FRT, Resolution Time: First response and resolution time

All metrics

How the chains work

Open a model to follow the numbers from one level to the next.

Money: from GMV to free cash flow

One ladder. Each line comes from the one above by subtracting a named cost. If a metric has no slot here, it is not a money metric.

GMVTurnover through the platform at buyer prices. Sellers' money, not yours.
less cancellations, refunds, fraud
Net GMVThe turnover that survived to fulfilment.
× take rate, or revenue recognised under your own model
RevenueCompany revenue. Gross for a reseller, net for an agent or marketplace.
less COGS: cost of goods, hosting, licences, delivery
Gross ProfitGross profit. Divided by revenue it becomes gross margin.
less other variable costs: payments, support, last mile
Contribution MarginWhat one order leaves behind before marketing.
less acquisition marketing
CM after CACCohort profit after paying for the cohort. This is where growth and bought growth separate.
less OPEX: engineering, brand, G&A
EBITDAOperating result before depreciation, interest and tax.
less depreciation, interest, tax
Net IncomeNet profit. The line a shareholder sees.
± working capital, less CAPEX
Free Cash FlowCash in the bank. Profit can be paper, cash cannot.

GMV

Gross Merchandise Value

GMV: lithography — All paid orders count toward turnover before deductions.

Σ price × quantity across paid orders

Turnover before deductions. A measure of the platform's scale, not its income.

→ rolls up into Net GMV

Show example

Orders in the month 1,200 Average order value $350 GMV = 1,200 × 350 = $420,000 Orders later refunded are still in GMV. The deduction happens one line below.

Net GMV

GMV after refunds and cancellations

GMV − cancellations − refunds − fraud

The turnover that survived to fulfilment. Take rate and revenue are computed on this base, not on gross GMV.

derived from GMV

Show example

GMV $420,000 Refunds 5% −$21,000 Net GMV = 420,000 − 21,000 = $399,000 Agree once which deductions count: cancellations before payment, refunds after it, chargebacks.

Take rate

Commission rate

Revenue ÷ Net GMV

The share of completed turnover kept by the platform. Compare it with GMV and seller retention; a higher rate alone does not prove a healthier business.

links Net GMV to Revenue

Show example

Paid orders (GMV) $420,000 Refunds −$21,000 Completed orders (Net GMV) $399,000 Platform revenue $39,900 Take rate = 39,900 ÷ 399,000 = 10% Use the same Net GMV base in the Revenue example below. Dividing by gross GMV would understate this rate.

Revenue

Top line

Net GMV × take rate, or Σ recognised revenue

The top line. In an agency model it is the commission, not the turnover.

→ rolls up into Gross Profit

Show example

GMV $420,000 Refunds 5% −$21,000 Net GMV $399,000 Revenue = 399,000 × 10% = $39,900 A reseller on the same data reports $399,000 of revenue and a huge COGS. The model moves the number tenfold.

COGS

Cost of Goods Sold

direct cost of delivering the product

In SaaS: hosting, CDN, licences, the infrastructure that keeps the product running.

subtracted from Revenue

Show example

SaaS example, monthly revenue $40,000 Hosting and CDN $6,200 Licences and APIs $3,100 Product support $2,700 Total COGS $12,000 Gross profit = 40,000 − 12,000 = $28,000

Gross Profit

Revenue after direct costs

Revenue − COGS

Money left after paying for delivering the product. Divided by revenue it becomes gross margin.

derived from Revenue and COGS

Show example

Same SaaS example: Revenue $40,000 COGS −$12,000 Gross profit = 40,000 − 12,000 = $28,000 Gross profit pays for marketing, salaries and everything else below this line.

Gross Margin

Gross margin

(Revenue − COGS) ÷ Revenue

The ceiling of the model. SaaS holds 70-85%, marketplaces lower, retail lower still.

derived from Gross Profit

Show example

Same SaaS example: Revenue $40,000 − COGS $12,000 = $28,000 gross profit Gross margin = 28,000 ÷ 40,000 = 70% Every new dollar of revenue brings 70 cents to cover everything else.

Contribution margin

CM1 / CM2 / CM3

CM1 = Revenue − COGS CM2 = CM1 − other variable costs CM3 = CM2 − marketing

The numbering is internal. Agree it once and write it down, or two reports will never match.

→ rolls up into EBITDA

Show example

Same SaaS example: Revenue $40,000 − COGS $12,000 CM1 $28,000 − payments and delivery $8,000 CM2 $20,000 − acquisition marketing $12,000 CM3 $8,000 Product support is already included in COGS in this example; do not subtract it again at CM2.

EBITDA

Earnings before interest, tax, depreciation, amortisation

CM3 − OPEX

A proxy for operating efficiency. Not cash flow.

