Payment Gateway Optimization for Bangladeshi E-commerce

Most Bangladeshi e-commerce brands lose more revenue to payment failure than to ad inefficiency, and the cost is invisible because nobody measures it properly.

A customer fills the cart. Reaches checkout. Selects payment. The transaction fails — gateway timeout, OTP not received, network drop, declined card, bKash app crash. The customer doesn't retry. The order is lost.

For most Bangladeshi e-commerce brands, this happens to 15-30% of checkout attempts. The brand sees the successful 70-85% in their conversion metrics. The 15-30% that failed get categorized as "abandoned cart" if they're tracked at all, and the brand assumes the customer changed their mind rather than understanding that the customer tried to pay and couldn't.

This post is about the layer of e-commerce optimization that gets dramatically less attention than ads, creative, or conversion rate optimization — but typically has higher impact when you actually fix it. Payment infrastructure. Gateway selection. Failed transaction recovery. CoD economics. The operational mechanics of how money actually gets from Bangladeshi customers to your business.

If you run e-commerce in Bangladesh and you don't have a clear view of your payment success rate by gateway by method by device, the optimization opportunity here is probably larger than whatever you're currently working on.

The Bangladesh payment landscape in 2026

Bangladeshi e-commerce payment splits across roughly five categories. The exact mix varies by brand and category, but the general distribution looks something like this:

Cash on delivery: 50-75% of orders for most brands. Still dominant despite years of industry effort to shift toward digital payments. The reasons are well-documented — customer trust in receiving products before paying, low credit card penetration, comfort with cash, and accumulated friction in digital payment experiences. CoD is structurally embedded in Bangladeshi e-commerce in ways that aren't going away soon.

Mobile financial services (MFS): 15-30% of orders. bKash dominates this category, with Nagad as a meaningful second and Rocket, Upay, and others taking smaller shares. The growth of this category has been substantial over the past five years and continues to compound.

Card payments (credit/debit): 5-15% of orders. Lower penetration than most international markets due to limited card-holder base. Skews toward higher-AOV purchases, urban customers, and brands serving more affluent demographics.

Internet banking: 1-5% of orders. Smaller share than MFS, used by customers who prefer their bank's online portal over MFS apps. Available through gateways like SSLCommerz, Aamarpay, and ShurjoPay.

Buy-now-pay-later and EMI: emerging. Growing but still small. ShopUp's Baki and similar offerings are creating new options. Card-based EMI through banks remains the larger BNPL-equivalent option.

The strategic implication: a Bangladeshi e-commerce brand needs to handle multiple payment categories well, not just one. The brands that optimize aggressively for digital payment success while ignoring CoD economics leave significant revenue on the table. The brands that focus only on CoD without building digital payment infrastructure miss the segment of customers who specifically prefer digital and won't order otherwise.

One thing that has surprised us across several Bangladeshi e-commerce brands is how different the payment mix can be from what the business owners initially assume. In many cases, management believes most customers are paying online, but when we look at the actual transaction data, Cash on Delivery still accounts for a much larger share of completed orders than expected.

We've also seen payment behavior vary significantly by product category. For example, customers buying lower-ticket items are often comfortable paying online, while higher-value purchases tend to generate a stronger preference for Cash on Delivery or other trust-based payment methods. This is one of the reasons we always encourage brands to analyze payment performance at a segment level rather than relying on overall averages.

The biggest lesson has been that optimizing payment gateways isn't just about adding more payment options—it's about understanding how different customer groups actually prefer to pay and removing friction from that process.

The gateway landscape

The payment gateway market in Bangladesh has consolidated meaningfully over the past few years. The major players for e-commerce specifically:

SSLCommerz. The largest and most established. Handles essentially all payment methods (cards, MFS, internet banking) through a single integration. Pricing tends to be standard market rate. Documentation is reasonable. Reliability has historically been strong. The default choice for most Bangladeshi e-commerce brands and probably the right choice for most brands starting out.

