5 exercises on the business metrics IT professionals encounter in product discussions.
Business metric vocabulary essentials
DAU/MAU: Daily/Monthly Active Users — engagement breadth
Churn rate: % of customers who cancel in a given period
CAC: Customer Acquisition Cost — what it costs to win a customer
LTV: Lifetime Value — expected revenue from a customer over their lifetime
0 / 5 completed
1 / 5
A PM says: "Our DAU/MAU ratio is 0.45." What does this mean, and is it good or bad?
DAU/MAU Ratio — the stickiness metric
DAU/MAU = Daily Active Users ÷ Monthly Active Users
A ratio of 0.45 means 45% of your monthly users open the app on any given day.
Benchmarks:
Ratio
Signal
0.20 (20%)
Low — users open monthly but not often
0.25 (25%)
Average for many apps
0.45 (45%)
Strong — indicates high daily habit formation
0.50+ (50%+)
Excellent — WhatsApp, Facebook Messenger range
Contextual phrases:
"Our DAU/MAU stickiness ratio is [X%]."
"We're targeting a 30%+ DAU/MAU ratio by Q3."
"DAU has been growing faster than MAU — our engagement is improving."
2 / 5
A product update states: "Monthly churn rate dropped from 4.2% to 2.8% after the onboarding improvements." How would you describe this in a presentation to stakeholders?
Communicating churn improvements professionally
The professional phrasing here has three layers:
Absolute change: "1.4 percentage points" — the direct difference (not %)
Business impact: "extend average customer lifetime and improve LTV" — connects the metric to revenue consequences
Important vocabulary distinction: "1.4 percentage points" ≠ "1.4%". Going from 4.2% to 2.8% is a drop of 1.4 percentage points, but a 33% relative reduction.
Why D is wrong: Churn is the rate of customers who leave. 2.8% churn means 2.8% of customers cancelled — not 97.2%.
Churn vocabulary:
"Monthly churn rate of [X%]" — most common SaaS metric
"Gross churn" vs. "net churn" — gross = raw cancellations; net accounts for expansions
3 / 5
An engineer asks: "Is our LTV:CAC ratio healthy?" The company spends €120 to acquire a customer (CAC) and the expected lifetime revenue is €480 (LTV). What's the ratio, and how do you describe it?
LTV:CAC Ratio — the unit economics test
LTV:CAC = 480 / 120 = 4:1
This means for every €1 spent acquiring a customer, the expected lifetime return is €4.
Benchmarks:
Ratio
Signal
Below 1:1
Losing money per customer — unsustainable
1:1 – 3:1
Break-even or marginal — difficult to grow
3:1
Acceptable baseline for SaaS businesses
4:1+
Healthy — room for growth investment
Professional phrasing:
"Our LTV:CAC is 4:1 — within the healthy range."
"We're targeting a 5:1 LTV:CAC ratio by end of year."
"The unit economics are solid — each customer pays back acquisition cost within 6 months."
4 / 5
A data analyst says: "The A/B test reached statistical significance after 14 days with p = 0.03." What does p = 0.03 mean in plain English?
p-value — what it means in plain English
The p-value is the probability that the observed difference between test and control could have occurred by random chance if there were actually no real effect.
p = 0.03 means:
There is a 3% probability this result is due to random variation
With a standard threshold of p < 0.05, this result is statistically significant
We can confidently conclude the treatment had a real effect
Standard thresholds:
p < 0.05: significant (industry standard)
p < 0.01: highly significant
p > 0.05: not significant — the result may be noise
Important caveat: Statistical significance ≠ practical significance. A 0.01% conversion lift might be statistically significant but commercially irrelevant.
A/B test vocabulary:
"The test reached significance after [N] days."
"The lift was [X%] with a p-value of [Y]."
"The control and variant are statistically indistinguishable — we need more data."
5 / 5
A product manager presents: "The feature improved conversion from trial to paid by 2 percentage points, from 8% to 10%." Which follow-up question demonstrates the strongest analytical thinking?
Critical thinking about conversion metrics
A 2 percentage point improvement from 8% to 10% is a 25% relative lift — which sounds impressive. But the right analytical questions are:
Sample size: A 2pp lift on 50 users is noise. On 10,000 users, it's meaningful.
Statistical significance: Was there a control group? What was the p-value?
Seasonality: Did the time period coincide with a promotion, pricing change, or seasonal demand shift?
Attribution: Was the improvement actually caused by the feature, or by something else that changed at the same time?
Why A is insufficient: Simply echoing the metric back as a percentage ("went up by 2%") doesn't add analytical value — and technically it went up by 2 percentage points, not 2%.