Guming (Trans): Delivery Down; Coffee and Dine-in Hold Up

The following is a Trans compiled by Dolphin Research for $GUMING(01364.HK) FY26 1H earnings call

I. Key takeaways from the results

1. Shareholder returns: repurchased about 39 mn shares for approx. HK$800 mn, with the shares recently cancelled. This buyback largely used the portion earmarked from the CB proceeds, though further open-market repurchases are not ruled out.

a. The allocation tied to the CB for buybacks is nearly exhausted, but the company may continue buying in the market. Management did not provide a fixed pace.

b. The CB was issued at 101% with an annual coupon of approx. 1%. Proceeds also fund talent upgrades and AI investment.

c. The controlling shareholder and concert parties have not sold a single share since listing, and have net bought about 26 mn shares (~HK$600 mn). Purchases were made both directly and via physical settlement of accumulated options.

2. Guidance: full-year per-store GMV remains at −5% to +5%. Per-store sales in Jul–Aug were up YoY and in line with internal expectations, but the company will not disclose monthly figures.

a. The Jul–Aug base last year was high. Q4 needs monitoring for the lingering mix impact from last year's heavy delivery contribution, and no standalone forecast is provided.

3. 1H revenue grew 30%+, core profit up 50%+, but management warned against linear extrapolation. Mix helped GPM: last 1H had higher sales of coffee machines and other equipment, which diluted GPM and has now decreased.

a. Opex ratio fell partly due to timing. COGS includes field staff costs for expansion and ops, and slower store openings reduced one-off expansion bonuses; meanwhile merchandise and franchise fee revenue reflect previously opened stores, with franchise fees amortized over three years.

b. There were no large-scale marketing campaigns in 1H. Spokesperson and related spend will be significant in 2H; fixed R&D and G&A costs provide operating leverage, but the goal is a relatively stable opex-to-revenue ratio, with absolute staff upgrade and AI spend still rising.

c. Core profit excludes accruals for withholding tax on onshore dividends remitted to Hong Kong. This tax will still be paid upon actual distribution, but it is non-core, and H-share vs. red-chip structures affect comparability; effective tax rate likely trends modestly higher over time and can be raised gradually in models.

4. Balance sheet and capex. Some loans were repaid in 1H and leverage was proactively lowered; these loans mostly arbitraged fixed deposits with dynamic spread management, all short-term and can be unwound quickly if needed.

a. As the balance sheet expands, leverage should continue to decline. Capex increase mainly reflects land payment for the Xiaoshan plot in Hangzhou, recognized as right-of-use assets; other spend is maintenance such as warehouses and vehicles.

b. The HQ building will require investment over the next few years. Management remains disciplined on maintenance capex.

II. Details from the earnings call

2.1 Management highlights

1. Store network and opening cadence. Net adds in 1H were close to 800, with gross openings around 1,300; both metrics were below last year and below initial guidance.

a. The relative mix in tier-2 cities did not change meaningfully. By region, net adds were fewer in legacy regions, while new regions maintained healthy per-capita openings.

b. Neither net adds nor gross openings are hard targets. The company will adjust cadence as needed and communicate promptly when adjustments are material.

2. Per-store GMV and delivery reporting. Reported GMV is influenced by delivery-related factors, but delivery subsidies have been rolling off since early Feb, bringing reported GMV closer to the internal standard sales measure.

a. There is no need to change disclosure definitions for now. Per-store performance stayed relatively stable in 1H; in a falling industry beta, management aims to outperform peers, which it believes it achieved.

3. Coffee and store upgrades. More than 13,500 stores are equipped with coffee machines, covering over 90% of the network; coffee penetration and per-store coffee sales mix continue to rise.

a. Gen-6 store remodels significantly boosted dine-in and were a key driver of in-store sales in 1H. However, the scale of remodels also pressured franchisee cash flow.

b. Earlier opening hours target the morning daypart that was previously under-penetrated. Morning consumers are more inclined to pick up in-store rather than order delivery, lifting dine-in mix.

c. Breakfast 'food' categories have not fully rolled out. There will be more pilots this year, especially in 2H.

4. Disclosure and communication. The company does not disclose monthly operating data and will only provide directional quarterly commentary after each quarter.

a. The call will not address new statistics that have not been previously communicated. Operations in Jul–Aug tracked expectations, but no specific numbers will be given today.

b. Multiple profit metrics are presented in the P&L. Management marked what it considers core profit, and investors can add back or deduct items from the base unit to build their own measures.

