How a CFO Cut $140K in Annual Payment Costs in 20 Minutes

DuneBet is a MENA-focused crypto casino operating out of Dubai under an offshore gaming license, serving approximately twelve thousand monthly active players across the Gulf region. The platform runs live casino and sports betting verticals, settling in USDT and AED through a mix of local payment partners and direct on-chain transfers, generating around $5M per week in gross gaming revenue.

Products used: Payment Analytics, Cost Optimization, Chain Performance Analysis

20 minutes | full payment cost audit, from first question to approval

3 | optimization levers identified in a single session

$140K | projected annual savings approved same day

---

Challenge

DuneBet's payment processing costs have climbed steadily for two quarters, but the volumes that would explain the increase simply aren't there. GGR grew 8% over the same period. Payment volume followed in roughly the same proportion. Yet the line item on Ahmed's P&L for payment processing fees had expanded 22% — a gap of fourteen percentage points that had no obvious explanation and no clean data trail.

For Ahmed, the optics were bad before the numbers even mattered. At a platform with $5M weekly GGR, a 22% cost overrun in a single expense category was the kind of thing that demanded a precise answer. Not a theory, not a "we think it's probably the Ethereum withdrawals" — a proof, with numbers, that he could put in front of the board and act on. The problem was that payment cost data lived in three places: the crypto wallet provider's admin panel, the local AED payment partner's monthly PDF invoices, and a spreadsheet his finance analyst maintained manually by reconciling the two. Getting a full picture meant at least two days of work and a conversation with the analyst that inevitably produced three slightly different totals.

"I kept asking my team where the extra cost was coming from and getting back different answers every time. One week it was the Ethereum gas fees. The next week it was the AED gateway commission. The actual answer required looking at six months of data across three sources at the same time, and we didn't have a clean way to do that."

— Ahmed Al-Rashid, CFO, DuneBet

The frustration wasn't just about money. Every week Ahmed couldn't explain the variance was a week the finance team looked like they didn't have control of their own cost base. With a board meeting scheduled in three weeks, he needed the answer — and the fix — before he walked into that room.

---

Solution

Ahmed opened Gaming Mind AI on a Tuesday morning with one goal: find out where the extra 14 percentage points of cost growth came from and whether anything could be done about it before the end of the quarter. He had no idea the conversation would be over before lunch, with a board-ready summary and two signed-off action items already in motion.

Here's how that session unfolded:

---

Ahmed: "Our payment processing costs grew 22% over the last two quarters but volume only grew 8%. Show me a full cost breakdown by payment method."

Payment Method	Q1 Fee Rate	Q2 Fee Rate	Volume Share Q2	Status
Ethereum On-chain	0.81%	1.34%	34%	🔴 Rate spike + volume drift
AED Local Gateway	1.6%	1.6%	22%	🟢 Stable
Binance Smart Chain (BSC)	0.22%	0.22%	25%	🟢 Stable
Tron On-chain	0.18%	0.18%	19%	🟢 Stable

⚠️ Gaming Mind flags: The cost problem is Ethereum-specific and compounded by a volume drift toward the most expensive chain. Ethereum's share of withdrawals grew from 29% to 34% as the player base expanded — amplifying the impact of the fee rate spike.

The breakdown lands the punch immediately. Gaming Mind surfaces four distinct payment buckets — Ethereum on-chain, Tron on-chain, Binance Smart Chain, and the AED local gateway — with the effective fee rate for each tracked across both quarters. Ethereum's effective rate jumped from 0.81% to 1.34% of transaction value, and its volume share actually grew from 29% to 34% of total withdrawals as the platform's player base expanded. Tron remained flat at 0.18%, BSC held at 0.22%, and the AED gateway sat at 1.6% but processed a stable 22% of volume. Gaming Mind flags immediately: the cost problem is Ethereum-specific, and it's been compounded by a volume drift toward the most expensive chain.

---

Ahmed: "Why did Ethereum fees jump so much? Show me the trend month by month."

Month	Avg Gas Cost / Withdrawal	Monthly Tx Count	Fee as % of GGR
Month 1	$4.10	3,820	0.39%
Month 2	$4.40	3,990	0.42%
Month 3	$8.70	4,210	0.87%
Month 4	$11.40	4,450	1.21%
Month 5	$9.80	4,620	1.08%
Month 6	$8.20	4,780	0.94%

⚠️ Gaming Mind flags: DuneBet's withdrawal processing runs on a fixed schedule — batches every three hours regardless of network congestion. During months 3–4 of peak gas, this meant withdrawals were executing at exactly the wrong times. Shifting the batch window by as little as 90 minutes could have avoided peak-congestion rates entirely.

