Summary

  • Zepp’s second-quarter revenue rose 6.9% to $63.5 million and gross margin improved 1.2 points to 37.4%. The operating loss nevertheless doubled to $12.3 million as selling and marketing expense rose to $18.3 million.
  • The strongest demand evidence is upstream of the income statement: high-end T-Rex models were about half of that family’s activations, a new $169 Active tier gained share, and the dearer Balance generation added activations without displacing the older model.
  • Supply constraints, channel deployment, e-commerce fees and launch campaigns make activation, shipment, revenue and profit different clocks. The January 2027 Bip price increase will test elasticity; it does not yet prove pricing power.

Zepp’s second-quarter disclosure is unusually generous for a consumer-device company. It gives prices and activation shares for individual product families, identifies which ranges were short of supply, and itemizes much of the increase in selling expense. Those data make the quarter more intelligible. They also expose why a premium-mix story can be commercially true and financially incomplete at the same time.

The headline looks constructive. Revenue reached $63.5 million, 6.9% above the prior year. Gross margin rose from 36.2% to 37.4%, despite higher memory costs and appreciation of the renminbi. Gross profit was roughly $23.7 million by applying the reported margin to revenue. Yet the GAAP operating loss widened from $6.1 million to $12.3 million, and the net loss attributable to Zepp increased from $7.7 million to $11.3 million.

The missing bridge is not one accounting adjustment. It is the sequence through which a wearable becomes an economic result.

An activation is a demand signal, not a shipment receipt

Zepp says T-Rex 3 Pro and T-Rex Ultra 2, offered in the United States at suggested prices of about $399 and $549, represented roughly half of global T-Rex-family activations. The $169 tier inside Active rose from approximately 22% of family activations in the first quarter to around 40% in the second and 57% through 25 August. Balance 2 and the new Balance 3 generation recorded July activations more than one-third above the monthly second-quarter average, while older-product activations remained broadly stable.

These are useful observations. A customer activating a device has gone further than clicking an advertisement. Activation can show that a premium price ladder is finding real wrists and that a new model is not merely replacing the old one. But it is not identical to Zepp’s shipment, a retailer’s sell-through, the price Zepp realized, a cash collection or an active user retained months later.

The denominator also matters. Half of T-Rex activations can describe a richer family mix without describing the T-Rex family’s share of Zepp revenue. A 57% share for Active’s new tier can rise while total Active volume falls, stays flat or grows; the release does not supply that denominator. Suggested retail prices describe the consumer proposition, not the company’s net selling price after channel discounts, taxes, returns and platform charges.

The right reading is neither cynical nor promotional. Activations are leading evidence. The income statement is the settlement layer.

Four product families are moving on four clocks

Management itself draws the boundaries. T-Rex has sustained a higher-end mix but not rapid unit growth. Balance is beginning to establish momentum and has not reached scale. Bip demand met a supply constraint in the quarter, although supply had recovered by the report date. Helio Strap was also constrained, with partial recovery expected in the third quarter and full recovery in the fourth. Cheetah and Helio Strap Pro remain earlier in professional credibility and market development.

That is not one product cycle. It is a portfolio of separate queues. A popular device without inventory cannot be recognized as revenue. Restored inventory without retailer replenishment can become working capital. A successful launch can lift activations while launch expense arrives first. A higher suggested price can raise gross margin, reduce volume or change the regional and channel mix.

The distinction is especially important because Zepp’s 2025 comparison already contained a different mix event. In the second quarter of 2025, lower-margin Bip 6 and Active 2 products carried more revenue while Balance 1 was cleared at reduced prices ahead of Balance 2. The move from 36.2% to 37.4% therefore says that the portfolio became more favourable against that base. It does not isolate how much came from premium products, how much was lost to memory costs and currency, or how much remains after channel economics.

Gross margin improved before the cost of creating demand

Selling and marketing expense rose $6.2 million year on year to $18.3 million. Zepp attributes $2.9 million to new-product campaigns, $1.6 million to e-commerce platform charges, $700,000 to athlete sponsorships, $500,000 to the HYROX partnership and $500,000 to physical retail and event activations.

That list matters more than a generic statement that the company invested in its brand. Some costs are campaign investments whose return may arrive later. E-commerce platform charges are closer to a toll on current channel volume. Sponsorships can build consideration but do not come with a disclosed conversion ledger. Physical retail may expand discovery while adding a fixed operating surface.

Revenue increased by about $4.1 million from the prior-year quarter. Selling and marketing expense increased by $6.2 million. Those two numbers should not be mechanically netted: current spending may support later sales, and platform fees can vary with current activity. They do show why the 1.2-point gross-margin improvement did not reach operating profit.

Research and development expense was broadly stable at $11.5 million. General and administrative expense rose to $6.2 million from $4.4 million, with substantially all of the increase attributed to foreign exchange. The operating gap is therefore not evidence that premiumization failed. It is evidence that product margin and the full cost of demand are different accounts.

The third-quarter guide is the next channel receipt

Zepp guided to third-quarter revenue of $68 million to $73 million. Against $75.8 million a year earlier, that implies a decline of roughly 10.3% to 3.7%. Management points to an unusually high comparison and says improving July and August demand will not appear immediately because production ramp, supply recovery and channel deployment take time.

That explanation is plausible and testable. A timing story should later produce inventory availability, replenishment orders and recognized sales. If the product remains activated but the revenue bridge does not close, the missing state could be regional mix, channel inventory, discounts, returns or weaker aggregate volume. The next quarter should not be judged only against the guide; it should show where the waiting demand moved.

Cash improved, but the mechanism needs a full bridge

Cash and restricted cash reached $106.3 million, up $3.1 million from March and $11 million from a year earlier. Inventory was $62.4 million, nearly flat sequentially and down from $79.9 million a year earlier. Zepp says working-capital management more than offset the quarterly net loss.

That is encouraging, but a cash balance is not operating cash flow. Receivables, payables, restricted cash and financing can move the same total. Debt increased $6.2 million sequentially, entirely in long-term debt, while $13.3 million of short-term obligations was converted to long-term. Extending maturity can reduce refinancing pressure; it does not reduce the debt principal. The useful receipt is a quarterly cash-flow bridge that separates customer collections, inventory release, supplier terms, restricted cash and financing.

January’s Bip price is an experiment, not a conclusion

Zepp plans to raise prices across the Bip family in January 2027. The stated objective is better pricing discipline and unit economics over a longer product life. The disclosure does not provide the size of the increase, affected markets, channel treatment or expected elasticity.

Bip is the cleanest forthcoming test because its supply has recovered and its buyer proposition is established. If realized prices rise while activations and replenishment remain resilient, the company will have stronger evidence of pricing power. If volume shifts to promotions, old inventory, another Zepp family or a rival, the list-price action will have measured the boundary instead.

Premium demand is real enough to monitor. It is not yet real enough to collapse all the clocks into one number.

Sources