→ rolls up into Net Income

Show example

CM3 $8,000 − engineering $4,500 − brand and G&A $2,000 EBITDA $1,500 EBITDA margin = 1,500 ÷ 40,000 = 3.75%.

FCF, Burn, Runway

Cash flow, burn, runway

Runway = cash on hand ÷ monthly net burn

Runway is measured in months. Below 12 it stops being a metric question and becomes a plan question.

the last line of the money ladder

Show example

Cash on hand $6,000,000 Net burn per month $500,000 Runway = 6,000,000 ÷ 500,000 = 12 months A 20% rise in burn costs two months of runway with no change in revenue at all.

Subscription: how MRR moves

In subscription, revenue is not a number but a balance of five flows. Read the composition of the movement, not the total.

Opening MRRThe base you start from.
plus New MRR and Expansion MRR: upgrades, seats, overage
less Contraction MRR and Churned MRR
Net New MRRNet growth for the period. Negative means marketing is pouring into a leaking bucket.
accumulated month over month
MRR → ARRARR = MRR × 12. A yearly projection, not a fact.

MRR

Monthly Recurring Revenue

MRR: lithography — Active subscriptions feed one normalised monthly intake.

Σ normalised monthly value of active subscriptions

Annual plans divide by 12. One-off payments never belong in MRR.

→ rolls up into Revenue

Show example

300 monthly × $120 = $36,000 50 annual × $1,200 ÷ 12 = $5,000 One-off setup fee $9,000 = 0 MRR = $41,000

ARR

Annual Recurring Revenue

ARR: lithography — The same current subscription level is projected across a year.

MRR × 12

The investor's language. Valuation usually starts as a multiple of ARR.

derived from MRR

Show example

MRR $41,000 × 12 = ARR $492,000 At a 6x multiple the valuation is ≈ $2,950,000 The 6× valuation multiple is an illustrative assumption, not a price implied by ARR alone.

NRR

Net Revenue Retention

NRR: lithography — One existing customer cohort grows, shrinks and loses members.

(start + expansion − contraction − churn) ÷ start

Measured on a fixed cohort, new customers excluded. Above 100% means growth without acquisition.

Expansion: extra recurring revenue from existing customers
Contraction: less recurring revenue from customers who stayed
Churn: recurring revenue lost when customers leave

the quality of MRR

Show example

Cohort a year ago $100,000 + expansion $18,000 − contraction $4,000 − churn $9,000 Total $105,000 NRR = 105,000 ÷ 100,000 = 105% NRR 105% with GRR 87% means expansion offsets losses. Check whether that expansion depends on a few large accounts.

GRR

Gross Revenue Retention

GRR: lithography — The ship carries only the retained part of its original cargo.

(start − contraction − churn) ÷ start

Capped at 100%. Shows the leak raw, with no upgrade makeup.

the pair to NRR

Show example

Start-of-year MRR for the same customer cohort: $100,000 Downgrades (contraction): −$4,000 Cancelled subscriptions (churn): −$9,000 GRR = (100,000 − 4,000 − 9,000) ÷ 100,000 = 87% NRR 105% with GRR 87% means the base leaks while large accounts pay more. Concentration risk.

SaaS Quick Ratio

Growth efficiency

(New + Expansion) ÷ (Contraction + Churned)

How much new MRR arrives per dollar lost. Below 1 the business is shrinking.

a read on Net New MRR

Show example

In one month, new customers add $15,000 MRR and existing customers add $6,000. Downgrades remove $2,000; cancellations remove $5,000. Quick Ratio = (15,000 + 6,000) ÷ (2,000 + 5,000) = 3.0 Every $1 of lost MRR was offset by $3 of new or expanded MRR in this month.

Bookings, Billings, Revenue

Three different numbers

Bookings = contracts signed Billings = invoices issued Revenue = revenue recognised

Mixing them in one report is the most common way to fool yourself a quarter ahead.

feed Revenue with a lag

Show example

A 12-month, $120,000 contract starts in January and is invoiced upfront. January bookings (contract value): $120,000 January billings (invoice): $120,000 January recognised revenue: $10,000 Deferred revenue after January: $110,000 This simple example assumes an even $10,000 of service is delivered each month.

Unit economics: a single customer

The whole business folded into one user. If it is negative here, scaling only speeds up the loss.