Six months of data makes the story concrete. Ethereum gas fees spiked sharply in months three and four — average cost per withdrawal climbed from $4.10 to $11.40 — and then partially retreated to $8.20 in month six without ever returning to their baseline. Gaming Mind overlays the transaction count data and catches something Ahmed's team had missed entirely: DuneBet's withdrawal processing runs on a fixed schedule, pushing batches every three hours regardless of network congestion. During the two months of peak gas, that schedule meant the platform was executing withdrawals at exactly the wrong times, paying peak-congestion rates that could have been avoided by shifting the batch window by as little as ninety minutes.

---

Ahmed: "What would costs look like if we'd been processing on Tron instead of Ethereum for those same withdrawals?"

Chain	Q2 Actual/Projected Cost	Saving vs Ethereum
Ethereum (actual — 34% of withdrawals)	$194K	—
BSC (counterfactual)	$83K	$111K
Tron (counterfactual)	$67K	$127K

⚠️ Gaming Mind flags: DuneBet defaulted to Ethereum for all USDT withdrawals above $500 — a threshold set 18 months ago and never revisited. Tron would have cost one-seventh as much for the same transactions, and the gap widened every time the Ethereum gas market spiked.

Gaming Mind runs the counterfactual directly against DuneBet's actual Q2 withdrawal data. If the 34% of withdrawals currently routed through Ethereum had been processed on Tron, total withdrawal costs for the quarter would have been $67K instead of $194K — a saving of $127K in a single quarter. BSC would have landed at $83K. The math is uncomfortable: DuneBet has been paying Ethereum rates for transactions where Tron would have cost one-seventh as much, and the gap widened every time the Ethereum gas market spiked. Gaming Mind notes this isn't a matter of player preference — the platform defaulted to Ethereum for all USDT withdrawals above $500, a threshold set eighteen months ago that was never revisited.

"I didn't expect to see a counterfactual that specific. It wasn't 'Tron is cheaper in general' — it was 'here is what your actual Q2 withdrawals would have cost on Tron.' That's the number I needed."

— Ahmed Al-Rashid

---

Ahmed: "Do players actually care which chain they receive on? Show me deposit and withdrawal preferences by chain."

Metric	Tron	Ethereum	Other
Players with Tron-compatible wallets	61%	39%	—
Deposit method (Tron)	53%	30%	17%
Withdrawal method (current routing)	19%	34% (above $500)	47%
D30 Retention (by deposit chain)	41%	38%	—

⚠️ Gaming Mind flags: The platform is receiving money on Tron and paying it back out on Ethereum — an asymmetry nobody had mapped. More than three in five players already hold Tron-compatible wallets, meaning chain switching requires no friction for the majority of the player base. Tron depositors also show marginally higher D30 retention (41% vs 38%), likely because lower fees make smaller, more frequent deposits practical.

This is where the data gets strategically interesting. Gaming Mind analyzes the wallet addresses behind DuneBet's deposits and withdrawals and finds that 61% of players already hold Tron-compatible USDT wallets — meaning more than three in five players could receive withdrawals on Tron today without any friction. Among depositing players, 53% already send funds via Tron. The platform is essentially receiving money on Tron and paying it back out on Ethereum, an asymmetry nobody had mapped before. Gaming Mind also flags a smaller but useful signal: players who deposit via Tron have a marginally higher D30 retention rate — 41% versus 38% for Ethereum depositors — likely because Tron's lower fees make smaller, more frequent deposits more practical.

---

Ahmed: "Is there a way to incentivize players to use Tron without killing the experience?"

Incentive Option	Quarterly Cost	Estimated Uptake	Quarterly Fee Saving	Net Quarterly Benefit
0.5% cashback on Tron deposits (cap $50/player/month)	$18K	40% of eligible players	$127K	$109K
Prompt only ("Recommended" label, no incentive)	$0	~25% uptake	$68K	$68K
No change (status quo)	$0	0%	$0	$0

⚠️ Gaming Mind flags: Gaming Mind recommends starting with the prompt-only option, measuring uptake for 30 days, and layering in cashback only if uptake stays below 30%. The prompt-only path generates $68K quarterly saving at zero incremental cost — the highest-ROI first move.

Gaming Mind models three incentive structures against DuneBet's actual player deposit behavior. A 0.5% cashback on all Tron deposits — paid in USDT, capped at $50 per player per month — would cost an estimated $18K per quarter at current deposit volumes if 40% of eligible players opted in. Against the $127K quarterly saving from chain switching, the net benefit is approximately $109K per quarter, or $436K annualized. A simpler prompt-only intervention — surfacing Tron as the "recommended" chain at the withdrawal screen with no financial incentive — is modeled at roughly 25% uptake, producing a $68K quarterly saving at zero incremental cost. Gaming Mind recommends starting with the prompt-only option, measuring uptake for thirty days, and layering in cashback only if uptake stays below 30%.

