王峰 AI × 经营 约一次诊断

总目 / 文章 / WF-0405

文章 · 存档

餐饮店不外卖无销量,上平台就拖垮,餐饮真的还有出路吗?

餐饮外卖硝烟弥漫。平台补贴大战如火如荼,所有商家都在高销量与真金白银之间徘徊。

一位博主分享的经历:投资15万元开外卖店一年营业额224万元却亏损9万元,这恐怕是当下无数餐饮老板的镜像。为什么会这样?因为几乎所有商家都忽略了精打细算的核心:在整体财务框架下实时测算收入成本风险。只别以为财务只是记记账报报税,那都只能算是

有这样才能避开陷阱真正赚到钱。

案例剖析:高营业额背后的亏损陷阱

博主年营业额224万元,约4.4万单,每单平均51元,其实客单价不低了,而且号称毛利70%,也不低了,要知道,毛利再高的话,用户就不干了,因为用户不傻。

成本账扒开了算算。

- 原料成本:约68万元,占比30%。这是基础但已属高效。

- 人工成本:32.4万元,占比14%,6人团队平均月薪4500元。

- 房租与固定成本:仅4万元,占比1.8%,得益于非堂食小型卫星店模式。

- 水电气与杂费:14万元,占比6%。

上述这些加起来大概52%,其实这些比例已经算低到不正常了,通常能堂食餐饮店的重要成本就是“三二一”:也就是食材原料占比30-35%、人工占比20%-25%、房租占比10%-15%。这个35:25:15是天花板了,任何一个超了都很难赚到钱的。加起来大概60%-75%。水电气5%,其他包装杂费5%,再加上设备装修的折旧摊销费5%,75%-90%的成本费用占比,如果不加盟什么的,估计至少还有5%的总部管理费分摊,也就是基本上能赚到5-20%。

但是,但是,你发现这里没有外卖费的。

根据博主计算:

- 线上成本大头:平台抽成51万元(23%)、补贴52万元(每单12元红包+配送减免)、推广费11万元(5%)。这些合计114万元,占营业额51%,直接拖垮利润。

最终博主从156万元毛利,扣到-9万元净亏损。

其实你有没有想过,外卖平台为什么把收费定格在23-25%?你对比一下前面没有外卖的利润,是不是刚好就在20%左右,也就意味着,外卖平台最希望你成为他们的交付工具,把你的利润全部吃掉以后养肥它自己。

过了几天发现,餐饮店还能转起来,好嘞,再给你搞更多五花八门的名堂来收费,再加5%、再加5%、再加5%、再加。。。。

许多老板只看表面销量,越补贴越能冲销量,但补贴本身就是可怕的隐形杀手。补贴大战中平台推万亿补贴,商家被动卷入:每单补贴12元,看似拉动销量却让实际到手价从51元降至不足40元。推广费更是流量税,不投曝光就无单。

外卖渗透率超60%,但商家净利率仅6-15%。高销量不等于高利润。销量高却因补贴赔本赚吆喝。

这个账很好算的,不上外卖基本上没什么销量,上了外卖基本赚不到钱。到底是要销量还是要赚钱?这根本不应当成为一个问题,而是要考虑如何针对每一项成本支出的所能带来的投入产出比和投入回报率进行测算和过程监管,简单点理解,就是回到整体财务框架下测算怎么赚钱。尽管这对老板来说有点难,但我只能告诉你,流量的红利全部都被平台吃透了,不会给商家留半毛钱,你作为商家想要赚钱,必须建立财务模型实时测算每项决策的影响。否则,用不了两三年中国的餐饮将会整体性哀鸿遍野。而平台,也因为高额成本不得不更高收费而导致遭所有商家抵制。中国人是时候开始反内卷了。

精打细算的核心:构建整体财务框架

以前餐饮店只要好吃不太贵,基本上都赚的盆满钵满。于是全部都开始考虑开加盟,因为加盟费多好赚啊,自己不用干啥,将以前的红利再次翻红赚一笔。

可是绝大多数,绝对是绝大多数餐饮店只要一加盟就拉胯。不是品牌不好,而是初始餐饮店的红利被消耗掉了,初始餐饮店的赚钱大多数靠运气,运气还能传承吗?