ARPU

Average Revenue Per User

Revenue ÷ active users

Counted over everyone, free users included. Falling ARPU with rising revenue means the base is being diluted.

compare with ARPPU to see paid conversion

Show example

Monthly revenue $40,000 Active users 60,000 ARPU = 40,000 ÷ 60,000 = $0.67

ARPPU

Average Revenue Per Paying User

Revenue ÷ paying users

The pair to ARPU. The gap between them is the paid conversion rate.

the pair to ARPU

Show example

Monthly revenue $40,000 from 1,600 payers: ARPPU = 40,000 ÷ 1,600 = $25 1,600 payers out of 60,000 actives = 2.67% paid conversion $25 × 2.67% ≈ $0.67 ARPU

AOV

Average Order Value

AOV: lithography — The value of one average order.

order value ÷ number of orders (GMV ÷ orders for a marketplace)

Average value of an order. Specify whether returns are included and whether the numerator is GMV or seller revenue.

→ can drive GMV

Show example

Marketplace: GMV $420,000 from 1,200 orders AOV = 420,000 ÷ 1,200 = $350 A later month: GMV $453,600 from 1,200 orders New AOV = 453,600 ÷ 1,200 = $378 The higher AOV is an observation. A shipping threshold may have contributed, but the example does not establish causality.

Frequency

Purchase frequency

orders ÷ buyers in the period

Orders per buyer in a defined period. Together with buyer count and AOV it explains GMV in a marketplace.

→ can drive GMV

Show example

In one month, 800 buyers place 1,200 orders. Frequency = 1,200 ÷ 800 = 1.5 orders per buyer GMV = 800 buyers × 1.5 orders × $350 AOV = $420,000

LTV

Lifetime Value

LTV: lithography — Each return adds margin to one customer relationship.

Quick estimate = monthly margin per customer ÷ monthly churn

Estimate lifetime value using margin, not revenue. The shortcut assumes stable monthly churn and an unlimited horizon; use cohort data or a fixed horizon for decisions.

→ rolls up into LTV/CAC

Show example

Monthly revenue per payer $25 Gross margin 70% Monthly margin = 25 × 70% $17.50 Monthly churn 5% Quick lifetime estimate = 17.50 ÷ 0.05 = $350 Expected margin in first 12 months ≈ $161 The $350 shortcut includes months beyond year one. The 12-month estimate sums each month's expected margin at 5% monthly churn.

CAC

Customer Acquisition Cost

CAC: lithography — Flyers and campaign work go into the cost of a new customer.

acquisition spend ÷ new customers

Only new customers below the line. Media and growth-team salaries both go above it.

→ rolls up into LTV/CAC

Show example

Media spend $60,000 Growth team $18,000 Creative $2,000 Total $80,000 New customers 400 CAC = 80,000 ÷ 400 = $200

Blended vs Paid CAC

Blended and paid

Blended = spend ÷ all new users Paid = paid spend ÷ paid-acquired users

Blended looks better and says less: organic hides what the channel actually costs.

a refinement of CAC

Show example

Separate scenario: 1,000 new customers arrive; 400 came from paid ads. Paid-channel spend is $80,000. Blended CAC = 80,000 ÷ 1,000 = $80 Paid CAC = 80,000 ÷ 400 = $200 Both values use paid-channel spend here. Allocate shared salaries and organic costs explicitly before comparing real channels.

LTV / CAC

Value to cost ratio

LTV ÷ CAC

Below 1 the business pays for the right to operate. Above 5 usually means underinvestment in growth.

the verdict of unit economics

Show example

Quick LTV estimate $350 ÷ CAC $200 = 1.75 If this team sets a target of 3.0, the gap is 1.25. A target is a planning choice, not a universal SaaS rule. With CAC unchanged, the shortcut reaches 3.0 at about 2.9% monthly churn.

CAC Payback

Payback period

CAC ÷ monthly margin per customer

Measured in months and hits runway directly. It matters more than LTV/CAC when cash is short.

→ drives Free Cash Flow

Show example

CAC $200 ÷ $17.50 margin per month = 11.4 months At 5% monthly churn, about 44% of the starting cohort has left by month 11.4. Read payback with retention, not alone.

Magic Number, Burn Multiple

Efficiency of the dollar invested

Magic = ΔARR ÷ prior-quarter sales spend Burn Multiple = net burn ÷ Net New ARR

Two views of growth efficiency: sales spend against new recurring revenue, and cash burn against net new ARR. Read both against your company's stage and cash position.

ΔARR: change in annual recurring revenue
S&M: sales and marketing spend
Net burn: cash spent minus cash received over the period

the joint between growth and money

Show example

Quarterly ΔARR $900,000 Prior-quarter S&M $1,200,000 Magic = 0.75 Annual net burn $2,000,000 Net New ARR $1,500,000 Burn Multiple = 1.33 The two ratios use different periods and denominators. Avoid treating a single threshold as a universal go/no-go rule.

