Payback Period Formula for AI Project DecisionsAI 项目回收期：简单、折现与不规则现金流完整计算

What is the payback period?什么是回收期？

Payback is a liquidity and exposure measure, not a complete value measure. It answers “how long is capital unrecovered?” but ignores cash flow after the crossing date. A shorter payback does not automatically mean higher net present value, stronger ROI, lower total cost, better strategic fit, or acceptable risk.

回收期衡量流动性与资金暴露时间，而不是完整价值。它回答“资本有多长时间尚未收回”，却忽略交叉日期之后的现金流。回收期更短并不自动表示净现值更高、ROI 更强、总成本更低、战略契合更好或风险可接受。

Choose the payback formula that matches the cash-flow pattern选择与现金流形态匹配的回收期公式

The familiar shortcut—initial investment divided by constant annual net cash inflow—works only when the post-investment net inflow is stable, begins immediately, and there are no material later investments. AI projects commonly violate all three conditions because implementation continues, adoption ramps, usage and review costs recur, and benefits arrive unevenly.

常见快捷公式“初始投资 ÷ 固定年度净现金流入”只有在净流入稳定、立即开始且后续没有重大追加投资时才成立。AI 项目通常同时违反这三个条件：实施仍在继续、采用率逐步爬坡、使用与审核成本持续发生，而且收益并不均匀。

Constant-flow simple payback =
  initial investment ÷ constant net cash inflow per period

Uneven-flow payback =
  full periods before cumulative net cash flow crosses zero
  + unrecovered balance before crossing
    ÷ net cash flow in the crossing period

Discounted net cash flow at period t =
  net cash flow at t ÷ (1 + annual discount rate)^(t ÷ periods per year)

Discounted payback =
  first time cumulative discounted net cash flow reaches zero

Net cash flow means cash benefits minus all incremental cash costs in the same period. Include subscription, consumption, support, human review, maintenance, retraining, and change when they continue after launch. Keep noncash capacity value separate unless there is an approved, evidenced route by which it changes cash or creates attributable value.

净现金流表示同一期间的现金收益减去全部增量现金成本。订阅、消耗、支持、人工审核、维护、再训练与变更若在上线后继续发生，就必须纳入。非现金产能价值应单独记录，除非存在经批准且有证据的路径，使其改变现金或创造可归因价值。

Distinguish payback period from ROI, NPV, IRR, and TCO区分回收期、ROI、NPV、IRR 与 TCO

Use the ROI example for complete return logic, the cost-benefit analysis formula for full-horizon value, and the total cost of ownership guide for the lifecycle cost boundary. Payback should sit beside these measures, not replace them.

完整回报逻辑可参考 ROI 算例，完整分析期价值可参考成本效益分析公式，生命周期成本边界可参考总拥有成本指南。回收期应与这些指标并列，而不能替代它们。

Define what “recovered” means before calculating a date计算日期前先定义“收回”的含义

A payback date is only as meaningful as its cash-flow boundary. Name the decision, baseline, proposed option, analysis horizon, currency, price basis, tax treatment, financing treatment, discount rate, period length, and timing convention. State whether the model uses project cash flow, budget cash flow, or a broader economic value flow; do not blend them.

回收日期的意义取决于现金流边界。应写明决策、基线、拟议方案、分析期、币种、价格口径、税务处理、融资处理、折现率、期间长度与时点约定，并说明使用项目现金流、预算现金流还是更广泛的经济价值流，不能混合使用。

Use incremental cash flow: proposed-option cash flow minus business-as-usual cash flow. If a current contract continues under both options, it is not an incremental benefit. If the project frees employee hours but payroll does not change, those hours are not automatically cash inflow. If released capacity supports measurable additional output, show that as a separately governed value case.

