---
name: egbe-os-saas-unit-economics
description: Calculate subscription customer value, acquisition costs and funnel requirements from grounded inputs.
metadata:
  egbe-when: "Offer when pricing or acquisition decisions need a viability check."
  version: "1.4.0"
  egbe-visible: "true"
  egbe-title: "SaaS unit economics"
  egbe-description: "Check subscription economics with measured data, explicit scenarios and traceable calculations."
  egbe-category: "Business operations"
  openclaw:
    requires:
      env: ["OS_API_KEY", "OS_URL"]
      bins: ["curl", "jq"]
    primaryEnv: "OS_API_KEY"
    emoji: "💰"
    os: ["linux", "darwin"]
---

<!-- egbe-catalog-updates:v1 -->
For work in a workspace with an installed, executable Egbe catalog, run
`python3 "<skills-directory>/egbe-os-quickstart/scripts/sync-catalog.py" --check` using the
actual installation path before starting work with this skill. `<skills-directory>` is the
project folder that contains this skill's directory, such as `.agents/skills`; find it with
file tools or `ls`, not inline interpreter code. Follow its result
with `<skills-directory>/egbe-os-quickstart/references/skill-updates.md`, then reread this
skill if updated. The check is read-only; the reference covers applying updates.
For public guidance without an installed catalog or command execution, skip
this local check and read the current published instructions directly. Reading
does not require a project or installation. For project execution that needs
connection or setup, use `GET /quickstart` on `OS_URL` to choose the appropriate
workspace or no-installation path before setup. The quickstart handles its own check.

Scheduled and other unattended runs have no one to approve inline code, so
`python -c`, `node -e`, heredoc scripts and code-execution tools may be blocked
there. Run bundled scripts as `python3 <script> ...` or `node <script> ...`, and
read files with file tools or plain shell commands.

<!-- egbe-connection-identity:v1 -->
Public instruction and playbook reads do not require credentials or connection
identity. The following applies to authenticated project requests, not reading
or applying public guidance without a connection.

For a BYOA workspace, retain its connection identity on **every authenticated
Egbe request**, including the examples and referenced workflows below. Load
`OS_URL` and `OS_API_KEY` from the project's Git-ignored `.env.local`. Set
`OS_AGENT_NAME` to your actual human-readable agent name, for example
`Claude Code` or `Codex`, not a slug such as `claude-code` or `CLAUDE_CODE`.
This is the name the owner sees in Dashboard. Never infer it from curl or reuse
another agent's identity. If your name is unknown, leave it unset. Quickstart saves the per-agent UUID in
`OS_CONNECTION_ID_<AGENT_SLUG>` (uppercase name with separators changed to
underscores, e.g. `CLAUDE_CODE`) and the optional owner-provided nickname in
`OS_WORKSPACE_NAME` in that file.

Many agent runtimes, including Claude Code, start a fresh shell for every
command, so shell functions and exported variables from an earlier command are
gone. Make each command that calls Egbe self-contained: load `OS_URL` and
`OS_API_KEY` in that same command, and run each authenticated curl example
through the project-local helper with your agent name set inline:

```bash
OS_AGENT_NAME="<your agent name>" python3 "<skills-directory>/egbe-os-quickstart/scripts/egbe-curl.py" --egbe-request /me/limits -s
```

`--egbe-request` loads the saved URL and API key from `.env.local` without
executing it, constructs the full URL, and adds the bearer token before calling
curl. Replace `/me/limits` with the API path from an example and keep its other
curl options; omit the example's URL and Authorization header. This mode rejects
additional destinations and redirects. Without this mode, the helper takes the
same arguments as `curl` and you must load credentials before shell expansion.
Do not edit global shell
configuration. Only in a shell that persists between your commands may you
instead export `OS_AGENT_NAME` once and define
`curl() { python3 "<skills-directory>/egbe-os-quickstart/scripts/egbe-curl.py" "$@"; }`.
Likewise prefix `OS_AGENT_NAME="<your agent name>"` on each command that runs a
bundled Python or JavaScript client.