The funnel: AARRR

Five stages, each with its own metrics. Always work the stage where the leak is biggest in people, not in percent.

AcquisitionImpressions, clicks, visits, installs. Metrics: CPM, CTR, CPC, CPI, CAC, channel share.
visit → signup conversion
ActivationThe user reached value. Metrics: signup rate, activation rate, TTV, onboarding completion.
activation → second visit
RetentionCame back, and came back again. Metrics: D1/D7/D30, cohort curves, churn, resurrection.
active → paying conversion
RevenuePaid. Metrics: conversion to paid, ARPU, AOV, expansion.
share who bring others
ReferralBrought someone new. Metrics: k-factor, viral cycle time, invite rate, NPS as a proxy.

CPM, CTR, CPC, CPA

Ad impressions, clicks and acquisition cost

CPM = spend ÷ impressions × 1,000 CTR = clicks ÷ impressions CPC = spend ÷ clicks CPA = spend ÷ chosen actions

These four numbers describe one paid-ad funnel. Define the action behind CPA (a signup here); CAC then uses paying customers.

CPM: cost per 1,000 ad impressions
CTR: click-through rate: share of impressions that became clicks
CPC: cost per click
CPA: cost per action: spend for one chosen conversion

→ rolls up into CAC

Show example

100,000 ad impressions cost $1,200. CPM = 1,200 ÷ 100,000 × 1,000 = $12 1,500 clicks → CTR = 1,500 ÷ 100,000 = 1.5% CPC = 1,200 ÷ 1,500 = $0.80 120 signups → CPA per signup = 1,200 ÷ 120 = $10 6 become paying customers → CAC = 1,200 ÷ 6 = $200 The same spend is divided by a different denominator at each step; name the action whenever you report CPA.

Activation Rate

Share who reached value

users who did the key action ÷ signups

The share of signups who complete a defined first-value action. A retention gap helps identify a candidate action; it does not prove the action caused retention.

→ rolls up into Retention

Show example

10,000 signups 4,000 reached the first export Activation = 40% D30 among activated 38% D30 among the rest 4% D30 means return on day 30. The 38% versus 4% gap makes first export a useful candidate for activation; check cohort mix before claiming causality.

TTV

Time to Value

TTV: lithography — The median wait until first value.

median time from signup to the key action

Median, not mean. The mean breaks on the tail of people who came back six months later.

a driver of Activation Rate

Show example

For users who completed their first export: Median time from signup 18 minutes Mean time from signup 9 days Users with TTV under 10 minutes: D30 retention 45% Users with TTV over 1 day: D30 retention 12% A few very late exports pull the mean far above the median. The retention comparison is observational.

Step conversion

Per step and end to end

Step conversion = users completing a step ÷ users entering it End-to-end conversion = final users ÷ starting users

End-to-end conversion hides where the break is. Keep both.

the structure of the funnel

Show example

100,000 visits → 10,000 signups 10% → 4,000 activations 40% → 600 payers 15% End to end = 0.6% People lost: 90,000 at the first step The first step has the lowest conversion (10%) and the largest absolute loss (90,000 visits). Diagnose traffic quality before changing the signup screen.

K-factor

Viral coefficient

invites per user × invite conversion

Above 1 growth sustains itself. In reality it is almost always below.

→ lowers CAC

Show example

Each acquired user sends 0.6 invites on average. 25% of invites turn into new users. K = 0.6 × 25% = 0.15 1,000 paid users bring about 150 invited users. At $200 CAC for the paid users, effective acquisition cost across all 1,150 users ≈ $174.

Channel Mix

Acquisition mix

share of new users, CAC and retention by channel

A channel is judged on three: volume, price, cohort quality. Two out of three is an incomplete decision.

a cut of Acquisition

Show example

One month's new customers, by source: SEO 40%, CAC $30, D30 retention 38% Paid social 35%, CAC $240, D30 retention 14% Referral 15%, CAC $12, D30 retention 41% Other 10% If paid-social LTV is $350, its LTV/CAC = 350 ÷ 240 = 1.46. Compare the same cohort window for CAC and retention. The LTV used here is an explicit assumption.

Engagement: how the product is used

The layer between activation and money. It answers whether a habit exists at all.

DAU / WAU / MAU

Daily, weekly, monthly actives

unique users with an activity event in the window

Write the definition of activity down. "Opened the app" and "completed an action" differ by multiples.

the base of every product metric

Show example

Two definitions for the same app: App opened today: DAU 12,000 Key action done today: DAU 4,300 Using 'app opened' consistently: MAU 60,000, WAU 38,000, DAU 12,000 Same product, two definitions, a 2.8x difference.