应使用增量现金流，即拟议方案现金流减去业务照常现金流。如果现有合同在两个方案下都会继续，它就不是增量收益；如果项目释放员工工时但薪酬没有变化，这些工时也不会自动成为现金流入。若释放产能支持了可测量的额外产出，应作为单独治理的价值路径展示。

Build a monthly AI project cash-flow schedule before calculating payback先建立 AI 项目月度现金流计划，再计算回收期

The hypothetical project deploys an AI-assisted data workflow. Month zero contains $240,000 of acquisition and implementation. Integration, migration, security, and training add $55,000, $35,000, and $15,000 in months one through three. Benefits begin in month four and ramp as eligible work, user adoption, and first-pass acceptance improve. Stable monthly gross cash benefit is $64,000; stable recurring cash cost is $18,000; stable net inflow is therefore $46,000.

假设项目部署一套 AI 辅助数据工作流。第 0 月包含 24 万美元采购与实施；第 1 至第 3 月的集成、迁移、安全与培训分别增加 5.5 万、3.5 万和 1.5 万美元。收益从第 4 月开始，随着合格工作量、用户采用率与一次通过率提升而爬坡。稳定期月度现金总收益为 6.4 万美元，持续现金成本为 1.8 万美元，因此稳定期净流入为 4.6 万美元。

The schedule makes three facts visible. Total transition outflow is $345,000, not only the $240,000 month-zero amount. Gross benefit must be reduced by recurring operation before accumulation. Payback occurs during month 14, not when the project first produces a positive monthly flow in month four.

该计划清楚显示三个事实：转型总流出为 34.5 万美元，而不只是第 0 月的 24 万美元；累计前必须用持续运营成本扣减现金总收益；回本发生在第 14 月内，而不是项目第 4 月首次产生正月度现金流时。

Calculate simple payback from cumulative monthly cash flow使用累计月度现金流计算简单回收期

At the end of month 13, the project still has $13,000 unrecovered. Month 14 contributes $46,000 of net cash inflow. If that inflow accrues approximately evenly during the month, divide the unrecovered balance by the crossing-month flow to estimate the fraction of month required.

在第 13 月末，项目仍有 13,000 美元尚未收回；第 14 月提供 46,000 美元净现金流入。如果该流入在月内近似均匀产生，可用尚未收回余额除以交叉月份现金流，估算所需月份比例。

Simple payback =
  13 full months
  + $13,000 unrecovered ÷ $46,000 month-14 net inflow

= 13 + 0.2826087
= 13.28 months

Approximate month-in-period =
  0.2826087 × 30.44 average days
  ≈ 8.6 days into month 14

The 13.28-month result is not valid if the $46,000 arrives as one discrete payment at month-end. In that case the investment remains unrecovered until the payment date, so the defensible answer is “payback occurs at the end of month 14.” Use interpolation only when the underlying cash benefit genuinely accrues through the period or when the estimate is explicitly described as an approximation.

如果 46,000 美元是在月末一次性到账，13.28 个月就不是有效的实际日期；此时投资会一直处于未收回状态，直到付款日，最可辩护的答案是“在第 14 月末回本”。只有当底层现金收益确实在期间内持续产生，或明确把结果描述为近似值时，才使用插值。

The constant-flow shortcut would also be misleading here. Dividing the $240,000 month-zero investment by the $46,000 stable monthly inflow produces 5.22 months, while dividing the full $345,000 transition outflow by $46,000 produces 7.50 months. Both omit the three-month implementation sequence and the five-month adoption ramp. The cumulative schedule correctly returns 13.28 months.

固定现金流快捷法在本例中同样会误导。用第 0 月 24 万美元除以稳定月度净流入 4.6 万美元得到 5.22 个月；用完整转型流出 34.5 万美元除以 4.6 万美元得到 7.50 个月。两者都忽略了三个月实施序列与五个月采用爬坡；累计计划才正确得到 13.28 个月。

Calculate discounted payback with a consistent monthly convention使用一致月度约定计算折现回收期

Discounted payback converts each monthly net cash flow to present value before accumulation. The example uses a hypothetical 10% effective annual rate. The equivalent effective monthly rate is approximately 0.7974%, calculated as (1.10)^1/12 − 1. Applying the annual rate with an exponent of month divided by 12 produces the same factors.

