Marketing metrics become dangerous when teams use the same word to mean different things.
“CAC” may mean ad spend divided by customers in one dashboard and fully loaded sales plus marketing cost divided by customers in another.
“LTV” may mean revenue in one model and gross profit in another.
“ROAS” may be reported as 4.0x in one meeting and 400% in another.
None of those choices is automatically wrong.
The problem is ambiguity.
The Marketing Metrics & Formulas Dictionary included with this article is designed to create a shared operating vocabulary for acquisition, funnel, lifecycle and unit-economics metrics.
It includes formulas, units, use cases, common traps and source references.
CAC: Customer Acquisition Cost
HubSpot defines CAC using sales and marketing costs divided by new customers.
A practical formula is:
`CAC = (Sales Costs + Marketing Costs) / New Customers`
The most common mistake is undercounting costs.
Teams may include:
- ad spend;
and exclude:
- sales salaries;
- marketing salaries;
- contractors;
- CRM;
- agencies;
- tooling;
- events.
The correct scope depends on the management question.
The important rule is consistency.
If the organization uses “fully loaded CAC,” document what is included.
If it uses “paid CAC” for channel analysis, label it differently.
Do not compare the two as if they were the same metric.
ROAS: Return on Ad Spend
Google Ads defines ROAS as conversion value divided by spend.
Formula:
`ROAS = Conversion Value / Ad Spend`
If a campaign generates $100,000 of attributed conversion value from $25,000 of spend:
`ROAS = 4.0x`
or:
`400%`
ROAS is useful for media efficiency.
It is not profit.
A campaign can have high ROAS and weak profit if:
- gross margin is low;
- discounts are large;
- fulfillment cost is high;
- returns are high.
ROAS also depends on attribution.
That leads to another metric.
Incremental ROAS
Google Ads describes incremental ROAS in Conversion Lift as incremental conversion value divided by total ad spend.
Formula:
`Incremental ROAS = Incremental Conversion Value / Ad Spend`
The difference is causal intent.
Standard ROAS asks:
How much attributed conversion value is associated with the spend?
Incremental ROAS asks:
How much additional value did the advertising produce relative to a control design?
The second question is harder.
It requires:
- experiments;
- holdouts;
- geography tests;
- credible lift measurement.
Use standard ROAS for recurring media management.
Use incrementality evidence for larger allocation questions when practical.
ROI: Return on Investment
ROI should include profit.
A generic formula:
`ROI = Net Profit / Total Cost`
Google Ads provides an example where ROI is calculated using revenue minus total costs relative to total cost.
ROI is broader than ROAS.
ROAS: media-specific revenue relationship.
ROI: economic return after costs.
Do not use the words interchangeably.
LTV / CLV
Shopify's 2026 CLV guide uses:
`CLV = Average Order Value × Purchase Frequency × Customer Lifespan`
Example:
AOV: $50
Purchases per year: 3
Customer lifespan: 2 years
CLV: $300
This is a revenue-based estimate.
For acquisition decisions, a gross-margin-adjusted LTV may be more useful.
Example:
Revenue LTV: $1,000
Gross margin: 70%
Gross-margin LTV: $700
That distinction matters because acquisition is paid with cash, not revenue.
LTV:CAC
Formula:
`LTV:CAC = LTV / CAC`
Stripe discusses a 3:1 ratio as a common SaaS benchmark, but benchmarks should be used carefully.
A 3:1 ratio can mean very different things depending on:
- payback;
- margin;
- churn;
- capital availability;
- growth stage.
A business with 5:1 LTV:CAC and a 30-month payback can still be cash constrained.
That is why the next metric matters.
CAC Payback
Stripe defines CAC payback as the time required to recover customer acquisition cost.
Formula:
`CAC Payback = CAC / Monthly Profit per Customer`
If:
CAC: $600
Monthly gross profit: $100
Payback: 6 months
Payback is a cash-efficiency metric.
Two companies with the same LTV:CAC can have very different payback.
Shorter payback lets a business recycle capital faster.
Conversion Rate
Generic formula:
`Conversion Rate = Conversions / Eligible Users`
The denominator matters.
Examples:
- purchases / sessions;
- leads / landing-page visits;
- activated users / signups;
- opportunities / qualified leads.
A metric is only interpretable when numerator and denominator are defined.
Do not say “conversion rate increased” without naming the stage.
Activation Rate
Activation measures the share of new users who reach first meaningful value.
Formula:
`Activation Rate = Activated New Users / New Users`
The hard part is defining “activated.”