Aamarpay. Strong competitor to SSLCommerz with similar coverage. Often competitive on pricing. Documentation and support have improved substantially. Some brands prefer their interface and reporting. Worth comparing against SSLCommerz on actual quotes rather than assuming SSLCommerz is automatically the best deal.

ShurjoPay. Newer entrant that's grown rapidly. Often aggressive on pricing for newer customers. Some brands report strong support responsiveness. Less established track record than the older gateways but worth considering particularly for smaller operations.

bKash for Business and Nagad Payment Gateway. Direct MFS integrations rather than aggregator gateways. Lower fees on MFS transactions if you process meaningful MFS volume. Worth considering as supplementary integrations alongside an aggregator gateway, particularly if your customer base skews heavily toward one MFS provider.

Stripe, PayPal, international gateways. Generally not optimal for Bangladeshi e-commerce serving Bangladeshi customers. Used by brands serving international customers from Bangladesh, but for domestic Bangladeshi commerce, local gateways are typically more cost-effective and provide better local payment method coverage.

From our experience at Ngital, there isn't a single payment gateway that's best for every e-commerce business. The right choice usually depends on the business model, average order value, customer demographics, and the level of technical support required.

What we've found is that reliability matters more than feature lists. A gateway can offer dozens of features, but if customers experience failed transactions or inconsistent payment confirmation, conversion rates suffer quickly. That's why we pay close attention to transaction success rates, settlement speed, support responsiveness, and how easily the gateway integrates with the client's existing systems.

One thing we've learned over the years is that merchants often evaluate gateways based on pricing alone. In reality, a slightly higher processing fee can be worthwhile if it results in a smoother checkout experience and fewer failed payments. For most growing e-commerce brands, customer trust and successful transaction completion have a much bigger impact on revenue than small differences in gateway fees.

The gateway selection question for most brands isn't "which is best in absolute terms" but "which fits our specific needs best given customer payment preferences, transaction volume, and operational priorities." The right answer varies. Brands processing under BDT 50 lakh monthly often do fine with a single primary gateway. Brands processing higher volumes typically benefit from multi-gateway architecture for redundancy and optimization opportunities.

The failed transaction problem

Failed transactions are the single largest revenue leak in Bangladeshi e-commerce payment infrastructure, and the brands that take this seriously consistently outperform brands that don't.

The categories of payment failures and what's actually happening:

Gateway timeout failures. Customer reaches the payment step, gets redirected to gateway, the gateway times out before completing. The customer is left in a confused state — was their payment processed or not? Many don't retry. The transaction is recoverable if you can identify the customer and reach out, but most brands don't have the tracking to do this.

OTP failures. Customer attempts to pay via card or MFS, OTP doesn't arrive in expected timeframe, customer abandons. This is particularly bad on mobile where switching to the SMS app and back can disrupt the payment flow. The OTP delivery problem isn't your gateway's fault directly but it affects your conversion rate.

Insufficient balance. Customer attempts MFS payment with inadequate balance. The transaction fails. Different MFS providers handle this differently — bKash gives a clear "insufficient balance" message; some others give generic failure messages that confuse customers.

Declined cards. Cards declined for various reasons including expired cards, exceeded limits, fraud detection, bank-side issues. Recovery rate varies dramatically based on how the decline is communicated.

App crashes during MFS flows. Particularly bKash app crashes during payment flows — happens more than the industry admits. Customer is in the middle of typing PIN, app crashes, customer returns to your site uncertain whether payment went through.

Network drops. Particularly common on slower mobile connections outside Dhaka. The payment flow requires multiple network handshakes; a single dropped packet can fail the entire flow.

3D Secure friction. Card payments increasingly require 3D Secure authentication. The flow adds steps and friction. Some customers abandon at the 3DS step.

The aggregate impact varies but typical Bangladeshi e-commerce brands lose 15-30% of attempted payments to some combination of these failures. The recovery opportunity is substantial.