Stickiness

DAU ÷ MAU

DAU divided by MAU. With average daily actives, it approximates average active days per monthly user; it is not the median user's behaviour.

derived from DAU/MAU

Show example

Average DAU 12,000; MAU 60,000 over 30 days. Stickiness = 12,000 ÷ 60,000 = 20% Average active days per MAU ≈ 20% × 30 = 6 days The same average can hide very different user patterns. Read it with retention cohorts.

Sessions

Count, length, depth

sessions per user, median length, actions per session

A long session is not automatically good. A utility wants it shorter, content wants it longer.

a cut of DAU

Show example

Sessions per active per day 2.4 Median length 3 m 10 s Actions per session 7 After simplifying export: length 1 m 50 s, sessions 3.1 After the export change, sessions were shorter and more frequent. Check task completion before calling that an improvement.

Feature Adoption

Feature uptake

breadth = feature users ÷ actives depth = uses per feature user

Two numbers per feature. Narrow and deep is a niche tool, wide and shallow is decoration.

explains Retention

Show example

CSV export breadth 9,000 ÷ 60,000 = 15% depth 12 times a month D30 with the feature 46%, without it 19% The D30 gap is an association. Users who choose CSV export may already be more engaged.

Power User Curve

Distribution of active days

histogram of users by 1-30 active days in a month

A smile with a right-hand hump means a habit core. A left peak only means the product is one-off.

a cut of MAU

Show example

Active days in a 30-day month: 1-3 days 60% of users, 18% of actions 4-19 days 28% of users, 34% of actions 20-30 days 12% of users, 48% of actions 12% of users produce nearly half the activity. Their churn hurts more than anyone else's.

Core Action

Core action

completions per unit of time

The thing the user came for. It usually becomes the North Star.

→ rolls up into North Star

Show example

40,000 weekly active users 17,000 exported at least once (42.5%) Each exporter completed 5 exports on average Core action = 17,000 × 5 = 85,000 exports/week Per active user = 85,000 ÷ 40,000 ≈ 2.1

Retention and churn

Retention is a cohort curve, not a number. A flat plateau matters more than a high D1.

Retention D1 / D7 / D30

The classic cohort

users returning on day N ÷ cohort size

State the model: N-day, unbounded or rolling. The same data yields different numbers under each.

→ sets LTV

Show example

Cohort of 5,000 signups D1 2,100 = 42% D7 1,150 = 23% D30 800 = 16% D60 760 = 15.2% D90 750 = 15% The curve approaches 15% in this cohort. Check more cohorts before treating that level as a stable base for LTV.

Churn Rate

Churn

lost in the period ÷ base at the start

Separate behavioural churn (stopped using) from billing churn (cancelled).

the inverse of Retention

Show example

Base on the 1st 1,600 Cancelled in the month 80 Churn = 80 ÷ 1,600 = 5% Annualised ≈ 1 − (1 − 0.05)^12 = 46% 5% a month sounds mild and means losing nearly half the base in a year.

Logo vs Revenue Churn

Customer churn and revenue churn

Lost customer count and lost revenue can diverge.

Logo = lost customers ÷ customers Revenue = lost MRR ÷ MRR

The gap shows who is leaving: small accounts or large ones.

→ rolls up into GRR

Show example

Start of month: 1,000 customers and $100,000 MRR. 50 customers leave → logo churn = 50 ÷ 1,000 = 5%. Their lost MRR is $2,100 → revenue churn = 2.1%. This suggests smaller accounts are leaving. Another month: 15 leave (1.5%) but $6,000 MRR is lost (6%). This suggests at least one larger account left.

Customer Lifetime

Expected lifetime

1 ÷ churn rate

Only valid when churn is stable. On a young base it overstates lifetime by multiples.

a multiplier inside LTV

Show example

If 5% of subscribers leave each month, the simple estimate is 1 ÷ 0.05 = 20 months. At 3% monthly churn: about 33 months. At 8% monthly churn: 12.5 months. Two percentage points of churn move lifetime by 13 months.

Resurrection

Resurrected users

returning after a dormant spell ÷ dormant users

The fourth flow in the user balance: new, retained, churned, resurrected.

a flow into MAU

Show example

MAU last month 58,000 + new 6,500 + resurrected 1,900 − churned 6,400 MAU this month 60,000 MAU gained 2,000. Without resurrected users it would have gained only 100, nearly flat.

Cohort table

The retention triangle

rows are cohorts by entry date, columns are age

Reading down a column shows product effect, reading across a row shows cohort quality.

the instrument for Retention

Show example

Cohort M1 M2 M3 January 42% 24% 19% February 44% 27% 22% March 47% 31% 26% M3 is 19% for January and 26% for March. Investigate onboarding and channel mix before attributing the change to one cause.