Bad definition: account created.
Better: customer completed the behavior correlated with receiving product value.
Activation is product-specific.
Retention Rate
One simple cohort formula:
`Retention Rate = Retained Customers / Starting Cohort`
Retention should usually be cohort-based.
Aggregate active-user counts can hide churn because new users replace old ones.
Specify:
- cohort;
- time window;
- activity definition.
Example:
30-day retained account rate.
Churn Rate
Formula:
`Churn Rate = Customers Lost / Starting Customers`
Revenue churn and logo churn are different.
Logo churn: customer count lost.
Revenue churn: recurring revenue lost.
For SaaS, expansion can offset contraction, which leads to NDR.
Net Dollar Retention
A simplified formulation:
`NDR = Ending Recurring Revenue from Existing Cohort / Starting Recurring Revenue from Existing Cohort`
NDR includes:
- expansion;
- contraction;
- churn.
It excludes new-customer revenue from the measured cohort.
This makes it a useful indicator of how the existing customer base evolves.
AOV
Average order value:
`AOV = Revenue / Orders`
AOV is useful for:
- merchandising;
- bundling;
- CLV analysis.
AOV can increase while profit falls if discounting or product mix changes.
Pair it with margin.
Purchase frequency
Shopify defines purchase frequency as orders divided by unique customers within a period.
Formula:
`Purchase Frequency = Orders / Unique Customers`
This is a major component of ecommerce CLV.
An LTV initiative can improve:
- AOV;
- frequency;
- lifespan.
Each lever behaves differently.
CPL
Cost per lead:
`CPL = Lead Generation Cost / Leads`
CPL is useful for B2B and lead-gen campaigns.
But a low CPL can be dangerous if lead quality is weak.
Always pair it with:
- qualification rate;
- opportunity rate;
- customer rate.
CPA
Cost per action:
`CPA = Spend / Actions`
The action must be named.
Examples:
- purchase CPA;
- signup CPA;
- qualified-lead CPA.
The closer the action is to business value, the more useful it becomes.
CTR
Click-through rate:
`CTR = Clicks / Impressions`
CTR is a media signal.
It can indicate:
- creative relevance;
- search intent;
- audience response.
It does not prove business value.
High CTR can produce poor conversion quality.
CPC
Cost per click:
`CPC = Spend / Clicks`
Useful for media-market pricing.
Not a business outcome.
A more expensive click can be economically better if it converts at a higher rate or produces stronger customers.
CPM
Cost per 1,000 impressions:
`CPM = Spend / Impressions × 1,000`
Useful for:
- reach;
- auction pressure;
- media pricing.
Do not compare CPMs without considering audience, placement and objective.
Revenue per visitor
Formula:
`Revenue per Visitor = Revenue / Visitors`
This metric combines:
- conversion;
- order value.
It is especially useful for ecommerce and monetized media.
It can still hide margin differences.
Lead-to-customer rate
Formula:
`Customers / Leads`
This helps distinguish:
- lead volume;
- lead quality;
- sales conversion.
For long sales cycles, align cohorts by acquisition period.
Use a metric dictionary operationally
The downloadable workbook contains:
- metric name;
- category;
- definition;
- formula;
- unit;
- use;
- common trap;
- source.
A metric should also have internal governance fields in your own organization:
- owner;
- authoritative source;
- refresh cadence;
- segmentation rules.
The dictionary is a starting point.
Avoid benchmark dependency
Benchmarks can provide context.
They should not become targets automatically.
CAC varies by:
- category;
- margin;
- price;
- geography;
- sales motion.
ROAS varies by:
- attribution;
- repeat purchase;
- product mix.
Payback varies by:
- capital structure;
- business model.
The most useful benchmark is often your own trend.
Compare:
- cohort to cohort;
- channel to channel;
- quarter to quarter.
Metric hierarchy
A strong measurement system distinguishes:
Business outcomes
- revenue;
- profit;
- pipeline;
- retained revenue.
Unit economics
- CAC;
- LTV;
- payback;
- LTV:CAC.
Behavioral outcomes
- activation;
- conversion;
- retention.
Operational signals
- CTR;
- CPC;
- CPM;
- open rate.
This prevents the organization from optimizing the easiest metric.
The dictionary is designed to support exactly that discipline.
When teams agree on definitions, they can disagree about decisions productively.
Download the Marketing Metrics & Formulas Dictionary
When they disagree on definitions, the debate is often about two different realities.