The recovery layer most brands don't build:

Failed transactions should trigger immediate recovery flow. The customer is still on your site with intent to buy. The window for recovery is minutes, not hours or days.

Tactical implementation: when payment fails, the customer sees a clear failure page that:

Explains what happened in plain language (not generic "transaction failed"). Offers immediate retry with the same payment method. Offers alternative payment methods (different gateway, MFS instead of card, CoD as fallback). Captures customer contact for follow-up if all immediate retries fail.

The follow-up matters substantially. Customers who fail at payment but provide contact information should receive WhatsApp or SMS within 15-30 minutes offering assistance and an easy completion path. Conversion on this recovery flow can recover 20-40% of failed transactions that would otherwise be permanently lost.

Most Bangladeshi e-commerce brands have none of this infrastructure. The customer fails, sees a generic error, and disappears. The brand counts it as abandonment without realizing the customer was actually trying to pay.

One of our e-commerce clients noticed a recurring pattern where customers would reach the payment stage but never complete the transaction. Initially, the team assumed these were abandoned purchases and focused most of their efforts on acquiring more traffic.

After digging into the data, we found that a significant portion of those customers had actually attempted to pay but experienced failed or incomplete transactions. The business had no structured process for identifying or recovering those orders.

We helped the team create a simple recovery workflow. Failed payment attempts were tracked separately, customer service received alerts for high-value orders, and customers were contacted through SMS, phone calls, or messaging channels with instructions to complete their purchase. The business also reviewed checkout friction points and simplified parts of the payment process.

The result was that orders that would previously have been written off as lost sales started converting into completed purchases. More importantly, the company gained visibility into a problem that had been hidden inside its overall abandonment numbers. It was a good reminder that not every lost order is a marketing problem—sometimes it's an operational one.

CoD economics that brands miss

Cash on delivery looks like a fully successful payment from most brand dashboards. The order is placed. The conversion fires. The campaign optimizer counts it as a win.

The reality is substantially more complex. CoD's actual economics involve several layers most brands don't track properly.

Delivery success rate. Not every CoD order results in successful delivery. Refused at door, wrong address, customer not reachable, customer changes mind. Across Bangladeshi e-commerce categories, 15-35% of CoD orders fail to deliver successfully — varying by category, geography, customer segment, and the delivery partner used.

Time to cash receipt. Even successful CoD deliveries take days to weeks to result in cash actually reaching the brand. Your delivery partner collects cash on delivery, processes it through their system, and remits to you on their settlement schedule — typically 3-14 days depending on partner. The cash flow implications differ substantially from digital payments where money arrives within 1-3 days.

Cost of returns and handling. Failed CoD deliveries don't just lose revenue — they incur costs. The delivery partner charges for the failed attempt. The product comes back to your warehouse, requires processing, may require restocking or quality checks. The per-order cost of failed CoD is meaningful and often unmeasured.

Customer quality differential. Customers who use CoD have measurably different quality profiles than customers who pay digitally. CoD customers have higher refusal rates, lower repeat purchase rates, and lower lifetime values on average. This is industry-wide pattern, not brand-specific.

The strategic implication: brands optimizing toward "orders placed" without distinguishing between CoD and digital are over-investing in CoD acquisition that has worse downstream economics than the dashboards suggest. Brands that track delivered revenue rather than placed revenue make different and better media allocation decisions.

The fix on the tracking side: implement delivery-confirmed conversion events flowing to Meta, Google, and TikTok rather than order-placed events. The platforms then optimize toward audiences that actually pay rather than audiences that just place orders. The CPA looks worse on paper (because you're counting fewer "conversions") but the actual customer acquisition cost in delivered-revenue terms typically improves substantially within 60-90 days.

I covered the technical implementation of this in Conversion API Setup Across All Major Platforms. The operational implementation requires integration between your delivery confirmation system and your tracking infrastructure.