所以,餐饮店赚钱不是靠运气而是靠数据,是真正的财务数据。整体财务框架包括收入预测、成本分解、现金流模拟和风险评估。还再用Excel简单加减啊,不行啊。以为有了ERP就能实时监管了啊,也不行啊。有数据不是重要的,能动态实时分析实时决策才是应当有的日常工作场景。

市场瞬息万变:原料价格波动、平台规则调整、门店人员异动等等。如果商家能够真的每天复盘,就能及早发现补贴支出过高而导致投入产出虽然好看但投入回报过低,如果在过程中就选择及时止损,尽管看起来影响销量了,但至少还不会赔钱。

如果你是商家,那我就建议你

- 至少每周重新审视几个关键指标,例如经营毛利率>35%(食材、房租、人工)、净利率>10%(含折旧摊销)。

- 至少每周重新审视平台活动的各项苛捐杂税到底都是什么名堂,别被销量蛊惑,每一条收费都要测算回报率。

- 当平台推出活动大力推广的时候,冷静,冷静,冷静。还是测算,别盲目跟风,划算就投,不划算就别投,或者象征性的投一点,把平台当做广告位。

如何操作?你只要每天都关注这五点,基本上就不会太拉胯:

- 用房租反算日营业额的最低额,租什么样的场地就应当卖出什么样的销量,否则就白白浪费了房租;

- 用日销量测算最多有多少人工费可以付;人工若涨薪,就需测算“多雇一人,多产出多少单?”人工出勤管理是非常重要的现场管控;

- 食材成本是供应体系好坏的重要体现,当然预制菜成本基本上都能控制在25%以下,而且工业化生产的核心就是标准化低成本高安全。其实不一定非要预制菜,只要供应链流转到门店之前尽可能的标准化就能够实现更低成本更高品质。

- 平台扣点的管控就是品类销量管控,指望从平台销售中赚到钱是挺有挑战的,所以必须要控制销售品类和销量,太便宜一定赔钱,太贵会摸黑自己。

- 水电气杂费每天关注永久性不超过10%,测算一下必要的水电气在正常营业状态下各项消耗的具体单位值,例如冰箱每天耗电多少,每一份餐食消耗多少煤气等等,将难以量化的总额成本具象化。

这些管控是最基本的,如果这些都是放羊状态,那基本上想要赚钱就是梦幻了。

补贴大战:参与还是后撤?测算决定一切

6年外卖平台补贴战将会继续升级:美团、饿了么、抖音本地生活互抢市场,推出“9.9元套餐”“0元配送”等活动。商家面临抉择:参与高销量不赚钱,还是后撤低销量赚钱?

不参与补贴大战或让销量至少降低30%,但单均利润升至10元以上。低销量赚钱的逻辑是精而专,聚焦高价值客户靠口碑复购。

如何决策?详细测算:

1. 模拟场景:用财务模型计算,参与补贴销量增50%但成本增40%;不参与补贴销量降20%成本降30%。

2. 关键指标:计算盈亏平衡点。销售收入=固定成本/(1-变动成本率),参与补贴的影响会让变动成本率大幅提升,销量就必须要更加的高。

3. 风险评估:补贴战是短期机会,用短跑的速度去持续长跑一定拖垮。测算补贴期后用户留存率若仅10%则不值。

4. 我的帮助:我不仅能帮你构建这些测算模型,更加可以帮你持续追踪跟进实时把控异常状况,在你面临要不要参加百亿补贴洪流的时候,我也能帮你测算加入与否,即便是加入也要有限度,确保你的利润能在预期范围内。

有一点请相信我,平台的测算能力一定远超所有商家,它一定会踩着你的盈亏平衡点来收你的各种名目费,让你食之无味弃之可惜。所以,你就要比平台大数据测算更加精明,否则就只能被它算计。

