Marketing Analytics: The Five Numbers That Matter

Customer acquisition cost (CAC) is what you spend to win one customer. Harvard Business School Online gives the basic formula as marketing cost divided by the number of customers acquired in the same period.

CAC = acquisition spend in the period / customers acquired in the period

Three decisions make the number useful:

1. Decide what counts as spend. At minimum, include media and advertising spend. A fully loaded CAC also includes marketing and sales salaries, agencies, tools and commissions. Chargebee defines CAC as the total cost of gaining a customer, including marketing and sales expenses. Pick a definition and keep it, so the number is comparable month to month.
2. Calculate it by channel as well as overall. A blended CAC hides the difference between a channel that wins customers cheaply and one that loses money. Lean Labs notes that analyzing acquisition data shows which channels and campaigns drive conversions and which fail to return your investment.
3. Match the timing. If your sales cycle is two months long, this month's customers came from spend in earlier months. Use a lag, or compare over quarters, so a slow cycle does not distort the number.

A related figure is the CAC payback period: how many months of gross profit from a customer it takes to recover what you spent to acquire them. Chargebee describes it as the time it takes to earn back the money invested in acquiring customers. A short payback matters most for businesses with little cash, because money spent on acquisition is tied up until it is recovered.

A funnel is the sequence of steps a person takes from first contact to purchase. Adobe describes funnel analysis as measuring how customers progress through key steps, from initial engagement to conversion and beyond, with the aim of finding where they drop off.

A typical funnel for a business-to-business service:

Calculate the conversion rate between each pair of stages:

Stage conversion rate = people who reached the next stage / people who entered this stage

Stage-by-stage rates matter more than the single end-to-end rate, because they show where to act. If visitor-to-lead conversion is healthy but qualified leads rarely become opportunities, the problem is lead quality or sales follow-up, not the website.

Conversion rate also beats traffic as a measure of quality. Zimmer gives the example of a page with 1,000 visitors converting at 10%, which produces more than a page with 5,000 visitors converting at 1%. Before spending more to get traffic, check whether the traffic you already have converts.

Write down the definition of each stage and do not change it lightly. A "lead" that means a newsletter signup one quarter and a demo request the next makes trends meaningless.

Customer lifetime value (LTV) is the total profit a customer brings over the whole relationship. Harvard Business School Online describes it as the value a customer contributes over the time they are engaged with a company, counting the profit contribution of each transaction, not just one sale.

Two common ways to calculate it:

Simple:          LTV = annual contribution margin per customer x average customer lifetime in years
Subscription:    LTV = (average revenue per account x gross margin) / churn rate

The first formula is the one Harvard Business School Online uses. The second is common for subscription businesses. Wall Street Prep explains that dividing by churn estimates lifetime: a 5% annual churn rate implies an average lifetime of 20 years, because one divided by 5% is 20.

Points that keep LTV honest:

• Use profit, not revenue. Multiply by gross margin. A customer who generates a lot of revenue at a thin margin is worth less than the revenue suggests.
• Be cautious with young businesses. If you have been operating for a year, you do not yet know how long customers stay. Use a conservative estimate and revise it as data arrives.
• Calculate it by segment and channel. Customers from some channels stay longer than others. Knowing this changes how much each channel is worth.
• Watch retention closely. Retention drives lifetime value more than anything else. Zimmer notes that high churn can signal that marketing is attracting the wrong audience or setting the wrong expectations.

The LTV to CAC ratio compares what a customer is worth with what it cost to win them. Harvard Business School Online calls it perhaps the most common metric for measuring an early-stage venture's success.

LTV to CAC ratio = customer lifetime value / customer acquisition cost

Harvard Business School Online works through an example in which a customer is worth 345 in contribution margin over three years and cost 125 to acquire, giving a ratio of about 2.76. It quotes Professor Christina Wallace: "A good rule of thumb is that an LTV-to-CAC ratio of three or higher is attractive and indicates a scalable business."

How to read the ratio, drawing on Harvard Business School Online and Wall Street Prep's interpretation guide:

The last row surprises people. Wall Street Prep notes that a very high ratio can mean a company should spend more on acquisition, because it may be missing growth that competitors will take.

To improve the ratio, Harvard Business School Online suggests either raising LTV, through better customer profitability and satisfaction, or lowering CAC, by choosing channels that reach customers where they are with a clear value proposition.

Treat the ratio as a check, not a target to optimize alone. It is only as good as the two numbers underneath it, and both contain estimates.

The fifth number connects marketing activity to revenue. It answers which campaigns and channels contributed to the deals you closed. Zimmer describes return on marketing investment as revenue attributed to marketing, minus marketing spend, divided by marketing spend.

Return on marketing investment = (revenue attributed to marketing - marketing spend) / marketing spend

The hard part is the word "attributed". Most customers touch several channels before buying. An attribution model decides how to divide credit among them. HubSpot describes the main models:

HubSpot notes that single-touch models are simpler but miss the full picture, while multi-touch models give a fuller view of how channels work together. Adobe lists the limits every model shares: data gaps, people switching devices, offline interactions such as events and phone calls, and attribution bias, where teams adjust models to make their own channels look better.

Practical rules for a small team:

• Pick one model and keep it so trends are comparable. Last touch is a reasonable start for short sales cycles. First touch or U-shaped suits longer cycles where discovery matters.
• Add a "How did you hear about us?" field to signup or checkout. Self-reported answers catch word of mouth, podcasts and events that tracking misses.
• Use consistent tracking links with campaign tags so every channel is labeled the same way.
• Test incrementality when spend is large. Pause a channel in one region or period and see whether revenue really drops.

A dashboard should fit on one screen and be reviewed on a fixed rhythm.

Monthly review:

• CAC overall and by channel, with the trend over the last six months.
• Conversion rate at each funnel stage, compared with the previous month.
• Revenue attributed to marketing, by channel.

Quarterly review:

• LTV by segment, updated with current retention data.
• LTV to CAC ratio overall and by channel, and CAC payback period.
• A decision for each channel: increase, keep, fix or cut.

Rules for the review:

1. Start from the five numbers. Open the detailed reports only to explain a change.
2. Record every decision and its reason, so the next review can check whether it worked.
3. Agree definitions with sales and finance, so everyone reports the same numbers. Zimmer advises working with sales and customer success so metrics reflect the complete customer journey, not just the marketing portion.

Many small teams build this in a spreadsheet fed from their billing system, CRM and analytics tool. An AI analyst, such as the data and finance agents available on Quanta Intelligence, can help pull the figures together and explain changes, but the definitions and the decisions should stay with the people accountable for the budget.