The CoD reduction question:

Some brands try to reduce CoD share through pricing incentives (digital payment discounts) or friction (CoD surcharges, deposit requirements). The results vary substantially.

What works: small digital payment incentives (1-3% discount for prepaid) shift some customers, particularly those who were near-indifferent between methods. Free or expedited shipping for prepaid orders shifts more. Combining incentives with educational content about digital payment safety has compounding effect.

What doesn't work well: aggressive CoD surcharges or limits. They lose customers more than they shift behavior. Bangladesh's CoD preference is structural enough that brands trying to force the issue typically reduce conversion more than they improve mix.

The realistic medium-term target for most brands: shifting from 70-75% CoD toward 50-60% CoD over 18-24 months through sustained but moderate incentive structures, while maintaining or growing total order volume.

Mobile financial services optimization specifically

MFS payments — primarily bKash, secondarily Nagad — deserve their own treatment because they're the growth category and the optimization opportunities differ from cards or CoD.

The integration architecture decision:

You can accept MFS through your payment gateway aggregator (SSLCommerz, Aamarpay, etc.) or through direct integration with bKash for Business / Nagad's merchant APIs.

Aggregator route: simpler integration, single relationship, slightly higher per-transaction fees on MFS. Right choice for most brands starting out or operating at moderate volumes.

Direct integration route: more complex implementation, separate operational relationships, lower fees on MFS transactions. Worth considering when MFS volume becomes substantial enough that fee savings justify the integration complexity.

The crossover point varies. Brands processing under BDT 10 lakh monthly in MFS transactions typically don't gain enough from direct integration to justify it. Brands processing BDT 30 lakh+ monthly in MFS typically do. Brands in between are case-by-case.

The OTP and PIN flow optimization:

MFS payment flows involve multiple steps — entering MFS number, receiving OTP or being redirected to MFS app, entering PIN, confirming payment. Each step is a friction point where customers drop off.

The brands optimizing well here:

Pre-fill MFS number from customer profile when available. Use deep linking to MFS apps where supported, reducing manual app-switching friction. Provide clear visual progress indicators so customers know how many steps remain. Time out gracefully with clear recovery options rather than generic failure. Capture payment intent before MFS flow starts, so failed payments can be recovered.

These are modest individual improvements but compound substantially. A checkout flow optimized for MFS friction typically converts MFS payments 5-15% better than an unoptimized flow.

bKash Pay versus standard bKash payment:

bKash now offers multiple payment integration options. bKash Pay (their newer flow) provides smoother experience than standard merchant payments in many cases. Worth evaluating which option your gateway integration is actually using and whether upgrading produces measurable conversion improvement.

From what we've seen at Ngital, simply adding bKash as a payment option doesn't automatically improve online payment adoption. The brands that get the best results are usually the ones that make bKash highly visible throughout the customer journey, not just on the final checkout page.

One pattern we've noticed is that customers are much more likely to complete a payment when they already know their preferred payment method is available before they reach checkout. We've seen merchants improve payment completion rates by highlighting bKash availability on product pages, promotional campaigns, and even within ad creatives.

We've also learned that the technical implementation matters. A smooth handoff between the website and bKash, clear payment instructions, and fast payment confirmation create a noticeably better customer experience. On the other hand, confusing redirects, delayed confirmations, or unclear error messages often lead to customer support requests and abandoned orders.

Perhaps the biggest takeaway is that bKash works best when it's treated as part of the overall conversion journey rather than just another payment gateway. The merchants that understand how their customers prefer to pay tend to get far more value from the integration than those who simply enable it and hope for the best.

Operational metrics most brands don't track

The payment optimization opportunities I've described above can't be acted on without proper measurement. The metrics that should be in your operational dashboard but probably aren't:

Payment attempt rate. Not just successful payments — total attempted payments including failures. Without this, you can't measure success rate.