Perception and support

Survey metrics do not replace behavioural ones. Their place is early diagnosis and segment cuts.

NPS

Net Promoter Score

NPS: lithography — Advocacy and criticism, with comments behind the score.

% promoters (9-10) − % detractors (0-6)

Range −100 to 100. The value sits in the comments and the cuts, not in the number.

a proxy for Referral

Show example

500 responses Promoters 275 = 55% Passives 125 = 25% Detractors 100 = 20% NPS = 55 − 20 = 35 Paying users NPS 48 Free users NPS 12

CSAT

Customer Satisfaction

satisfied responses ÷ all responses

Tied to one interaction: a ticket, a purchase, a release.

a cut of quality

Show example

430 ratings of 4-5 out of 500 = 86% Survey response 500 ÷ 1,800 tickets = 28% Only 28% of tickets produced a rating. Nonresponse can bias the result in either direction; compare response mix over time.

CES

Customer Effort Score

CES: lithography — An observer sees the effort needed for one task.

rating of "it was easy to get my task done"

A survey of how easy a specific task felt. Compare it with repeat use or churn for the same respondents; the score alone does not prove future behaviour.

a driver of Retention

Show example

Mean score 5.8 of 7 Share 6-7 64% Share 1-3 11% Churn among 1-3 responders 14%/mo Churn among 6-7 responders 2% In this sample, low-effort respondents churn less. That is an association, not proof that changing the survey score will reduce churn.

PMF Score

The Sean Ellis test

share answering "very disappointed" without the product

Share of respondents who would be very disappointed without the product. The 40% line is a heuristic; sample active users and inspect segments.

the gate before scaling

Show example

300 active users answered. Very disappointed 132 = 44% Somewhat disappointed 120 = 40% Not disappointed 48 = 16% Of 100 analysts, 61% were very disappointed. Of 100 students, 18% were very disappointed. The other 100 respondents account for 53 such answers. The analyst segment is stronger in this example. Validate its sample size and retention before deciding where to scale.

Contact Rate

Share of users contacting support

unique users contacting support ÷ active users

Shows how many people need help. Count tickets separately for support cost: one person may open several tickets.

can raise COGS through support cost

Show example

1,500 of 60,000 active users contacted support. Contact rate = 1,500 ÷ 60,000 = 2.5% They opened 1,800 tickets. At $4.20 per ticket, monthly support cost = $7,560. The top issue caused 38% of tickets: 684. Maximum avoidable cost if all 684 vanish ≈ $2,873. Actual savings will be lower if some contacts remain or support staffing does not change.

FRT, Resolution Time

First response and resolution time

median and p90 across tickets

Report percentiles. A support average always lies in the company's favour.

FRT: first response time: from ticket creation to first reply
p90: 90% of tickets are at or below this time

a driver of CSAT

Show example

FRT median 2 h 10 m FRT p90 11 h 40 m Resolution median 9 h Resolution p90 52 h CSAT when FRT under 1 h 94% CSAT when FRT over 8 h 61% The CSAT gap is a correlation: difficult tickets can both take longer and earn lower scores.

Marketplace and platform

A two-sided market is measured symmetrically: demand, supply and the quality of the match.

Liquidity

Supply liquidity

transactions ÷ listings posted

The share of supply that finds a buyer. The core health metric of a marketplace.

→ can influence GMV

Show example

3,400 deals ÷ 12,000 listings = 28% Core cities 41% Everywhere else 19% The 19% figure outside core cities is a prompt to investigate seller outcomes, not a universal failure threshold.

Fill Rate

Demand coverage

fulfilled requests ÷ all requests

The share of demand that got an answer. Liquidity seen from the buyer's side.

→ can influence Net GMV

Show example

8,600 of 10,000 requests = 86% Morning 94% Peak 6-8 pm 71% The peak-hour drop suggests a timing mismatch. Check available supply by hour before concluding why requests were missed.

Search-to-Fill

Search to transaction

transactions ÷ search sessions

The quality of search and matching, not the size of the catalogue.

→ rolls up into Conversion

Show example

Before: 3,400 deals ÷ 41,000 search sessions = 8.3% 11% of sessions showed no results. After synonym expansion, 3,936 deals from 41,000 sessions = 9.6%. The before/after comparison suggests an improvement; use an experiment to isolate the effect of synonyms.

Supply / Demand

Balance of sides

active sellers ÷ active buyers

Shows the tilt of the market and which side the budget belongs to.

→ sets Liquidity

Show example

4,200 active sellers and 26,000 active buyers Supply/demand ratio = 1 seller per 6.2 buyers The team chose a test corridor of 1:4 to 1:8. Furniture is at 1:14, so inspect unfilled requests there. The corridor is an example target, not a benchmark for every marketplace.