配送抉择:自建还是平台?划算需测算

配送是外卖成本的黑洞。博主用平台骑手减免4元/单,支出17万。但平台配送费5-10%,加高峰拥堵效率低。平台配送就是方便,但抽成高补贴易超支。

如果自建配送招聘骑手覆盖方圆5公里。优势:固定工资成本可控,人员也可以调配,控制时效提升满意度。缺点:初始投资高管理复杂。

哪个划算?详细测算:

1. 成本对比:平台:每单配送费8元(含减免)。自建:骑手月薪5000,覆盖日100单,每单成本5元。如果订单少则成本畸高。

2. 效率模拟:方圆5公里,自建平均送达20分钟,平台30分钟。测算时效提升复购率增15%。如果订单多则效率畸低。

3. 规模匹配:小店(日单<100)宜自建能省钱;大店(日单>200)平台更稳。

4. 综合框架:计算总成本。博主若自建省17万补贴,但增人工10万。净省7万转亏为盈。

但是,这里就需要做一个最为重要的决策了。你是清醒的认知你这个店上限多少,还是被动的被平台牵着鼻子走销量?

这是所有商家都不得不面对却绝大多数选择逃避的问题。

如果这一点你自己不决策而丢给平台去诱导你,你就会变成平台的交付工具疲于奔命。

让我助你精打细算赚到真金

餐饮店的战场从来不是销量竞赛,而是利润保卫战。

只有精打细算在整体财务框架下实时测算、随时纠偏,才能化险为夷。博主案例证明忽略测算高销量成空欢喜,补贴大战配送选择等机会若无数据支撑也会变成陷阱。

共享CFO的价值不是给你记记账、搞搞合规、筹划筹划税啥的,哪些你的财务都会干。

如果你希望经营公司赚利润不靠运气,而是把企业经营这件事真正跑顺、跑稳、跑久,共享CFO这套服务就是为你准备的。

Is There Still a Way Out for Restaurants?

Can't Sell Without Delivery, Platforms Can Drag You Down — Is There Still a Way Out for Restaurants?

Food delivery is like a battlefield now. Platforms fight with big subsidies, and every restaurant owner is stuck between “high sales” and “real cash in hand.”

One blogger shared this: they invested 150,000 RMB to open a delivery-only food shop, made 2.24 million RMB in sales in one year, but still lost 90,000 RMB. This is probably the mirror of many restaurant owners today. Why does this happen? Because almost everyone misses the core of careful math: under a full financial picture, you must calculate income, cost, and risk in real time. Don't think finance is just doing bookkeeping and filing taxes. That is only the smallest part.

Only then can you avoid traps and truly make money.

Case Breakdown: The Loss Trap Behind High Sales

The blogger's yearly sales were 2.24 million RMB, about 44,000 orders. Each order was about 51 RMB on average. That is not a low average price. They also said gross margin was 70%, which is also not low. You should know: if gross margin is even higher, customers won't accept it, because customers are not stupid.

Let's open the cost book and do the math.

Food ingredients cost: about 680,000 RMB, 30%. This is basic, and already pretty efficient.

Labor cost: 324,000 RMB, 14%. A 6-person team, average monthly pay 4,500 RMB.

Rent and fixed cost: only 40,000 RMB, 1.8%, because it is a small “satellite shop” with no dine-in.

Utilities and small extra costs: 140,000 RMB, 6%.

All of these add up to about 52%. Honestly, this is already low in an “almost not normal” way. Usually, the big costs of a dine-in restaurant are the “3-2-1 rule”: food ingredients 30%–35%, labor 20%–25%, rent 10%–15%. This 35:25:15 is already the ceiling. If any one goes over, it becomes very hard to make money. Together that is about 60%–75%. Utilities 5%, packaging and other small costs 5%, plus equipment and decoration depreciation and amortization 5%. That is 75%–90% in cost. If you are not a franchise, you may still have at least 5% “head office management fee” to share. So you can maybe make 5%–20% net profit.

But, but, do you see what is missing here? There is no delivery platform fee.