The helper adds the saved identity headers only for the configured Egbe URL;
other destinations receive no extra headers. For an Egbe request it exits 22 on
an HTTP status of 400 or above and writes a status and request ID line to stderr
after every response; `<skills-directory>/egbe-os-quickstart/references/connection.md`
describes both. It rejects mixed destinations and redirects when attaching
identity. If the helper or this agent's UUID is missing, follow `GET /quickstart`
first. Before catalog installation, download and inspect
`GET /skills/quickstart/resources/scripts/egbe-curl.py` into the workspace's temporary
setup directory and use that path. After installation, switch to the installed helper.
For other Python, JavaScript or MCP clients, explicitly configure the same
`X-Egbe-Connection-Id`, `X-Egbe-Agent-Name`, and optional `X-Egbe-Workspace-Name`
headers on Egbe requests. A client that cannot attach identity headers remains
usable as an unidentified connection; do not claim its later activity is identified.
Deployed-product requests use their separate product credentials and no agent identity.


# Subscription unit economics

Calculate the economics needed for the current question using product evidence,
applicable benchmarks and explicit assumptions. Take the question and acceptable
risk from the caller's concrete task or plan. This skill supplies
calculations and validity checks, not a standing target for LTV/CAC, channel mix,
spend or conversion.

Reuse the product's existing `egbe-docs/product/economics.md` or linked test
calculation. Preserve measured values, sources and relevant prior assumptions;
do not overwrite them with a fresh benchmark forecast. Save only the calculations
needed for the task, not a compulsory full metric table or new master report.

## Establish the applicable inputs

Read the supplied brief, file, pricing page and existing records. Confirm which
product is meant only if ambiguous. Use an available browser when prices are
client-rendered; if unavailable, request the consequential missing price or leave
that scenario conditional. Do not infer current prices from search snippets.

These formulas apply to recurring monthly/annual subscription revenue. Identify
the actual billing, customer and cost units. A newsletter subscription and a
workflow SaaS can both recur but need different retention and cost evidence.
One-time purchases, transaction take-rates, ad revenue, and pure usage billing
need another model: explain the mismatch and do not force them into perpetual
monthly ARPU/churn. In a hybrid business, isolate the supported subscription
component and state what the calculation excludes.

Reuse observed cohort data when definitions, maturity and periods fit. Distinguish
observed values, externally observed benchmarks, product estimates and chosen
scenario assumptions. Do not invent a tier mix, churn adjustment, geography
discount, target ratio or conversion split to fill a table. Ask only for missing
inputs that would change the current decision; otherwise show a range, conditional
calculation or unavailable metric and explain what remains useful.

## Ground research and costs

Research only decision-relevant gaps. Before using Egbe search, read the live
`GET /search/skill`; the configured `OS_URL` and project `OS_API_KEY` need `tools`
scope. Credentials stay out of artifacts. This is an example query, not a
mandatory research stage or claim about the category:

```bash
OS_AGENT_NAME="<your agent name>" python3 "<skills-directory>/egbe-os-quickstart/scripts/egbe-curl.py" --egbe-request "/search/web/brave" -fsS \
  -H "Content-Type: application/json" \
  -d '{"query":"monthly customer churn <relevant subscription category> cohort benchmark <current-year>","count":5}' \
  | jq '.data.results[] | {title, url, snippet}'
```

Results use `data.results` and `snippet`; OS does not expose Brave's raw
`freshness` option. Verify publication dates and definitions on original pages.
An unavailable search source is a coverage gap, not permission to invent figures
or discard valid product data already available.

Compare category, buyer, product maturity, billing interval, acquisition channel,
geography, date, and measurement definition before applying a benchmark. Keep
reported ranges and sample limitations. A best-in-class statistic is not typical;
neither the low end nor midpoint of a published range is automatically a product
estimate. Triangulate consequential uncertain inputs where useful, without a
fixed source quota or averaging incompatible channels/cohorts. Explain any chosen
summary and adjustment. No automatic geographic or maturity multipliers apply.

Build product-specific COGS from applicable costs: hosting, storage/database,
third-party/AI usage, payment fees, support, and recurring content production,
localization or licensing where relevant. Use actual contracts/usage or labeled
estimates at a stated scale. Do not assume future volume discounts or optimized
support costs already exist. Separate fixed and variable costs and avoid counting
them twice; scenario comparisons can expose scale sensitivity. Include payment
fees at the actual billing frequency and applicable fee schedule, not a universal
processor percentage. Preserve links or other provenance for cost inputs.