折现回收期先把每个月净现金流转换为现值，再进行累计。本例使用假设的 10% 有效年折现率；等效有效月利率约为 0.7974%，计算为 (1.10)^1/12 − 1。使用“月份 ÷ 12”作为年折现率指数会得到相同折现因子。

Discounted payback =
  13 full months
  + $36,356 unrecovered PV ÷ $41,159 month-14 PV flow

= 13 + 0.8833
= 13.88 months

The 0.60-month gap between simple and discounted payback is specific to this cash-flow shape and rate. It is not a universal adjustment. A higher rate, slower inflow, later benefits, or larger future costs will generally widen the gap. With a zero rate, discounted payback equals simple payback.

简单与折现回收期相差 0.60 个月，这是本现金流形态与折现率共同产生的结果，不是通用调整值。更高折现率、更慢流入、更晚收益或更大未来成本通常会扩大差距；当折现率为零时，折现回收期等于简单回收期。

Do not mix effective, nominal, real, and monthly rates不要混用有效、名义、实际与月度利率

A rate label must identify both economic meaning and compounding convention. “10%” is incomplete unless the reader knows whether it is an effective annual rate, a nominal annual rate compounded monthly, a real rate applied to constant-price cash flows, or a nominal rate applied to inflated cash flows.

利率标签必须同时说明经济含义与复利约定。仅写“10%”并不完整，读者必须知道它是有效年利率、按月复利的名义年利率、应用于不变价现金流的实际利率，还是应用于含通胀现金流的名义利率。

The HM Treasury Green Book explains that discounting and inflation adjustment are distinct and that real discount rates should be applied to real values. A company, government body, or investor may use a different approved rate; the control is internal consistency, not copying the illustrative rate on this page.

英国财政部《Green Book》说明折现与通胀调整是不同过程，实际折现率应应用于实际价值。企业、政府机构或投资者可能采用不同的批准利率；关键控制是内部一致，而不是照搬本文的示例利率。

Match the reported precision to how cash actually arrives让报告精度与现金实际到达方式匹配

Interpolation assumes the crossing-period net flow is spread evenly. That may be reasonable for continuously avoided labor or transaction cost measured across many events. It is not reasonable for a renewal avoided on one date, a milestone payment, a grant receipt, a year-end rebate, or a benefit dependent on one customer contract.

插值假定交叉期间的净现金流均匀分布。对于跨大量事件持续避免的人工或交易成本，这可能合理；但对于某日避免的续约费、里程碑付款、补助到账、年终返利或依赖单一客户合同的收益，就不合理。

NIST Handbook 135 notes that interpolation can produce a non-integer payback result but the underlying data often do not support that precision. The responsible report includes the period, method, timing assumption, and uncertainty—not only a decimal such as 13.28.

NIST Handbook 135 指出，插值可以产生非整数回收期，但底层数据通常不支持这种精度。负责任的报告应同时包含期间、方法、时点假设与不确定性，而不只是 13.28 这样的一个小数。

Use dated cash flows when invoices and benefits are irregular当发票与收益不规则时使用日期现金流

Monthly buckets are adequate when timing within a month is immaterial. For irregular dates, discount each cash flow using the elapsed fraction of a year from a common valuation date. Build a running cumulative present value in chronological order. The payback date is the first date on which that running total is nonnegative.

当月内时点影响不大时，月度分组足够。对于不规则日期，应从共同估值日开始，按经过的年份比例折现每笔现金流，再按时间顺序建立累计现值。累计总额首次非负的日期就是回收日期。

PV of dated cash flow i =
  cash flow i
  ÷ (1 + annual rate)^((date i − valuation date) ÷ day-count basis)

Running cumulative PV at date k =
  Σ PV of all cash flows dated on or before k

Excel's XNPV function documentation describes present value for a cash-flow schedule that is not necessarily periodic and uses corresponding payment dates. XNPV returns NPV for the selected series; it does not directly return a payback date. To find discounted payback, calculate running XNPV or equivalent cumulative PV after each dated cash flow and identify the first nonnegative result.