2.2 Q&A

Q: With slower openings, how are franchisees' cash receipts and profitability trending, and can store openings re-accelerate next year?

A: Franchisee cash flow is pressured by three items: coffee machine installments, Gen-6 remodels, and tighter store quality controls. Coffee machines bought last year were on installments, with higher outflows in year two; large-scale Gen-6 remodels hurt cash flow more for franchisees with more stores; from Nov last year, store quality standards were tightened, especially in Q1.

Profit pressure was evident in 1H. May was hit by weather and last year's delivery battles, which lifted delivery mix and dragged store profit; after Jun, delivery mix fell and store profits improved sharply rather than mildly, with Q1 already above last year and real recovery in Jun–Aug. Jul–Aug data is undisclosed, but profitability improved clearly.

Q: Can full-year per-store GMV land within guidance, and will Q3 beat expectations?

A: Guidance maintained at −5% to +5% for full-year per-store GMV. Per-store sales rose in Jul–Aug and met internal expectations, but specifics will not be disclosed given last year's high base in those months.

Q4 needs monitoring. Even after the summer peak last year, delivery mix remained heavy, so the company will not provide a separate Q4 outlook.

Q: Where does delivery mix stand now?

A: Delivery accounted for 60% of GMV in last year's Q3 and has fallen by about 10 ppt this year. No specific figures will be given, only that mix is much healthier than last year; note pricing differs between delivery and dine-in, so delivery mix by GMV vs. by cups will not match.

Q: What measures drove the dine-in recovery, how is store-level execution, and where could delivery mix settle medium to long term?

A: Focus on stable pricing, quality, concentrated marketing, and the Gen-6 plus coffee combo; no long-term delivery mix target was set. After delivery subsidies faded, dine-in did not fully backfill immediately and delivery did not drop quickly, so the key is to lift in-store sales and ensure franchisee cash receipts without changing the price system or resorting to blind promotions.

Product quality improved significantly this year, with coffee contributing meaningfully. New customer acquisition and repeat rates rose notably vs. last year, and most new customers were directed to dine-in; marketing budget is not much higher than last year, but spend is now concentrated rather than fragmented.

Overall GMV did not decline. Dine-in offset the delivery decline; management stressed there is no silver bullet, and without coffee and other combinations growth would be hard to sustain.

Q: With a notable GPM lift in 1H, how do you see 2H and next year, and any changes to 2–3 year opex and net margin guidance?

A: The company is not chasing higher GPM and aims to keep it within a range over the coming years; avoid linear extrapolation. GPM is assessed on a full-year basis rather than by half-year, consistent since listing.

Net margin is similar. There is operating leverage in certain costs, but management has no intent to keep pushing down opex ratios; absolute spend on talent upgrades and AI will rise, while opex as a share of revenue should be stable, leaving long-term leverage potential.

Effective tax rate should be stable with an upward bias over time due to the broader tax environment, and can be raised gradually in models. Other expense ratios should stay stable near term and potentially drift down slowly over the long term.

Q: Coffee still has promotional subsidies; once stabilized, how do you see per-cup profits for franchisees and the company?

A: Promotions were tied to campaigns over the past two weeks, and store coffee prices will be lifted thereafter. Prices were temporarily lowered to support concentrated marketing; after the coffee launch events, prices will normalize and GPM should improve materially.

On a pure materials basis, coffee bean GPM is lower than some tea drinks. However, coffee is simpler to make with lower store labor, and after factoring labor, store-level coffee profitability is not worse than milk tea; positioning remains high quality with strong value.

Q: Over the next 12 months, what is the prioritization of category innovation, and where are coffee, HPP juices, and breakfast baking?

A: Coffee mix has exceeded 20% with a target of 25%–30%, while HPP juice is a five-year second growth curve. The 20% goal set last year has been achieved, and there is room to further segment coffee.

HPP juice is a five-year story with little short-term analytical value or stock price linkage. It will not be 100% within the listed company at the outset, though the listed entity will own a relatively high stake; profits or losses will be consolidated before revenue in accounting, with further communication when scale or funding needs arise; breakfast baking is near the end of the exploration phase and will pair with coffee.

Q: Has coffee already reached 25%–30%, and where is the steady-state?

A: Currently stable above 20%, with 25%–30% targeted between 2H this year and 1H next year. During campaigns or promotions (such as the past two weeks), mix can exceed 25%; without campaigns it is around 20%, and has not yet stably stood at 25%–30%.

No terminal mix is pre-set. Management does not insist on 40% or 50%; it depends on consumer demand and competitive strength; 25%–30% is a mid-term target rather than the end state.

