Picking the denominator · ZopDev University

Outcome

By the end of this lesson, you will be able to pick the right business denominator for your org’s unit economics, avoid the common bad-denominator traps, and explain why the choice of denominator shapes the entire cost narrative.


Tier	Engineer
JTBD	”Pick a denominator that turns the cloud bill into a business-meaningful efficiency number: not a vanity metric.”
Personas	FinOps Lead · Engineering Leader · Product Leader · Finance Partner
Prerequisites	T0, Foundations · M4.1, Maturity ladder
Time	9 minutes
Bloom verb	Pick (Evaluate), Avoid (Apply), Explain (Understand)

1. Concept

Unit economics means cost divided by something the business counts.

The thing you divide by is whatever your company is actually in the business of producing or serving: an order, a customer, a thousand requests.

It is the difference between “we spent $40K last month”, which nobody can do anything with, and “we spent $8 per active user, down from $9 last quarter”, which is a conversation.

CANDIDATE DENOMINATORS (by business model):
  MAU / DAU                        Engagement-driven products
  Orders / transactions             E-commerce, marketplaces
  API requests (per 1K)            API-first products
  Signups / activations            Growth-stage SaaS
  Sessions                          Content/media
  Tenants / paid seats              B2B SaaS
  GMV                              Marketplaces
  Inferences served                  ML products
  Bytes processed                    Data products

Picking criteria

The right denominator depends on what the business sells:

BUSINESS MODEL                                BEST DENOMINATOR
──────────────────────────────────────────────────────────────────
SaaS: monthly subscription                    MAU or paying users
SaaS: per-seat B2B                            active seats / tenants
SaaS: usage-based pricing                     primary usage unit
                                              (API calls, runs, etc.)
Consumer mobile / web                          DAU + MAU (both)
E-commerce                                     orders, GMV
Marketplace                                    GMV, completed transactions
API platform                                   API requests (per 1K typical)
Ad-supported                                   impressions, sessions
ML product (inference)                         inferences served
ML platform (training)                         model training jobs
Data products                                  bytes processed,
                                              datasets processed
Content / media                                streaming minutes,
                                              content downloads
Internal platform                              consumer team's tickets,
                                              builds, deploys

Trade-offs of each denominator

DENOMINATOR              DRIFT BEHAVIOR
──────────────────────────────────────────────────────────────────
MAU                       Grows organically; revenue tied; clear

DAU                       Volatile; depends on engagement; harder
                          to interpret trends

Orders                    Tied to revenue; cyclical (holiday peaks)

Signups                   Trial conversion affects this; can grow
                          without revenue

GMV                       Tied to revenue but volatile

Bytes / requests          Easy to measure; less business meaning
                          unless tied to revenue

Paid seats                Direct revenue tie; B2B's cleanest unit

What NOT to pick

Some denominators are tempting but bad:

AVOID                                          WHY
──────────────────────────────────────────────────────────────────
"Engagement" (vague)                          Subjective; not measurable
                                              consistently

Page views (unless tied to revenue)            Easy to measure but
                                              not causal to value

"Active users" without definition              "Active" varies; results
                                              not comparable

Lines of code, deploys, or eng-metrics        Measures activity, not
                                              business value

Anything not easily measurable                Bad denominator = bad
                                              metric

Multiple denominators averaged                 Composite metrics are
                                              uninterpretable

A common trap: picking “MAU” without defining what makes a user “active.” If the definition drifts (“active = logged in once” vs “active = used a feature”), the trend is meaningless. Lock the definition before publishing the metric.

The MAU example

PROFILE:    B2B SaaS, $40K/mo cloud, 5,000 MAU
  Cost-per-MAU: $40,000 / 5,000 = $8.00

TREND watching:
  $8.00 → $7.50 → $7.20 (positive: efficiency gains)
  $8.00 → $9.00 → $11.00 (negative: cost growing faster than user
    base; investigate)

The trend matters more than the absolute. Two orgs can have very different cost-per-MAU based on workload nature; the trend within an org is what reveals efficiency direction.

Denominator definition

Lock the definition in writing:

EXAMPLE: Monthly Active Users (MAU) at Acme:

  Definition:    A user who logged in AND took at least one
                  meaningful action (created, edited, or shared
                  content) within the 30-day window ending on
                  the report date
  Counted by:    distinct user_id
  Source:        analytics platform; daily aggregates
  Window:        rolling 30 days
  Last reviewed: 2026-04-15
  Owner:         data team

This definition will not change without org-wide notice. If we
decide to redefine, we will:
  - Document the change with the date
  - Recompute historical values under both definitions for
    overlap period
  - Communicate to stakeholders

Without the locked definition, six months in someone redefines MAU to include logged-out browsers; the number doubles overnight; cost-per-MAU halves; the improvement is illusory.