Excel 的 XNPV 函数文档说明了如何对不一定定期发生、且有对应付款日期的现金流计算现值。XNPV 返回所选现金流序列的净现值，并不会直接返回回收日期。要寻找折现回收期，应在每笔日期现金流后计算累计 XNPV 或等价累计现值，并识别首次非负结果。

Choose and disclose the day-count basis used by the model. The Microsoft function uses a 365-day year. If finance uses another approved convention, reproduce it consistently rather than combining spreadsheet defaults with a different contract or treasury basis.

模型必须选择并披露日计数基础。Microsoft 函数使用 365 天年度；如果财务采用其他批准约定，应一致复现，而不能把电子表格默认值与不同的合同或资金口径混用。

Use cash flow for cash payback and report capacity separately现金回收期只使用现金流，产能应单独报告

AI business cases often start with hours saved. Hours are an operational effect, not automatically cash. Payroll remains a cash outflow unless headcount, overtime, contractors, approved hiring, or another budget line changes. Counting every saved hour as cash inflow can pull payback forward without creating the liquidity that the metric is supposed to measure.

AI 商业案例常从节省工时开始。工时是运营影响，并不会自动成为现金。除非人员、加班、承包商、批准招聘或其他预算科目发生变化，否则薪酬仍是现金流出。把每个节省工时都计作现金流入，会在没有创造真实流动性的情况下人为提前回收期。

The same discipline applies to quality, speed, and risk. Monetize one evidenced causal endpoint, not every intermediate step. If fewer errors reduce rework labor and also reduce incident payments, those may be separate effects; if an avoided incident estimate already includes remediation labor, adding that labor again duplicates value.

质量、速度与风险也应遵循同样纪律。只货币化一个有证据的因果终点，而不是链条上的每个中间步骤。如果错误减少同时降低返工人工与事件付款，两者可能是独立影响；如果避免事件估算已经包含补救人工，再次加入这部分人工就会重复计算。

Stress-test the inputs that move the payback date压力测试会改变回收日期的输入

The base payback is not a promise. It is the consequence of transition cost, ramp timing, benefit realization, recurring cost, and the discount convention. Change one assumption at a time to identify sensitivity, then combine causally related variables into coherent scenarios. The 36-month NPV below uses the same 10% effective annual rate.

基础回收期不是承诺，而是转型成本、爬坡时点、收益兑现、持续成本与折现约定共同作用的结果。应先逐项改变假设以识别敏感性，再把有因果关系的变量组合成一致情景。下表 36 个月 NPV 使用相同的 10% 有效年折现率。

The simple payback does not change when only the discount rate changes; discounted payback and NPV do. The adoption-delay case has the latest discounted payback among these tests because positive flows move later. The 20% realization case produces the lowest NPV because it reduces positive flow throughout the remaining horizon. These are different decision signals.

只改变折现率时，简单回收期不会变化，而折现回收期与 NPV 会变化。在这些测试中，采用延迟使正现金流整体后移，因此折现回收期最晚；兑现降低情景在剩余分析期持续压低正现金流，因此 NPV 最低。它们是不同的决策信号。

Do not add every downside mechanically. A delayed rollout can also delay subscriptions, reduce early consumption, or change staffing. Build each combined scenario as a complete cash-flow schedule so linked effects move together and the same consequence is not counted twice.

不要机械叠加所有下行因素。上线延迟也可能推迟订阅、降低早期消耗或改变人员安排。每个组合情景都应建立完整现金流计划，使关联影响共同移动，并避免同一后果重复计算。

Turn an approved payback hurdle into measurable thresholds把批准的回收期门槛转化为可测量阈值

Suppose the organization requires payback by the end of month 18. The base model has $217,000 of cumulative undiscounted surplus and $166,209 of cumulative discounted surplus at that date. Those amounts are headroom, not guaranteed contingency. Translate them into thresholds for investment, recurring net inflow, or adoption timing.

假设组织要求在第 18 月末之前回本。基础模型在该日期有 217,000 美元累计未折现盈余和 166,209 美元累计折现盈余。这些金额是余量，不是有保证的应急准备金；应把它们转化为投资、持续净流入或采用时点阈值。

The minimum stable inflow is not $345,000 divided by 18. The project has uneven transition and ramp flows, so the threshold preserves those actual periods and solves only for the stable month-nine-through-month-18 inflow. On the discounted basis, the required nominal monthly inflow is higher because later amounts have lower present value.