Payment success rate by gateway by method. Where are failures concentrated? Different gateways have different failure patterns; different payment methods within the same gateway have different patterns.

Payment success rate by device type. Mobile typically has higher failure rates than desktop. Specific mobile manufacturers and browsers have specific patterns.

Time-to-payment-success. From checkout start to payment completion. Long times indicate friction; short times indicate smooth flow.

Recovery rate from failed payments. Of customers who fail a payment, how many ultimately complete? This is the ROI measurement for your recovery flow investments.

Cash flow timing by payment method. When does money actually arrive in your accounts after order placement? Should be tracked separately for CoD, MFS, cards.

Delivery success rate by payment method. CoD delivery success will be lower than prepaid delivery success. The gap is meaningful for unit economics.

Refund and chargeback rates by payment method. Different methods have different refund patterns and operational costs.

Most Bangladeshi e-commerce dashboards track aggregate revenue and aggregate order count without these granular metrics. The optimization opportunities live in the granular metrics; the aggregate metrics are too coarse to drive operational decisions.

Building this dashboard isn't complex. It requires pulling data from your payment gateway, your delivery partner, and your e-commerce platform into a single view. The work is modest for the operational value it produces.

The gateway negotiation point most brands miss

Gateway pricing in Bangladesh is more negotiable than most brands realize. The published rates are starting points, particularly for brands with meaningful processing volume.

The factors that affect what you can negotiate:

Monthly processing volume. Brands above BDT 50 lakh monthly typically have substantial negotiating leverage. Above BDT 2 crore monthly, custom pricing becomes routine rather than exceptional.

Payment method mix. Higher-margin payment methods (cards) have more room for negotiation than lower-margin ones (MFS). Your overall blend affects what gateways can offer.

Settlement terms. Faster settlement comes at cost; slower settlement saves money. Most brands accept default settlement terms without negotiating.

Multi-gateway leverage. Brands with relationships across multiple gateways have substantially better negotiating positions than brands locked into single-gateway arrangements.

Long-term contracts. Gateways offer better rates for multi-year commitments. Trade-off between flexibility and cost.

The practical move: every 12-18 months, get fresh quotes from your gateway's competitors. Use real quotes as negotiating leverage with your current gateway. Most brands that do this systematically see 10-25% reduction in effective payment processing costs over 2-3 years compared to brands that stay on default pricing throughout.

One thing we've observed at Ngital is that payment gateway negotiations are rarely won through aggressive bargaining alone. The merchants that tend to get the best terms are usually the ones that can demonstrate consistent transaction volume, strong growth potential, and a healthy payment success rate.

We've seen businesses approach gateways asking for lower fees and get very little movement. But when they come to the discussion with clear transaction data, projected volume, and a long-term partnership mindset, the conversation often becomes much more productive. In some cases, the outcome isn't even a lower rate—it might be faster settlement times, dedicated account support, technical assistance, or access to features that weren't previously available.

Another pattern we've noticed is that gateways are generally more flexible once a merchant has proven transaction history. Early-stage businesses often focus heavily on negotiating fees, while larger merchants tend to focus on operational benefits that have a bigger impact on revenue and customer experience.

In practice, the strongest negotiating position usually comes from having good business fundamentals and reliable transaction volume rather than simply asking for better pricing.

The 12-month payment optimization roadmap

For a Bangladeshi e-commerce brand looking to take payment optimization seriously, a realistic sequenced approach:

Months 1-2: Measurement foundation. Build the metrics dashboard. Document current payment success rates by gateway by method by device. Establish baseline. Identify where the biggest leaks are.

Months 2-3: Failed transaction recovery infrastructure. Implement immediate recovery flow at checkout. Build follow-up sequence for failed payments. Set up tracking to measure recovery rate.

Months 3-5: Delivery-confirmed conversion tracking. Integrate delivery confirmation with conversion tracking. Switch ad platform optimization to fire on delivery rather than order placement. Measure the impact over 60-90 days.