## Calculations and validity

Use consistent currency, periods, population and gross/net revenue definitions.
Keep full precision internally and round for presentation. Rates below are
fractions, not percentage-point values.

| Quantity | Calculation | Conditions |
| --- | --- | --- |
| Monthly revenue per paying customer | recurring monthly revenue / paying customers | Same population and period; label a per-user denominator differently. Normalize annual revenue to monthly when comparing earned recurring revenue, not cash receipts. |
| Scenario ARPU from tiers | `sum(weight_i * monthly_price_i)` | Supported or explicitly proposed mix; nonnegative weights sum to one. Include discounts and billing mix where material. With no credible mix, show relevant tier/scenario values instead. |
| Gross margin | `(revenue - COGS) / revenue` | Positive revenue and compatible costs/periods. Negative margin is possible; do not clamp it or force a benchmark band. |
| Monthly customer churn | churned eligible customers / customers exposed to churn | Use a coherent cohort and exposure window; distinguish customer churn, revenue churn, refunds and net revenue retention. |
| Constant-hazard monthly equivalent | `1 - (1 - annual_customer_churn) ** (1/12)` | Only a conversion under constant monthly survival, not a measured monthly cohort rate or annual-contract cancellation cadence. |
| Approximate expected lifetime | `1 / monthly_customer_churn` months | Constant churn model with `0 < churn <= 1`; zero observed churn in a short/small cohort does not establish infinite lifetime. |
| Approximate gross-profit LTV | `ARPU * gross_margin / monthly_customer_churn` | Same steady-state model; excludes discounting, expansion and varying retention/cost. Label as a model, not guaranteed lifetime value. Use cohort/horizon contribution when this model does not fit. |
| Observed acquisition cost | attributable acquisition cost / acquired customers | State media-only vs fully loaded cost and comparable attribution/cohort window. No observed customers means undefined CAC, not zero or a fabricated finite number. |
| CPC | paid-media spend / measured ad clicks | Same source and window; zero clicks leaves CPC undefined. A click is not automatically a product visit. |
| Conditional required click-to-customer rate | `CPC / proposed_CAC` | Positive proposed CAC and consistent media-only cost basis. This is a requirement, not a forecast; greater than one is infeasible under the assumptions. |
| Conditional affordable CPC | `proposed_CAC * supported_click_to_customer_rate` | The conversion rate needs appropriate cohort evidence or an explicit scenario. It is not a provider bid recommendation. |

An affordable CAC is a plan decision, not LTV divided by a built-in constant.
If the plan chooses a positive LTV/CAC target `r`, its implied ceiling is `LTV/r`;
if it chooses a payback horizon, relate acquisition cost to contribution during
that horizon and state retention/cash-flow assumptions. Do not require either
target when the question can be answered without it. LTV-based sustainability
and a limited learning budget are different quantities.

Funnel stage probabilities can be multiplied only for linked conditional stages
of the same cohort: `P(signup | click) * P(paid | signup)` yields click-to-paid
when those stages cover the relevant path. Add arrival loss if starting at clicks
and the measured funnel starts at visits. Separate cohorts and unrelated period
totals do not establish these rates. A total conversion rate alone cannot identify
the stage split; leave it unidentified unless evidence or a labeled scenario
supplies an independent stage rate. There is no fixed signup/payment ratio.

Check finite values, positive denominators, rate bounds, weight sums and units.
An impossible required rate, nonpositive contribution or incompatible model is
an informative limitation, not a reason to hide the result or declare the whole
product unviable. Distinguish a mathematical error from an unsupported assumption.
Sensitivity cases should vary consequential assumptions coherently, not combine
every favorable cost with every unfavorable market statistic by rule.

## Return a traceable answer

Show the relevant observed/scenario metrics with units, scope and calculation
inputs. Record source URLs or file references, collection dates, original ranges,
definitions, chosen assumptions and material exclusions beside the calculation
or in a short source log. Do not label modeled LTV a proven minimum, a required
conversion rate an expected result, or provider CPC a guaranteed future cost.

Explain which assumption changes the answer and which evidence would improve it.
Return the supported calculation and implications to the caller; do not
change prices, budgets or campaigns as part of this calculation. Match the
founder's language and preserve unknowns instead of requiring every metric.