最低稳定流入并不是用 345,000 美元除以 18。项目包含不均匀转型与爬坡现金流，因此阈值保留这些实际期间，只求解第 9 至第 18 月的稳定流入。按折现口径，所需名义月度流入更高，因为越晚发生的金额现值越低。

Assign each threshold to an owner and a source. Procurement monitors committed investment, product operations monitors eligible usage and acceptance, finance verifies realized cash effects, and the decision owner controls scope. A switching value without monitoring is merely a sensitivity result; with ownership, it becomes a control.

每个阈值都应指定负责人和数据来源。采购监控承诺投资，产品运营监控合格使用与验收，财务验证实际现金影响，决策负责人控制范围。没有监控的切换值只是敏感性结果；有了责任归属，它才成为控制。

Do not let a short payback hide weak full-horizon value不要让短回收期掩盖薄弱的完整周期价值

Payback stops counting when cumulative flow reaches zero. It assigns no value to later cash flow and does not show how long benefits last. Two options can have the same payback but radically different NPV, or a shorter-payback option can create less total value than a later-payback option.

回收期在累计现金流达到零时停止计数，不再为后续现金流赋值，也无法说明收益持续多久。两个方案可以拥有相同回收期却有完全不同的 NPV；回收更快的方案也可能比回收较晚的方案创造更少总价值。

Option A pays back one year sooner yet has slightly negative NPV at 10%; Option B pays back later but creates about $106,500 of NPV. Choosing A solely because of payback would sacrifice full-horizon value. The right decision depends on liquidity, risk, constraints, and value—not one metric.

方案 A 早一年回本，但按 10% 折现率计算的 NPV 略为负值；方案 B 回本较晚，却创造约 106,500 美元 NPV。仅因回收期选择 A 会牺牲完整周期价值。正确决策取决于流动性、风险、约束与价值，而不是一个指标。

NIST's Life Cycle Costing Manual for the Federal Energy Management Program explains that simple and discounted payback ignore costs, savings, and residual value after payback and should not be used to select among mutually exclusive alternatives. Use payback as a screen or exposure measure, then complete the full-horizon analysis.

NIST 的《联邦能源管理计划生命周期成本手册》说明，简单与折现回收期都会忽略回本后的成本、节省与残值，不应单独用于在相互排斥的方案之间做选择。应把回收期作为筛选或暴露指标，再完成完整周期分析。

Define a good payback period from the decision context根据决策背景定义合理回收期

There is no universal “good” payback period. A 12-month hurdle can be too strict for durable infrastructure and too loose for an experimental tool with a six-month contract, unstable demand, or high switching risk. Set the hurdle before reviewing the favored project's result to reduce outcome-driven bias.

不存在通用的“合理”回收期。12 个月门槛对耐用基础设施可能过于严格，对合同仅六个月、需求不稳定或切换风险较高的实验工具又可能过于宽松。应在看到偏好项目结果之前设定门槛，以减少结果驱动的偏差。

A hurdle is a governance rule, not evidence that every project just inside it is attractive or every project outside it is poor. Record who approved the hurdle, which method it applies to, the measurement horizon, and how exceptions are evaluated.

门槛是治理规则，并不能证明所有刚好达标的项目都具有吸引力，也不能证明所有未达标项目都很差。应记录谁批准门槛、门槛适用于哪种方法、测量分析期多长，以及例外如何评估。

Report “no payback within the horizon” when the curve never crosses累计曲线未穿越零点时，应报告“分析期内未回本”

If cumulative net cash flow remains negative through the approved horizon, the payback period is not a negative number and should not be manufactured by dividing investment by an average negative or speculative future inflow. Report the ending unrecovered balance and state that no payback occurs within the horizon.

如果累计净现金流在批准分析期内始终为负，回收期不是负数，也不能用投资除以负的平均流入或投机性未来流入来制造日期。应报告期末尚未收回余额，并说明分析期内没有回本。

If a longer useful life is credible, extend the whole model—benefits, costs, maintenance, risk, refresh, and exit—not only the positive cash flow. The horizon should be chosen from service life and decision need, not lengthened until the project happens to pass.

