The refurbished device market in Southeast Asia is not one market. It is a collection of city-level and channel-level markets that share a similar supply side but differ substantially in demand structure, price sensitivity, and preferred device tiers. Operators who treat it as a single market and price accordingly are leaving money on the table in some cities and losing deals in others.
This matters practically because most buyback pricing infrastructure in the region was designed around national averages. A supplier in Jakarta sets a national price list, ships to distributors in Surabaya, Medan, and Makassar, and expects the same margins to hold. They rarely do, and the gap is not random. It tracks predictably against city-level income distribution, smartphone penetration rates, and the presence or absence of strong informal resale networks.
Why City-Level Demand Diverges
Jakarta is the most developed secondary market in Indonesia. Supply of refurbished devices is relatively high because Jakarta is both a primary upgrade market and a collection point for trade-ins from carrier retail. The density of formal and informal resellers means competition is strong, price discovery is faster, and buyers are more sophisticated about condition grades. A Grade B device that sells without negotiation in Surabaya might need a 5 to 8 percent discount in Jakarta to move at the same pace because Jakarta buyers have seen more inventory and have more comparison options.
Surabaya operates with a tighter supply of mid-range refurbished devices and a buyer base that is less concentrated in the upper income brackets that drive premium model demand in Jakarta. Demand for the IDR 1.5 million to IDR 2.5 million refurbished handset band is proportionally stronger here than in Jakarta. A model that is slow-moving in Jakarta at a given price might clear faster in Surabaya at the same price because it falls in a more attractive tier for that market.
Manila has a different structure again. The Philippine secondary market has historically been dominated by pasalubong imports and gray-market handsets, which depresses formal refurbished channel pricing in categories where the gray market supply is strong, particularly for iPhones. The formal refurbished channel for Android mid-range, where gray market supply is weaker, is more competitive and commands closer to the regional average price for comparable condition grades.
These differences are not speculative. They are visible in listing price variation across local marketplaces for the same model, same grade, and same storage tier. The variation is not noise. It is structural, and it is large enough to matter for pricing decisions.
What Demand Data Actually Measures
When we talk about demand data for refurbished devices, we mean several things that are often conflated.
Listing price data is what appears on marketplace platforms: the price a seller is asking. It tells you what sellers think the market will bear, not what buyers are actually paying. Listings that sit unsold for three weeks at a given price are not demand signals. They are supply at a price the market rejected.
Transaction price data is what devices actually cleared at: the final sale price after any negotiation. This is harder to collect but far more useful. The gap between listing price and transaction price for a specific model and grade in a specific city tells you how much negotiation pressure exists and in which direction. When transaction prices consistently come in below listing prices by more than 5 percent, the listing prices are not anchored to real demand.
Velocity data is how quickly devices move at a given price. A model that clears in three days at IDR 1.1 million and in six days at IDR 1.2 million is priced differently than one that clears in two days at IDR 1.1 million and two days at IDR 1.2 million. The latter has stronger demand and should command a higher acquisition offer at intake, because the sell-side risk is lower.
Most operators in Southeast Asian refurbished markets are working with listing price data and calling it demand data. The distinction matters for pricing decisions that affect margin.
The Grade-Demand Interaction
Demand is not uniform across condition grades for the same model. This is obvious in principle but rarely operationalized in pricing systems.
In Jakarta mid-range Android resale, the Grade A to Grade B spread is typically wider than the underlying condition difference would suggest, because the Grade A tier attracts buyers who are making a considered purchase and are willing to pay a meaningful premium for near-new condition. The Grade B tier attracts buyers who are primarily price-sensitive. The two buyer populations are not on the same demand curve.
In smaller cities with lower average transaction values, the Grade A premium is narrower because fewer buyers have the budget to chase cosmetic perfection. Grade B and Grade B-plus devices move as quickly or faster than Grade A at a compressed spread, because the buyer population is predominantly price-sensitive.
This interaction has a direct implication for intake pricing: in Jakarta, it is worth paying up for Grade A devices because the sell-side premium is real and the demand is there to clear them quickly. In a city where the A/B spread is compressed, overpaying for Grade A at intake means you will recover less of that premium at sell. The optimal intake offer matrix is city-specific, not national.
Model-Level Demand Variation
Southeast Asia's secondary market is heavily weighted toward Android mid-range, specifically the price-performance leaders in the IDR 1 million to IDR 3 million retail equivalent band. In this segment, consumer brand loyalty is weaker than in premium segments, and the demand curve shifts faster when a successor model lands at a competitive price point.
A specific model might have strong secondary market demand in month three after its retail launch, when early upgraders start selling them. Demand peaks, then plateaus, then begins declining about six months after launch as the successor model enters the pipeline and buyers start anticipating it. This cycle is compressed in the mid-range segment compared to premium flagships, where brand loyalty holds secondary prices longer.
For operators who are buying and selling the same model over a multi-month period, this demand curve is a predictable pricing input. A device bought at intake in month three should be priced with the awareness that it will be competing against successor model supply in months eight to ten. Static pricing that ignores this trajectory leaves money on the table in the early period and takes losses in the late period.
What a Regional Operator Actually Needs
The practical ask here is not a perfect demand signal or a real-time price feed. It is a system that is wrong about demand in a predictable direction rather than a random one.
A static national price list updated monthly is wrong about city-level demand, and it is wrong in a random direction: it might be too high in Jakarta for Grade B units this week and too low in Surabaya for the same units next week. That random error is what causes the uneven margins and the inventory that sits in one location while another is understocked.
A pricing system that incorporates city-level listing velocity, recent transaction price signals, and model-level demand curves is not perfect either. But its errors are systematic and correctable. When you can see that Surabaya Grade B mid-range handsets are clearing in four days at your current price, you know the demand signal is good. When you see they are sitting for twelve days, you know the price is off and in which direction to adjust.
That kind of visibility, by city and by grade, is what changes the margin profile for a regional operator. The data to build it exists in the markets already. The gap is the infrastructure to collect it at the right granularity and connect it to the pricing decision at intake.