Q: What share of coffee revenue is flavored coffee, is its GPM better, and what R&D do rivals struggle to replicate?

A: No flavored coffee mix was disclosed, and management acknowledged there are no absolute moats that peers cannot replicate today. Relative advantages lie in product management, with raw materials, in-store processes, and equipment maintenance ensuring consistent quality; online reviews frequently praise solid beans; fruit tea has visible moats, but coffee is not yet at management's desired level.

The logic of flavored coffee resembles milk tea, swapping tea bases for coffee; core products like Americano are also improving in reputation. GPM does not necessarily rise with higher ticket prices, as higher-priced items carry higher costs; GPM differences are small across price tiers, though absolute gross profit per cup is higher; after labor, coffee store-level profitability is now roughly on par with milk tea.

Q: What lessons from Nanjing are transferable, and what still requires a city-by-city approach?

A: Provincial-capital expansion had been paused, and Nanjing is the first test of a restart, with no ready-made playbook. Management believed it lacked mastery in provincial capitals, so it reset the strategy and adopted new methods in Nanjing to tackle new problems; some tactics from other provincial capitals are reusable, but prior play relied on legacy approaches.

Whether to use more self-operated stores is also being explored in Nanjing, with no conclusion yet. Brand influence and teams have improved in recent years, but a handful of store successes prove little; whether success can be replicated in other provincial capitals remains unknown, and Nanjing should not be used for short-term forecasts but is more helpful for a 2–3 year view.

Q: Does Nanjing's strength raise expansion outlook for tier-1 and new tier-1 cities, and are larger offline stores incremental remodels or small-batch pilots?

A: No change to expansion outlook for higher-tier cities. Larger offline stores remain ongoing experiments, and management will update when there is scale or progress.

Q: How many stores now carry baking, and what share of daily sales does it contribute; within the RMB 1,500 breakfast sales goal, how do coffee and baking split?

A: Baking has launched in over 1,000 stores, with R&D testing concluded and implementation entering expansion. Basic work such as equipment selection and product exploration is done; given limited store coverage, contribution to total-store metrics will take time to show.

The RMB 1,500 breakfast sales is a forward-looking target. Management did not break down the split between coffee and baking, nor provide a long-term steady-state mix.

Q: There were no continued buybacks after Jul 9; at what price would buybacks resume?

A: No trading strategy will be discussed, and buybacks are only considered when the price is clearly below fair value; the CB-allocated buyback capacity is nearly used up. This does not preclude further open-market repurchases.

The CB serves three additional purposes: enhance liquidity, broaden investor reach, and, via delta hedging by CB investors, dampen stock volatility as they buy on declines and sell on rallies. Management emphasized the controlling shareholder will not sell easily and views prices below fair value as attractive opportunities.

Q: What is the opening cadence for the next two years; can net adds still be 3,000 per year?

A: The 2025 plan will be finalized internally in Oct–Nov, and long-term white space is at least 30,000 stores. Reaching 40,000 or more is possible, depending on market changes and category expansion, but any 2–3 year numbers set now would be no better than investors' own estimates.

Neither net adds nor gross openings are mandatory KPIs; the sole goal is long-term value. This year's adjustments show that what seems foreseeable at the start of the year can change by midyear, and any material changes will be communicated promptly.

Q: Are legacy market and store optimizations on track, and how is franchisee feedback?

A: Legacy store optimization is on plan in both scale and effectiveness, with consensus across the system. New-store adjustments are concentrated in legacy regions, with strong profit uplift and franchisee feedback; management will stay the course, and with fewer land grabs in the industry this year, it is a good window to push adjustments.

Closures were slightly higher in 1H, but the closure rate (closures divided by starting stores, or plus new openings) remains very manageable. A persistently ultra-low closure rate implies either insufficient opening aggressiveness or keeping underperforming stores afloat, which hurts both customer and franchisee sentiment; management prefers to close early.

Q: What is the strategy for northern markets, including provincial capitals such as Jinan?

A: Northern markets have proven viable, while provincial capitals remain a nationwide challenge and will be a key focus next year. From 2025 to 2026, the northern system will advance significantly; performance in newly opened Hebei and Shaanxi exceeded expectations, comparable to Guangdong and Guangxi at the start, due to changed opening strategies and higher standards for site selection and in-store requirements.

County-level cities such as Linyi and Jining are performing well, with profitability reaching or exceeding 55%. Provincial capitals are a different problem: by company standards, no other brand has achieved sustained success there; the approach is to keep store counts modest, drive higher units per store, capture city hotspots, serve delivery users well, and leverage brand power.

