起源故事 · 把财务报表压成四个数
Origin Story · Compressing the Financial Statement Down to Four Numbers
Goldratt 主张:与其让经理淹没在分摊成本、单位毛利的迷宫里,不如把决策压缩成三问 ——
「这个改善能让 T 上升吗?能让 I 下降吗?能让 OE 下降吗?」三个回答只要不全是「否」,方向就对了。
由这三个量直接拼出公司级指标:净利润 NP = T − OE,投资回报 ROI = NP / I。
这套口径的杀手锏是 —— 它天然反对局部优化,逼你永远从「整家公司多赚了多少」来评判每一个动作。
Goldratt argued that instead of drowning managers in allocated cost and unit margin, decisions should collapse into three questions:
"Does this change raise T? Does it lower I? Does it lower OE?" As long as the three answers aren't all "no", the direction is right.
From those three quantities the company-level metrics fall straight out: Net Profit NP = T − OE and ROI = NP / I.
The killer feature of this language is that it structurally rejects local optimization — it forces every action to be judged on how much more the whole company made.
1 四块表,T 一动全跟着动 Four Gauges — Move T, They All Move
健康healthy2 四条公式:T 是共用的分子 Four Formulas, One Shared Numerator: T
NP、ROI、生产力、周转 —— 四条公式的分子里都站着 T。砍 OE 只帮到 NP 和生产力,降 I 只帮到 ROI 和周转,唯有增 T 同时改善全部四个。这就是「改善决策三问」里把增 T 排第一的数学根据。 NP, ROI, Productivity, Turns — every formula has T in the numerator. Cutting OE only lifts NP and Productivity; lowering I only lifts ROI and Turns. Only growing T improves all four at once. That's the mathematical reason "grow T" sits first in the three-question decision filter.
3 现实里的 NP 与 ROI 决策 NP & ROI Decisions in the Real World
改善决策三问:评估任何改进项,先问它对 T、I、OE 各有什么影响,再看 NP 和 ROI 净变化,而不是只看局部省了多少。
The three-question filter: for any proposed improvement, first ask how it moves T, I, and OE, then look at the net change in NP and ROI — not how much was "saved" locally.
投资优先级:两个项目都要花 100 万,选 ROI 高的那个 —— 即同样投入下 NP 增量更大、I 占用更省的方案。
Ranking capital projects: when two projects both cost ¥1m, pick the higher ROI — the one that delivers more NP and ties up less I per yuan invested.
周转 = T / I:周转越高说明同样的钱跑出更多产出。压低库存 I,周转和 ROI 一起涨,现金流也松快。
Inventory Turns = T / I: higher turns mean the same capital is generating more output. Drive I down and turns and ROI rise together — and the cash flow gets a lot easier to breathe.
生产力 = T / OE:每花 1 元运营费换回多少有效产出。盲目降本若伤了 T,生产力反降 —— 表会告诉你真相。
Productivity = T / OE: how much Throughput each yuan of OE buys you. Blind cost cutting that nicks T actually drops productivity — the gauge tells the truth.
一句话In One Line
四个指标不是四件事,而是同一家公司从四个角度照出来的像。
NP 看绝对赚多少,ROI 看相对划不划算,生产力看运营效率,周转看资金跑得快不快。
它们的公式藏着一个共同信号:T 站在每一个分子上。
所以当资源有限、只能押一个方向时,TOC 给的答案永远清晰 —— 先增 T,因为它是唯一能让四块表同时变好的杠杆。
降 OE、压 I 是配菜,增 T 才是主菜。
The four indicators aren't four different things — they're four angles on the same company.
NP shows the absolute money kept, ROI shows whether the capital earned its keep, Productivity shows operational efficiency, Turns shows how fast the cash is cycling.
The formulas hide a single shared signal: T stands in every numerator.
So when resources are scarce and only one direction can be pushed, TOC's answer never wobbles — grow T first, because it's the only lever that moves all four gauges in the right direction at the same time.
Cutting OE and squeezing I are the side dishes. Growing T is the main course.
常见误用Common Mistakes
只盯一个指标做决策。NP、ROI、生产力、周转要一起看,单看 NP 高但 ROI 低,可能只是占用了太多资金。
Deciding on a single metric. Read NP, ROI, Productivity, and Turns together. A high NP with a low ROI usually just means too much capital is parked behind it.
为了 ROI 好看而拼命压 I,砍到伤了 T。I 是分母,但 T 在分子,压 I 不能以损 T 为代价。
Slashing I to flatter ROI until it bleeds T. I is the denominator, but T is the numerator — never trade T away to shrink I.
把会计净利当成现金。多堆库存能美化报表利润,却锁死现金、压低 ROI 与周转,TOC 视角看得很清楚。
Confusing accounting profit with cash. Piling up inventory dresses up book profit while freezing cash and dragging ROI and Turns down — the TOC lens sees right through it.