Months 4-6: Gateway optimization. Get fresh quotes from competing gateways. Negotiate current gateway pricing. Consider multi-gateway architecture if volume justifies. Evaluate direct MFS integrations.

Months 6-9: Checkout flow optimization. Optimize MFS payment flows specifically. Reduce friction in card payment flows. Implement device-specific optimizations. Test alternative checkout designs.

Months 9-12: CoD reduction program. Design and test prepaid incentive structures. Build customer education content. Refine the CoD-to-prepaid shift strategy based on test results.

Ongoing: Operational discipline. Monthly review of payment metrics. Quarterly gateway performance audits. Annual gateway negotiations. Continuous monitoring of failure patterns and recovery rates.

This isn't aggressive timeline; it's sustainable. The compounding effect over 12-18 months typically produces 8-15% improvement in delivered revenue from the same order volume — substantially more than most equivalent investments in advertising optimization produce.

We've seen this play out with several e-commerce brands in Bangladesh. The common pattern is that they initially focus on increasing traffic and ad spend, only to discover that a significant amount of potential revenue is being lost further down the funnel. Once tracking improves, failed payments are monitored, and delivery-confirmed data is incorporated into reporting, the business gains a much clearer picture of what is actually driving revenue.

In one case, a client was investing heavily in acquisition while assuming checkout abandonment was mainly a marketing issue. After implementing better payment tracking and recovery processes, they found that a meaningful portion of "lost" customers had actually attempted to complete their purchase. Recovering even a fraction of those transactions generated results much faster than increasing advertising budgets.

What stands out from these projects is that the biggest gains rarely come from a single change. They come from stacking improvements across tracking, payments, delivery confirmation, and media optimization. By the end of the process, the business is not only generating more revenue but also making decisions with far greater confidence because the underlying data is finally trustworthy.

What this looks like done right

A Bangladeshi e-commerce brand operating payment infrastructure well in 2026 has:

A primary gateway selected based on actual fit for their payment mix, with negotiated pricing rather than default rates.

Possibly a secondary gateway as backup, automatically routing to backup when primary fails.

Direct MFS integrations if volume justifies, alongside aggregator gateway.

Comprehensive measurement dashboard showing payment success rates, recovery rates, and unit economics by method.

Failed transaction recovery flow that recovers 25-40% of attempted-but-failed transactions.

Delivery-confirmed conversion tracking feeding ad platform optimization.

Prepaid incentive structure shifting payment mix modestly over time.

Quarterly operational review of payment infrastructure with specific improvement actions.

The compounding effect of these practices over 2-3 years is substantial. Brands that operate this way have meaningfully better unit economics than competitors operating on default payment infrastructure with no measurement layer.

Most Bangladeshi e-commerce brands have almost none of this. The optimization opportunity is large precisely because so few competitors are taking the work seriously.

If I were speaking to a Bangladeshi e-commerce founder today, I'd say this: before investing more money into acquiring new customers, make sure you're not losing the customers you already have.

Most founders are willing to spend months improving ads, creatives, or website design to gain a few extra percentage points in performance. At the same time, they often have limited visibility into payment failures, checkout friction, delivery success rates, or how much revenue is leaking after a customer decides to buy.

That's why I see payment optimization as one of the highest-return operational investments for many e-commerce businesses. It doesn't always get the same attention as marketing, but it directly affects whether revenue actually reaches your bank account.

My advice would be simple: if your payment success rates, failed transaction recovery process, and delivery-confirmed reporting aren't fully under control yet, I'd prioritize those before chasing the next marketing tactic or software tool. Growth becomes much easier when the foundation is working properly.

Ngital works with Bangladeshi e-commerce brands across Facebook Ads, Google Ads, TikTok Ads, and the broader operational infrastructure that makes e-commerce work — including payment optimization, conversion rate optimization, and tracking architecture. For payment infrastructure audits or e-commerce operational consulting, reach out at +8801601-654800.