETING-INTERACTIVE, and supported by GBG Loqate, the leading provider of address verification solutions, the event brought together leaders from across Indonesia’s finance, logistics, and retail sectors.

The discussion unearthed a critical, and often, underestimated challenge: the staggering operational and financial costs of inaccurate customer data. As Gunaseelan K, head of product at GBG Loqate, pointed out, the impact goes beyond the balance sheet, becoming “a real distraction from your actual business”.

The foundational fissures: Data silos and duplication

For many organisations, the data problem begins at the most fundamental level: data entry. Iyan Waer, head of IT, architecture and development at Adira Finance, shared a common, but costly scenario where sales agents, faced with repeat customers, find it easier to create new profiles rather than search for existing ones.

“What they would usually do is just input everyone as new customers because that’s the easiest method for them,” he explained. This simple shortcut creates a ripple effect of chaos. “In our system, we have a lot of duplicate data, and that is very challenging for us. When we need to do some kind of algorithms, we don’t know which one is the real ‘Pak Budi’ for example.”

This internal confusion often spills into high-level meetings. With different departments working from their own version of the truth, productivity grinds to a halt. “You bring your own data, and they bring their own data. So sometimes the meeting is not efficient because we are just arguing over which data is the best,” Waer lamented.

The challenge is compounded when businesses operate across regulatory boundaries. Mandiri Sekuritas’ vice president of digital marketing, Fadilla Zain, described the complexities of integrating banking and capital market platforms, which are governed by separate regulators.

“The data that they already have and that’s required by our regulators is actually different,” she noted. This forces the company into a constant balancing act, “juggling those two things to comply to the regulators while also delivering a good experience for our users”.

The pressure is similarly immense for a digital-first entity such as Bank Neo Commerce, the leading digital-based bank in Indonesia. Daniel Armanto, the bank’s chief technology director, confirmed its biggest challenge is ensuring the data is of the “highest quality” to conduct the rigorous, digitally native, know your customer (KYC) processes, required by regulators.

When data fails: Fraud, forfeited revenue, and logistical nightmares

Inaccurate data is not just an internal headache; it opens the door to significant financial losses from fraud and operational failure. Waer shared a chilling story of a sophisticated fraud syndicate that exploited the system by using the legitimate, verified documents, of real people. “The data we received when they were applying for a loan was valid, but this was actually a syndicate borrowing data from real people,” he recounted.

After making a few initial payments, the loans would default. When the collections team visited the address, the real individual had no knowledge of the loan, having been paid a small fee to lend their identity documents for a few days. The collateral was gone, and the loan was a total loss.

The experience forced Adira Finance to rethink its digital strategies and adopt a hybrid model. “We needed to go to ground, do a physical survey, and physically meet with customers because we couldn’t trust our digital methods entirely at that moment,” Waer stated, highlighting the limitations of relying on digital data alone.

This high-stakes reality is also felt in the logistics sector. Liberty Fauquex, director of technology for ASLI Satu Indonesia, explained how the company faces risks with international orders, where a bogus address can be disastrous.

If a shipment arrives at a port and is not collected, “it becomes the property of the state”, a costly and irreversible outcome. This necessitates a “manual” and time-consuming due diligence process for every suspicious order. A solution that could validate international customers “a lot easier and a lot faster, like the Loqate’s solution” he said, “would be a game changer”.

Even the F&B industry is not immune. Christian Wibowo, vice-president of product at Fore Coffee, described how syndicates of resellers, known locally as jastip, exploit new user promotions. “From the data, that’s a new user. That’s a real customer,” he said.

The company only uncovered the scheme through reports from their baristas, who noticed the same individuals picking up hundreds of drinks ordered under different new accounts. While the data appeared valid on the surface, it was being used in a way that undermined its business model, forcing it to explore solutions such as CCTV analysis to identify and block these actors.

The address conundrum: A landmark is not a location

Among the myriad data challenges discussed, the issue of address verification emerged as a uniquely pervasive and frustrating obstacle. It’s a problem every consumer has faced: the delivery driver who calls asking for a nearby landmark. “They shouldn't have had to call me,” Fauquex asserted. “I put that in when I ordered the goods.”

Gunaseelan pointed out that even official, government-issued documents, may not always be completely accurate. Holding up his own Malaysian ID card, he noted: “There’s ambiguity in my address as it lacks a complete street name.”

If a business relies solely on this information for KYC processes, it is basing its entire relationship on inaccurate data. The problem is amplified in Indonesia, where, as Zain said, ID card images can be so poor that agents resort to editing them in social media apps just to make the text legible for optical character recognition systems – a stark example of a costly manual workaround for a systemic data quality issue.

Luwize Tan, who heads commercial efforts for GBG Loqate, shared his own frustration as a consumer. While trying to onboard with a digital app in Indonesia, his address at a hotel was rejected because the “reference data format was totally different” from the provider’s database. The only solution was a manual video call to prove his location. These small frictions, replicated at scale, lead to abandoned carts, frustrated customers, and lost revenue.

The ripple effect: Distractions, hidden costs, and the search for a cure

The cumulative effect of these daily data battles is a significant “distraction” from core business functions, argued Gunaseelan. “Your IT department was not hired to do addresses,” he said. “You end up focusing your resources on [something that] is not your job. It’s not your industry.”

The time and money spent cleansing data, verifying identities, and manually resolving issues could be invested in revenue-generating activities. “If you could shift it left,” he proposed, “and find out the bad data upfront, you could spend 50% more of that time on your actual business.”

The financial impact can be colossal, often stemming from a tiny data error. Wibowo recounted a painful experience where inaccurate inventory forecasting, caused by baristas not following the correct standard operating procedure (SOP), led to a massive overstock of condensed milk. The company was facing a loss of “hundreds of millions” of rupiah. The crisis forced it to rapidly develop and market a new product just to use the stock before it expired. The key lesson, he said, was the urgent need to “automate some of the SOP” to eliminate the risk of human error in its data pipeline.

For Niek Budiasih, head of IT at SeaInsure Indonesia, the challenge is one of pure economics. When providing low-cost insurance products through an eCommerce platform, the cost of performing a regulatory-mandated KYC process check can be greater than the insurance premium itself. This puts the business in a difficult position, forced to find innovative, cost-effective ways to deliver affordable, compliant insurance products, without incurring losses on policies sold.

As the discussion concluded, a clear consensus emerged: the manual workarounds and reactive problem-solving that have long defined data management are no longer sustainable.

The strategic imperative, as Gunaseelan suggested, is to “shift left” – tackling data quality not as an afterthought, but at the very first point of entry. By verifying information upfront, businesses can stop being data janitors and refocus their valuable resources on their core missions.

For Indonesian businesses navigating the complexities of the digital age, a foundational investment in intelligent data verification is not just a competitive advantage, it is the cornerstone of building lasting customer trust, operational resilience, and sustainable growth.

This article was sponsored by GBG Loqate. To learn more, visit www.loqate.com.