如果更长使用寿命可信，应延长整个模型，包括收益、成本、维护、风险、更新与退出，而不能只延长正现金流。分析期应根据服务寿命与决策需要确定，不能为了让项目刚好通过而不断延长。

Replace forecast payback with realized cash evidence after launch上线后使用实际现金证据更新预测回收期

At approval, payback is forecast. After launch, maintain two curves: the approved baseline and the latest actual-plus-forecast. Do not rewrite the baseline to make performance appear on plan. Reconcile actual invoices, commitments, payroll effects, usage, accepted output, and realized cash benefits each month.

审批时的回收期只是预测。上线后应保留两条曲线：批准基线与最新“实际加预测”。不能通过重写基线让绩效看似符合计划。每月应核对实际发票、承诺、薪酬影响、使用量、合格产出与实际现金收益。

Report forecast payback, current forecast-to-payback, and realized payback as different fields. Realized payback is observed only when cumulative verified cash flow crosses zero. If the model includes valued capacity or economic benefit, label that recovery measure separately so it cannot be mistaken for cash payback.

预测回收期、当前最新回收预测与实际回收期应作为不同字段报告。只有累计已验证现金流穿越零点时，才能观察到实际回收。如果模型包含估值产能或经济收益，应单独标记该回收指标，避免被误解为现金回收期。

Use the ROI Calculator to estimate gross labor potential—not payback使用 ROI 计算器估算人工总潜力，而不是直接计算回收期

The InfiniSynapse Data Analysis ROI Calculator estimates hours saved, gross annual labor-cost savings, and extra query capacity from team size, query volume, time per query, hourly cost, and an efficiency assumption. It does not include the InfiniSynapse license or a complete project cash-flow schedule. Treat its output as one directional gross-benefit input, not as net savings, ROI, or a payback date.

InfiniSynapse Data Analysis ROI Calculator 根据团队人数、查询量、每次查询时间、小时成本与效率假设，估算节省工时、年度人工总成本节省与额外查询产能。它不包含 InfiniSynapse 许可成本，也不包含完整项目现金流计划。因此应把结果视为一项方向性的总收益输入，而不是净节省、ROI 或回收日期。

Build a review-ready payback analysis in twelve steps用十二步建立可审阅的回收期分析

1. State the decision and baseline.陈述决策与基线。 Name the proposed option, business-as-usual alternative, owner, approval date, and what changes because of the investment.写明拟议方案、业务照常替代方案、负责人、审批日期，以及投资导致的变化。
2. Choose the cash-flow perspective.选择现金流视角。 Use project, budget, investor, or broader economic flow consistently; keep noncash outcomes separate.一致使用项目、预算、投资者或更广泛经济现金流，并把非现金结果分开。
3. Set the horizon and period.设定分析期与期间。 Choose months, quarters, or exact dates based on cash-flow timing, useful life, and decision precision.根据现金流时点、使用寿命与决策精度选择月、季度或具体日期。
4. Define the price and rate convention.定义价格与利率约定。 Record currency, real or nominal cash flow, effective or nominal rate, compounding, taxes, and financing treatment.记录币种、实际或名义现金流、有效或名义利率、复利、税务与融资处理。
5. Map every transition outflow.映射全部转型流出。 Place acquisition, integration, migration, internal labor, training, controls, disruption, and parallel operation when they occur.按发生时点放入采购、集成、迁移、内部人工、培训、控制、中断与并行运行。
6. Build the benefit mechanism.建立收益机制。 Connect eligible demand, adoption, acceptance, unit effect, utilization, and cash realization without double counting.连接合格需求、采用、验收、单位影响、利用与现金兑现，并避免重复计算。
7. Include recurring cost.纳入持续成本。 Subtract licenses, usage, cloud, data, review, support, governance, maintenance, and retraining in each relevant period.在相关期间扣除许可、使用、云、数据、审核、支持、治理、维护与再训练。
8. Calculate net and cumulative flow.计算净流量与累计流量。 Use benefit minus cost for each period, accumulate in chronological order, and locate the first nonnegative balance.每期使用收益减成本，按时间顺序累计，并定位首次非负余额。
9. Calculate discounted payback.计算折现回收期。 Discount each period consistently, accumulate present values, and locate the discounted crossing independently.一致折现每期现金流，累计现值，并独立定位折现交叉。
10. Control interpolation.控制插值。 Use a fractional period only when flow accrues sufficiently evenly; otherwise report the discrete payment date or a range.只有现金流足够均匀应计时才使用分数期间，否则报告离散付款日期或区间。
11. Test scenarios and switching values.测试情景与切换值。 Vary adoption, realization, transition cost, recurring cost, timing, and rate; calculate the threshold that violates the approved hurdle.改变采用、兑现、转型成本、持续成本、时点与利率，并计算违反批准门槛的阈值。
12. Report payback beside full-horizon value.把回收期与完整周期价值并列报告。 Show NPV, ROI or net benefit, TCO, noncash outcomes, risk, evidence quality, and realized-versus-forecast controls.同时展示 NPV、ROI 或净收益、TCO、非现金结果、风险、证据质量及实际与预测控制。