According to the blogger's math:

The big online costs: platform commission 510,000 RMB (23%), subsidies 520,000 RMB (12 RMB coupon per order + delivery discounts), promotion fee 110,000 RMB (5%). Together, that is 1.14 million RMB, which is 51% of sales, and it directly kills the profit.

In the end, the blogger went from 1.56 million RMB gross profit down to -90,000 RMB net loss.

Have you ever thought about why delivery platforms set the fee at 23%–25%? Compare it to the profit without delivery. Isn't that profit about 20%? That means the platform wants you to become their “delivery tool.” They eat your profit and make themselves fat.

After a few days, they see your shop can still run. Great. Then they add more and more new charges in many strange ways: plus 5%, plus 5%, plus 5%, plus….

Many bosses only look at the order count. The more subsidies, the more orders they can push. But subsidies are a scary hidden killer. In the subsidy war, platforms throw “huge subsidies,” and merchants get pulled in. Each order gets a 12 RMB subsidy. It looks like it helps sales, but it makes the real money you receive drop from 51 RMB to under 40 RMB. Promotion fees are also a “traffic tax.” If you don't pay, you don't get views, and you get no orders.

Delivery penetration is over 60%, but merchant net profit is only 6%–15%. High sales do not mean high profit. Big sales can still mean you lose money just to make noise.

This math is not hard. Without delivery, you basically have no sales. With delivery, you basically can't make money. So do you want sales, or do you want profit? This should not even be a question. The real point is: for every cost, you must calculate the input-output and return, and you must manage the process. Simply put: go back to the full financial model and calculate how to make money. This is hard for many bosses, but I can only tell you: the traffic bonus is fully eaten by the platform. They will not leave even one cent for you. If you want to make money, you must build a financial model and calculate the impact of every decision in real time. Otherwise, in two or three years, China's restaurant industry will be full of pain everywhere. And platforms, because of their own high costs, will charge even more, and then all merchants will fight back and reject them. It is time for Chinese people to stop this crazy “involution.”

The Core of Careful Math: Build a Full Financial Framework

Before, as long as your food was tasty and not too expensive, you could make a lot of money. Then everyone started thinking about franchising, because franchise fees are easy money. You don't need to do much, and you can “cash in again” on the old good times.

But most restaurants — yes, absolutely most — collapse once they franchise. It's not because the brand is bad. It's because the early easy profit is gone. The first shop often makes money mostly because of luck. Can luck be passed on?

So, making money in restaurants is not about luck, it is about data — real financial data. A full framework includes sales forecast, cost breakdown, cashflow simulation, and risk checks. Just doing simple plus and minus in Excel is not enough. Thinking ERP alone can manage everything in real time is also not enough. Having data is not the key. The key is being able to analyze in real time and decide in real time — that should be your daily work scene.

The market changes fast: ingredient prices move, platform rules change, staff changes, and more. If a merchant truly reviews every day, they can find early that subsidy spending is too high. It may look good on “input-output,” but the real return is too low. If you cut losses in time, it may hurt sales, but at least you won't lose money.

If you are a merchant, my advice is:

At least every week, review key numbers, like gross margin >35% (food, rent, labor), net margin >10% (including depreciation and amortization).

At least every week, review all the platform “extra taxes” and fees. Don't get fooled by sales. Every single fee must be tested for return.

When the platform pushes a big campaign hard, stay calm, calm, calm. Still calculate. Don't blindly follow. If it's worth it, do it. If not, don't. Or just do a tiny bit, and treat the platform like an ad spot.

How to do it? If you watch these five points every day, you usually won't fall apart:

Use rent to calculate the minimum daily sales you must hit. If you rent that kind of place, you must sell that level of orders, or you waste rent.

Use daily orders to calculate the maximum labor cost you can pay. If wages go up, calculate: “If I hire one more person, how many more orders can we produce?” Attendance and labor management is very important on site.

Ingredient cost shows how good your supply system is. Pre-made food cost can often be controlled under 25%, and the core of industrial production is standard, low cost, and safer. It does not mean you must use pre-made food. If you standardize as much as possible before items reach the shop, you can still get lower cost and better quality.