Multiple denominators

For complex orgs, multiple denominators serve different audiences:

PRIMARY (for leadership):  cost-per-MAU
SECONDARY (engineering):    cost-per-1K-API-request
TERTIARY (finance):         cost-per-ACV-dollar

Each tells a different story; together they paint the full picture.
Use one as the headline; others as supporting metrics.

The risk of multiple: they can give contradictory signals (cost-per-MAU improving, cost-per-API-request worsening). Pick one as primary; explain when others are referenced.

How ZopNight uses denominators

ZopNight’s unit economics overlay supports arbitrary denominators via the Unit Metrics configuration (M3.5.L5-L6). The customer defines the denominator (name, label, source), wires the data ingest (Push API / CSV / Pull API), and the overlay activates on cost trend charts.

Most customers configure 2-3 unit metrics; one becomes the headline.

2. Demo

A SaaS company’s denominator journey:

ORG:    B2B SaaS at Series B
CURRENT setup: just total cloud cost ($40K/mo) reported monthly

PROBLEM: leadership sees the bill grow with the business but
cannot tell if growth is healthy

DENOMINATOR EXPLORATION:
  Option 1: MAU (5,000) → $8/MAU
    Pros: clear; tied to business growth
    Cons: not all MAUs equal (some are heavy ML users)

  Option 2: paying users (1,200) → $33/paying user
    Pros: directly tied to revenue
    Cons: smaller number; ratio more volatile

  Option 3: 1K API requests (250K/mo total) → $160/M requests
    Pros: closely tracks infrastructure usage
    Cons: not all audiences understand API-request volume

DECISION:
  Primary: $33 per paying user (for leadership + finance)
  Secondary: $160 per million API requests (for engineering)
  Both tracked monthly; reported in the cost review

DEFINITION lock:
  paying user = active subscription at end of month, paid >$0
  API request = one HTTP call to *.api.acme.com, status 2xx
  Both definitions reviewed annually

OUTCOME (after 6 months):
  Cost grew 25% in absolute terms
  Paying users grew 35%
  Cost-per-paying-user dropped 7% (good story for leadership)
  Cost-per-M-API dropped 12% (good story for engineering)

Same cloud bill, two narratives, both positive.

3. Hands-on (5 min)

For your product, pick the denominator:

BUSINESS MODEL:    __________
PRIMARY REVENUE METRIC: __________

CANDIDATE DENOMINATORS (3 options):
  1. __________    Pros: __________ Cons: __________
  2. __________    Pros: __________ Cons: __________
  3. __________    Pros: __________ Cons: __________

CHOSEN PRIMARY:    __________
SECONDARY (optional): __________

DEFINITION (one sentence, specific):
  __________________________________________________________
  __________________________________________________________

WHO OWNS the definition:    __________
REVIEW CADENCE:              annually / quarterly

DATA SOURCE for the denominator:
  □ Existing analytics platform
  □ Data warehouse
  □ Manual export
  □ Other: __________

If you cannot write the definition in one specific sentence, the denominator isn’t ready. Lock it first.

4. Knowledge check

Q1

A SaaS company with monthly subscription pricing. Best denominator:

A. Bytes processed
B. Page views
C. MAU or paying users
D. Compute hours consumed

Show answer

Correct: C. Both work; pick based on what leadership values SaaS pricing is tied to the subscriber base; MAU reflects the engaged customer base, paying users reflects revenue directly. MAU or paying users for SaaS. Either reflects business value.

Q2

An e-commerce platform’s best denominator:

A. Page views
B. Sessions
C. Product catalogue size
D. Orders (or GMV)

Show answer

Correct: D. Revenue-generating events Page views measure activity, not value; an order is the unit of value. Orders or GMV for e-commerce. Tied to revenue.

Q3

Cost-per-MAU trending from $7 to $11. Interpretation:

A. Cost is growing faster than the user base
B. Healthy growth
C. Optimal
D. Evidence of some successful upselling

Show answer

Correct: A. Either an efficiency loss (waste accumulating) or pre-revenue growth (new features added before users monetize). Investigate to determine which; the metric alone doesn’t say. Unit-cost trend up = inefficiency or pre-revenue growth. The fix depends on the cause.

5. Apply

Define your denominator at Settings → Unit Metrics. Configure data ingest per M3.5.L6. The overlay activates on Reports → Cost Trend and Reports → Unit Economics.

For new orgs, start with one denominator. Add secondary metrics after the primary is stable.

Glossary terms touched

Unit economics · Denominator · MAU · Cost-per-X · Vanity metric

Outcome

1. Concept

Picking criteria

Trade-offs of each denominator

What NOT to pick

The MAU example

Denominator definition

Multiple denominators

How ZopNight uses denominators

2. Demo

3. Hands-on (5 min)

4. Knowledge check

Q1

Q2

Q3

5. Apply

Related lessons

Glossary terms touched

Start with the bill.