Repeat Rate

Repeat buyer share

buyers with 2+ orders ÷ all buyers

Whether demand returns without advertising. It lowers the load on CAC directly.

→ rolls up into Frequency

Show example

2,900 of 9,800 buyers placed at least two orders. Repeat rate = 2,900 ÷ 9,800 ≈ 30% These buyers generated 58% of GMV. Their repeat orders do not require a new-customer CAC, but retention and service still cost money.

Concentration

Turnover concentration

GMV share held by the top 10% of sellers

Dependency risk on a few players. Read it alongside their retention.

→ a risk to GMV

Show example

4,200 sellers generated $2 million GMV in a month. Top 10% = 420 sellers; together they generated 62% of GMV. The top 5 generated 19% = $380,000. One of those five generated $80,000. If that seller leaves, up to $80,000 monthly GMV is exposed. Exposure is not necessarily a permanent loss: some buyers may move to other sellers.

Time to First Transaction

Time to first deal

median from signup to first transaction

How fast a new participant reaches the market. The two-sided version of TTV.

→ rolls up into Activation

Show example

Buyers median 2 days Sellers median 6 days Sellers with a deal in week one: 90-day retention 64% Everyone else 17% The retention gap is observational. Early transactions may reflect better matching or stronger seller intent.

Leakage

Off-platform transactions

estimated share of deals done around the platform

How much take rate escapes the till. Measured by survey and indirect signals.

→ a leak out of Revenue

Show example

Survey: 600 sellers estimate that 14% of deal value moves off-platform. Estimated total deal value: $420,000/month. Off-platform value ≈ 420,000 × 14% = $58,800. At a 10% take rate, foregone revenue ≈ $5,880. Separately, 23% of chats contain a phone number. The survey share and chat signal are estimates, not a count of verified off-platform transactions.

Reliability and delivery speed

Technical metrics enter the product tree exactly where they move conversion and churn.

Uptime, SLO, Error Budget

Availability and error budget

successful requests ÷ all requests error budget = 1 − SLO

99.9% is 43 minutes of downtime a month. The error budget is what releases spend.

SLO: the team's target for successful service requests
Error budget: the share or time allowed to miss that target

→ hits Churn

Show example

SLO 99.9% → budget 43 m 12 s a month Incident 1 12 m Incident 2 9 m Spent 21 m = 49% of the budget 99.95% → 21 m 36 s 99.99% → 4 m 19 s The minutes assume a 30-day month and a service that is either fully up or down. Request-based SLOs may use a different budget.

Latency p50 / p95 / p99

Response time percentiles

percentile of the response-time distribution

Users live in the tail. Average latency describes nobody's experience.

p50: half of requests are this fast or faster
p95: 95% of requests are this fast or faster
p99: 99% of requests are this fast or faster

→ a driver of Conversion

Show example

p50 180 ms p95 1.4 s p99 3.2 s Mean 310 ms 1.2M requests a day p99 = 12,000 requests slower than 3.2 s At 1.2 million requests a day, roughly 12,000 requests sit at or beyond p99. Those are requests, not necessarily 12,000 distinct people.

Crash-free rate

Crash-free sessions and users

sessions without a crash ÷ all sessions

The share of app sessions without a crash. Also track users affected: a small group can experience repeated crashes even when the session rate looks high.

→ hits Retention

Show example

Crash-free sessions 99.2% Crash-free users 94.1% Of 60,000 users, 3,540 saw a crash Their D30 is 9%, everyone else 34% Crash-exposed users had lower D30 in this example. Investigate device and cohort mix before claiming the crashes caused the whole gap.

Core Web Vitals

Largest Contentful Paint, Interaction to Next Paint, Cumulative Layout Shift

LCP ≤ 2.5 s, INP ≤ 200 ms, CLS ≤ 0.1 at p75

Thresholds apply to the 75th percentile of real users, not to a lab run.

LCP: when the main visible content finishes loading
INP: how quickly the page responds to an interaction
CLS: how much visible content shifts unexpectedly

→ a driver of Acquisition and Conversion

Show example

Real-user 75th percentile: LCP 3.1 s → needs improvement (good: ≤2.5 s) INP 140 ms → good (≤200 ms) CLS 0.04 → good (≤0.1) After image compression, LCP was 2.2 s; signup conversion moved from 10% to 11.4%. The before/after change does not prove the faster LCP caused the conversion change.

DORA: speed

Deployment Frequency, Lead Time for Changes

releases per week, time from commit to production

The speed of feedback. Rare releases make every release risky.