Platform commission control is really category-and-volume control. Making money from platform sales is hard, so you must control what you sell and how much you sell. Too cheap means you lose money. Too expensive means you hurt yourself too.

Watch utilities and small daily costs so they never go over 10%. Calculate unit use: how much power the fridge uses per day, how much gas one meal uses, and so on. Turn “hard to measure” total costs into clear, real numbers.

These controls are the minimum. If you let everything run wild, then making money is just a dream.

Subsidy War: Join or Pull Back? Calculation Decides Everything

In 2026, the delivery platform subsidy war will keep getting worse: Meituan, Ele.me, and Douyin Local Life will fight for market share, with deals like “9.9 RMB set meals” and “0 RMB delivery.” Merchants must choose: join for big sales but no profit, or pull back for lower sales but profit?

If you don't join, your sales may drop at least 30%, but profit per order can rise to over 10 RMB. The logic of “lower sales but profit” is to be focused and strong: serve high-value customers, build word-of-mouth, and get repeat orders.

How to decide? Calculate in detail:

Scenario simulation: use a financial model. If you join subsidies, sales rise 50% but costs rise 40%. If you don't join, sales drop 20% and costs drop 30%.

Key metric: calculate the break-even point. Sales revenue = fixed cost / (1 - variable cost rate). Joining subsidies pushes variable cost rate much higher, so sales must be even higher to survive.

Risk check: subsidies are a short-term chance. Running a long race with sprint speed will crush you. Calculate: if after subsidies end, customer retention is only 10%, then it is not worth it.

My help: I can not only help you build these models, but also help you track and control abnormal changes in real time. When you face the choice of joining a “huge subsidy flood,” I can help you calculate yes or no. Even if you join, it must be limited, so your profit stays within your expected range.

Believe me on this: the platform's calculation power is stronger than any merchant. They will step right on your break-even line and charge you many kinds of fees, so you can't live well, but you can't leave either. So you must be smarter than their big-data math, or you will only be used and trapped.

Delivery Choice: Build Your Own or Use the Platform? You Must Calculate

Delivery is a black hole of cost. The blogger used platform rider discounts of 4 RMB per order and spent 170,000 RMB. But platform delivery fees are 5%–10%, and rush-hour traffic makes it slow. Platform delivery is easy, but commission is high and subsidies can go out of control.

If you build your own delivery team and hire riders to cover a 5 km area: Pros: fixed wage costs are controllable, people can be scheduled, speed can be controlled, and satisfaction can rise. Cons: high start cost and complex management.

Which is better? Calculate in detail:

Cost compare: Platform: 8 RMB delivery cost per order (after discounts). Self-build: rider monthly pay 5,000 RMB, can cover 100 orders per day, cost per order 5 RMB. If orders are low, cost per order becomes crazy high.

Efficiency simulation: Within 5 km, self-build average delivery time 20 minutes, platform 30 minutes. Calculate: faster delivery raises repeat buy rate by 15%. If orders are too many, efficiency can drop badly.

Scale match: small shop (daily orders <100) should self-build to save money; big shop (daily orders >200) platform is more stable.

Full framework: calculate total cost. If the blogger self-built, they could save 170,000 RMB in subsidies, but add 100,000 RMB labor cost. Net saving 70,000 RMB, turning loss into profit.

But here comes the most important decision. Do you clearly know the real ceiling of your shop, or do you let the platform pull you by the nose with order volume?

Every merchant must face this, but most choose to avoid it.

If you don't decide and let the platform “guide” you, you become the platform's delivery tool, always busy and always tired.

Let Me Help You Do the Careful Math and Make Real Money

The real battle of restaurants is never a sales race. It is a profit defense war.

Only by doing careful math — under a full financial framework, calculating in real time, and correcting fast — can you turn danger into safety. The blogger's case proves it: if you ignore the math, high sales become empty happiness, and choices like subsidies and delivery can become traps without data support.

The value of a shared CFO is not to do bookkeeping, simple compliance, or some tax planning. Your finance team can already do that.

If you want your company to make profit not by luck, but by making the business run smooth, run steady, and run long, this shared CFO service is made for you.