Use a spreadsheet layout that exposes every assumption使用能够暴露每项假设的电子表格结构

Keep inputs, calculations, results, and sources separate. One row per period makes timing visible and allows reconciliation to invoices, budgets, and operational drivers. Do not hard-code totals into result cells or hide gross benefit and recurring cost inside one manual net number.

输入、计算、结果与来源应分开。每期一行能让时点可见，并允许与发票、预算和运营驱动因素核对。不能把总额硬编码进结果单元格，也不能把总收益与持续成本隐藏在一个手工净额中。

Find the first row where column H is nonnegative for simple payback and the first row where column I is nonnegative for discounted payback. Use the prior row's unrecovered balance and the crossing row's flow only when interpolation is justified. Add checks that cumulative totals equal independent benefit-minus-cost totals and that dates are ordered.

简单回收期寻找 H 列首次非负行，折现回收期寻找 I 列首次非负行。只有插值合理时，才使用前一行尚未收回余额与交叉行流量。还应设置检查，确保累计总额等于独立的收益减成本总额，并确保日期按序排列。

Avoid ten payback-period errors that change the answer避免十种会改变结果的回收期错误

Frequently asked questions about payback period关于回收期的常见问题

Primary references for payback, discounting, and dated cash flow回收期、折现与日期现金流的主要参考

• NIST Handbook 135: Life Cycle Costing Manual for the Federal Energy Management Program — general simple and discounted payback formulas, interpolation caution, and limits of payback for alternative selection.NIST Handbook 135：《联邦能源管理计划生命周期成本手册》——简单与折现回收期通用公式、插值精度警示及使用回收期选择方案的限制。
• HM Treasury Green Book 2026 — full-lifetime appraisal, discounting, real versus nominal consistency, risk, sensitivity, and balanced value-for-money decisions.英国财政部《Green Book 2026》——完整生命周期评估、折现、实际与名义口径一致性、风险、敏感性与平衡的价值判断。
• Microsoft Support: XNPV function — present value for cash flows that are not necessarily periodic and use corresponding payment dates.Microsoft Support：XNPV 函数——使用对应付款日期，对不一定定期发生的现金流计算现值。

This page is an educational planning example, not financial, accounting, tax, investment, procurement, or legal advice. The figures and rates are hypothetical. Use current contracts, verified cash evidence, applicable rules, approved organizational methods, and qualified finance reviewers for a real decision.

本页是用于学习的规划示例，不构成财务、会计、税务、投资、采购或法律建议。所有数字与利率均为假设。真实决策应使用现行合同、已验证现金证据、适用规则、组织批准方法与合格财务审阅者。

Calculate simple and discounted payback from uneven cash flows用不均匀现金流计算简单与折现回收期

Enter an upfront investment and twelve monthly net cash flows. The calculator interpolates inside the crossing month and applies an effective monthly rate for discounted payback.

输入初始投资和 12 个月不均匀净现金流。计算器会在跨越回本点的月份内插值，并以有效月利率计算折现回收期。