---

Ahmed: "Now look at the withdrawal batching. If we time batches around low-gas windows, what does that save?"

Batch Window (UTC)	Relative Gas Price	DuneBet Current Batch?	Potential Q2 Saving
3–5 AM	−38% to −52% vs peak	No	$22K
9 AM	+35% vs baseline	Yes 🔴	—
12 PM (Noon)	+28% vs baseline	Yes 🔴	—
3 PM	+31% vs baseline	Yes 🔴	—
8–10 PM	Moderate	No	~$8K partial

⚠️ Gaming Mind flags: DuneBet's 9 AM, 12 PM, and 3 PM UTC batches overlap almost perfectly with the three highest-congestion windows of the day. The 3–5 AM UTC window consistently runs 38–52% below peak rates. Shifting a single batch to that window requires only a configuration update — no player-facing changes.

Gaming Mind pulls six months of Ethereum gas price data and maps it against DuneBet's current batch schedule. The platform's 9 AM, 12 PM, and 3 PM UTC batches overlap almost perfectly with the three highest-congestion windows of the day — mid-morning European trading activity, the New York open, and the afternoon cross-regional peak. By contrast, the 3 AM to 5 AM UTC window, corresponding to late Dubai night and early morning Asia, consistently runs 38% to 52% below peak gas rates. Shifting a single batch to that window would have saved an estimated $22K in gas fees in Q2 alone — and requires no player-facing changes whatsoever, only a configuration update to the withdrawal scheduler.

---

Ahmed: "Give me a summary of the three actions I can take, with cost and saving estimates, that I can put in front of the board."

Lever	Implementation Cost	Quarterly Saving	Annual Saving	Risk	Effort
1. Shift nightly batch to 3–5 AM UTC	$0	$22K	$88K	Low	Config update only
2. Tron withdrawal prompt (no incentive)	$0	$68K	$272K	Low	Product UI change
3. 0.5% Tron deposit cashback (capped $50)	$18K/qtr cost	$127K	$436K gross	Low–Medium	Campaign + dev
Total (all three levers)	—	—	$140K net	—	—

⚠️ Gaming Mind flags: Combined annual saving across all three levers is $140K. Lever 1 can be implemented today with zero cost. Lever 2 requires a one-week product delivery. Lever 3 is approved for a 30-day pilot. Ahmed had a board-ready summary in hand before noon.

Gaming Mind produces a clean three-lever summary. First: shift the nightly withdrawal batch to the 3–5 AM UTC low-gas window — zero cost to implement, estimated $22K quarterly saving, low risk. Second: prompt players to select Tron at the withdrawal screen with no financial incentive — zero cost, estimated $68K quarterly saving based on 25% uptake, low risk. Third: introduce a 0.5% Tron deposit cashback capped at $50 per player per month — estimated $18K quarterly cost, estimated $127K quarterly saving at 40% uptake, low to medium risk. Combined annual saving across all three levers: $140K. Ahmed had a board-ready summary in hand before noon.

---

Results

Root cause identified in under ten minutes

The 22% cost overrun traced to two compounding factors: Ethereum gas price volatility combined with a volume drift toward the most expensive chain, amplified by a batch schedule that consistently hit peak-congestion windows. Neither factor was visible in any single existing report. Gaming Mind surfaced both in the first two exchanges.

Three optimization levers with clear ROI

All three actions — batch rescheduling, Tron withdrawal prompting, and Tron deposit incentive — were quantified against DuneBet's actual transaction history. Ahmed walked into his board meeting with specific numbers, not directional estimates, and with implementation effort rated for each lever.

$140K annual savings projection approved same day

Ahmed presented the analysis to the CEO at 2 PM the same day the session ran. The batch schedule change was approved for immediate implementation. The Tron withdrawal prompt was handed to the product team with a one-week delivery target. The deposit cashback incentive was approved for a thirty-day pilot in the following month. Total time from first question to CEO sign-off: under five hours.

Finance team recovered control of the cost narrative

The board meeting three weeks later was the first time in two quarters Ahmed could explain the payment cost line item with precision — what drove the increase, which levers had already been pulled, and what the trajectory looked like for the next two quarters. The analysis also produced a standing monthly report that monitors effective fee rates by chain, so the drift from Tron to Ethereum is caught early if it begins to recur.

"The number that got the CEO's attention wasn't the $140K. It was the fact that we were receiving money on Tron and paying it back out on Ethereum — an asymmetry we'd never noticed that had been costing us real money for over a year. That's the kind of thing you only see when you can look at six months of data across multiple chains at once."

— Ahmed Al-Rashid, CFO, DuneBet