Q: By late May–Jun, how many stores had been adjusted, how much does this aid same-store growth, and what is the plan for 2H?

A: The number of adjusted stores was not disclosed; management estimates a 1–2 ppt lift to same-store, with greater contribution next year. This metric had not been previously communicated and may not be convenient to share; the initial plan set a 1:1 ratio between new openings and optimizations, which is tracking plan, with slightly fewer adjustments in 2H.

Adjustments are scheduled more in off-peak seasons to minimize disruption, while new openings follow planning windows. Adjusted stores account for over 10% of the base, but their adjusted operating time is a smaller fraction of the year, limiting same-store impact; this is more like an AB test, where flat comps in a down beta environment imply growth, albeit with a small internal sample.

Q: Are coffee customers primarily new or existing, and what drives success in tea-coffee convergence?

A: New customers account for over 50%, with repeat rates among existing customers up sharply; coffee success is primarily product-driven. Management clarifies coffee mix is currently 20%+, with 25%–30% as a future goal; non-product factors are limited.

Research shows young consumers drink one beverage a day, alternating milk tea and coffee. Some have 20+ drinks a month, including five to six coffees; long-term tea-coffee convergence is consensus, and the end game is taste and value.

Milk tea is complex to make and heavy on training and labor. Coffee is traditionally simpler, though flavored coffee is getting more complex; coffee brands pivoting to milk tea often start with simpler products, and if those gain traction the company will follow, but complex products come with inherent barriers.

Q: How are online and offline ticket sizes YoY, and how do you see 2H trends?

A: Both ticket size and store cash receipt rates rose vs. last year; no specific 2H trend was provided. Management noted 1H pressure was precisely because promotions were the most conservative in the industry, with lower promo share per store than peers, leading to YoY growth in realized receipts.

Q: Is there a more asset-light way to bring platform-driven customers back to the company's own ecosystem?

A: Retention depends on experience and satisfaction rather than low prices. After amortization, Gen-6 stores are actually the highest-ROI approach.

Low-price promos are only short-term tools; retention comes from offline experience and brand satisfaction. The goal is to lift monthly orders per customer from seven upward and raise dine-in mix; the second consideration is store profit, as online promo spend is hard to sustain, whereas dine-in investments translate into higher, longer-lasting store margins.

The membership system and mini-program promotions are well developed, with more room for user reach. Gen-6 remodels cost about RMB 50,000, or around RMB 100,000 for higher-spec finishes, and amortized over three years, monthly costs are far lower than delivery platform fees (ex-traffic spend); better-decorated community stores can also attract customers without building flagships.

Q: Within dine-in, how much is in-store pickup, and how does it differ operationally?

A: Platform pickup is about 3 ppt and counted under delivery, so true in-store pickup is higher. Whether customers sit and for how long is not tracked, as the company does not use facial recognition and does not view this as a key KPI.

There are two pickup measures: in-house mini-program pickup and platform pickup. The latter is currently counted as delivery sales, so actual in-store pickup is higher than disclosed.

Q: Capex rose significantly in 1H; where did it go and what is the outlook?

A: Nearly all the increase came from land payments for the Xiaoshan plot in Hangzhou, with recognition in right-of-use assets. This aligns with prior communication about the Xiaoshan land payment earlier this year.

Other capex remains maintenance in nature, such as warehouses and vehicles, with limited scale and no obvious increase. The HQ building on this plot will require investment over the next few years, consistent with prior guidance.

Q: Were more new stores opened by new or existing franchisees this year?

A: The split between new and existing franchisees is broadly unchanged vs. prior years. In most years, existing franchisees contribute a higher share, which is tied to regional mix; in years with more new-region openings, new-franchisee share rises, whereas last year saw more openings in legacy regions, favoring existing franchisees.

Long-term, balance is preferred. The company neither wants to rely solely on new franchisees nor shut out fresh blood; while the mix varies slightly year to year, it is broadly stable.

Q: How are spokesperson and co-branding choices made, and how do you balance hype with brand equity?

A: The biggest change is shifting from fragmented marketing budgets to several concentrated campaigns. Management admits it is still learning in marketing and brand investment, with clear progress this year vs. before.

Specific selection criteria for spokespersons are not suitable for disclosure on an earnings call. Like the Nanjing pilots, this is iterative and not a fixed long-term playbook, and will be adjusted dynamically.

<End of text>

Risk disclosure and statement:Dolphin Research Disclaimer and General Disclosure