Deployment frequency: how often changes reach production
Change lead time: time from committed code to production

→ sets Cycle Time

Show example

For one service over four weeks: 56 production deployments = 14 per week Median commit-to-production lead time = 6 hours 85th percentile lead time = 31 hours Compare this service with its own past performance and failure rate, not a universal deployment target.

DORA: failures and recovery

Change fail rate, failed deployment recovery time, deployment rework rate

Change fail rate = deployments needing immediate intervention ÷ all deployments Recovery time = time to restore after a failed deployment Rework rate = unplanned repair deployments ÷ all deployments

Count both rollbacks and hotfixes that follow a failed production change. Recovery time and unplanned repair work show the cost of failure.

→ guards Uptime

Show example

In one month: 56 deployments; 4 caused failures. Change fail rate = 4 ÷ 56 = 7.1% Median recovery after those failures = 38 minutes. 5 of 56 deployments were unplanned fixes. Deployment rework rate = 5 ÷ 56 = 8.9% DORA currently tracks five delivery metrics across throughput and instability; this card covers three of them.

Cycle Time

Task cycle time

from work started to shipped, median and p85

Breaks into stages. Fix the longest one, not the most visible one.

includes Lead Time as one stage

Show example

One representative task took 6.2 days: Development 1.4 days Code review 2.9 days QA 0.8 days Waiting to ship 1.1 days Total 6.2 days Review share = 2.9 ÷ 6.2 ≈ 47% The example shows where one task waited. Calculate the median and p85 from all completed tasks, not by adding stage medians.

Escaped Defects

Defects found by users

production bugs ÷ releases, and their share of all bugs

The guard metric for release speed. It rises first when tests are cut for deadlines.

→ guards Deployment Frequency

Show example

In one quarter, 240 bugs were found in total. 41 were first reported by users: 41 ÷ 240 ≈ 17%. The team shipped 60 releases. User-found bugs per release = 41 ÷ 60 ≈ 0.7. Previous quarter: 9% and 0.3 per release.

North Star and building your own tree

The tree is built top down and driven bottom up. A team moves the bottom layer only.

Business goalProfit, market share, valuation. Changes once a year.
one metric that captures delivered value
North Star MetricNights booked, minutes listened, reports sent. Not revenue.
decomposes into a product or a sum
DriversUsually 3-5: base width, frequency, depth, quality.
each driver breaks into something controllable
Team input metricsWhat a team moves in a sprint: step conversion, TTV, feature adoption, p95.

North Star criteria

Captures user value, not the fact of payment.
Leads revenue: it moves before the money does.
Controllable by the team within a quarter.
Decomposes into factors that do not overlap.
Resists gaming: you cannot raise it by making the product worse.

Show example

Weekly active users 40,000 Share who exported at least once 42.5% = 17,000 Exports per exporting user 5 North Star = 17,000 × 5 = 85,000 exports/week The three factors explain the total. Assigning owners is a team decision, not part of the arithmetic.

Levels and owners

L0 money: CEO and board, horizon one year.
L1 North Star: product leadership, horizon one quarter.
L2 drivers: area owners, horizon one month.
L3 input metrics: the team, horizon one sprint.
L4 counters and events: analytics, refreshed daily.

Show example

One illustrative metric stack: L0 EBITDA $1,500/month (business result) L1 85,000 exports/week (North Star) L2 40,000 weekly actives × 42.5% exporters × 5 exports L3 Activation 40%, median TTV 18 min, p95 1.4 s L4 Raw events export_clicked and export_done L2 multiplies into L1. L3 and L4 help teams diagnose and move the drivers; they are not another multiplier in the same equation.

Traps

Eleven ways to get the right number and the wrong conclusion.

Vanity metrics. Cumulative signups, downloads, followers. They only go up and change no decision.
Goodhart's law. A metric that becomes a target stops being a metric. Pair every target with a guard: speed plus quality, conversion plus refunds.
Means instead of percentiles. Load time, session length, order value: anything with a long tail is read at the median and p90.
A ratio with no denominator. Conversion rose to 8% because traffic halved. Always show both parts.
Survivorship bias. Retention measured on active users is flattering by construction.
Simpson's paradox. A metric rises in every segment and falls overall because the mix changed.
Mixed cohorts. A young base hides churn: the leavers have not had time to leave.
Last-click attribution. The closing channel eats the credit of the channel that introduced you.
Peeking in A/B tests. Stopping the moment significance appears inflates the effect. Fix sample size and duration before the start.
Local optimum. A string of small conversion wins will not save a product with no retention.
Definition drift. "Active user" changed quietly and the yearly chart became meaningless. Keep definitions in one place, each with the date it changed.

Nothing found. Try another word: churn, CAC, p95, take rate.