Data analysis questions: 30 examples by department

Monday, 9 a.m. The dashboard is up on the big screen: sessions, clicks, impressions, followers. Everyone looks, nobody decides anything. Twenty minutes later the meeting ends with the most expensive phrase in business: "let's keep an eye on it." If that scene sounds familiar, the problem isn't your data, it's the questions nobody is asking it.

No dashboard answers anything on its own. It shows numbers; what turns a number into a decision is a well-built question. The gap between a team that "looks at reports" and one that decides with data lives almost entirely there.

Data analysis questions are formulations that connect a metric to a decision. A good one combines four elements: what to measure (metric), where (segment or cut), when (time period) and against what (a benchmark). For example: "what was our CAC by channel over the last 90 days compared to the previous quarter?" instead of "how is marketing doing?".

Why does the question matter more than the dashboard?

Because data without a question is decoration. A survey by the Brazilian analytics consultancy Beanalytic, covering more than 130 technology leaders, found that 78% of Brazilian companies are still in the early stages of data maturity. Most of them have reports. What's missing isn't another spreadsheet: it's knowing what to ask of it.

The prize for getting this right is substantial. According to McKinsey, companies that make intensive use of customer analytics are 23 times more likely to outperform competitors at customer acquisition and 9 times more likely at retention. It's not the volume of data that produces that result: it's the habit of interrogating it before every meaningful decision.

What separates a good question from a vague one?

Vague questions produce vague answers, and a vague answer can't carry a decision. Compare:

Vague question	Question that drives a decision
How are sales going?	At which funnel stage did we lose the most deals this month versus last month?
Is marketing working?	Which campaigns spent more than $1,000 in the last 30 days without a single conversion?
Is our cash position healthy?	If revenue drops 15% next quarter, how many months of runway do we have left?
Is Instagram growing?	How much of the traffic that becomes leads comes from Instagram, and at what cost per lead?

Notice the pattern: the right-hand column always carries a segment, a period and a benchmark. And every possible answer already suggests an action. If the answer to a question wouldn't change any decision you make, the question didn't need asking.

10 marketing questions to ask your data

A practical warning before the list: don't try to answer all ten at once. Pick the three most tied to this quarter's goal. To go deeper on the metrics behind them, see our guide to digital marketing KPIs.

1. Which channel brings the cheapest leads, and which brings the ones that become revenue? The cheap channel isn't always the one paying the bills.
2. What is CAC by channel over the last 90 days, compared to the previous quarter? Average CAC hides channels that are quietly getting expensive.
3. Which campaigns spent above a set threshold in the last 30 days without converting? The question that returns money fastest.
4. Which creative has the highest post-click conversion rate? High CTR with low conversion is a promise the landing page doesn't keep.
5. What share of leads comes from paid media versus organic, and how has that shifted in 6 months? It measures your dependence on ad budget.
6. What is the conversion rate at each funnel stage, and where is the biggest leak? Optimizing the wrong stage is wasted effort.
7. Which audience segment has the best ROAS, and does the budget reflect it? It almost never does.
8. Which search terms consume budget without converting? Adding negatives is silent, compounding optimization.
9. How much time passes between first touch and conversion? It defines your attribution window and the fair deadline for judging campaigns.
10. If the budget were cut 20% tomorrow, where would it come from with the least revenue loss? Teams that only answer this during a crisis answer it badly.

10 sales questions to ask your data

Here the raw material usually lives in the CRM, and the challenge is disciplined logging. If your sales funnel doesn't have clear stages yet, start there.

1. What is the conversion rate per pipeline stage, broken down by rep? It shows who needs coaching and on which skill.
2. At which stage do deals die most often, and after how many days? A stalled deal has an expiration date.
3. What is the average sales cycle by lead source? A referral and an ad-generated lead shouldn't be held to the same clock.
4. Which customer profile closes fastest and with the highest ticket? That's your ideal customer, drawn by data instead of intuition.
5. How many open opportunities have had no contact for over 14 days? An inflated pipeline yields optimistic forecasts and missed targets.
6. What is the average ticket per rep and per segment, and how has it evolved? A falling ticket with rising volume can mean too much discounting.
7. How many touches does it take, on average, to close? Most reps give up before that number.
8. What is this month's revenue forecast based on the current pipeline and historical rates? A forecast without historical conversion rates is a guess with a spreadsheet.
9. Customers from which source stay longest after closing? Selling well means selling to those who stay.
10. How much revenue is concentrated in your 10 largest customers? Heavy concentration is the risk nobody sees in a good month.

10 finance questions to ask your data

Finance tends to be the area with the most data and the fewest questions. These ten pair well with the essential financial indicators of any operation.

1. What is the contribution margin per product or service? Every portfolio has an item that only causes work; the data says which one.
2. How does the projected cash flow for the next 90 days compare with the actuals of the last 90? A projection that's never checked against reality never learns.
3. Which expenses grew faster than revenue over the last 6 months? Costs grow in silence; this question forces them to speak.
4. What is the average gap between paying suppliers and collecting from customers? That gap is the size of the cash cushion your operation demands.
5. What is the default rate by customer profile? Sometimes the problem isn't "the market" but one segment that deserves its own credit policy.
6. What is the monthly break-even point, and on which day of the month do you usually cross it? Knowing the day relocates the anxiety.
7. What does one day of stopped operations cost? Daily fixed cost is the number that gives urgency to every other decision.
8. What is the LTV/CAC ratio per segment? Growth that doesn't pay for itself is a loss with nice marketing.
9. If revenue drops 15%, how many months of cash remain? The stress test banks run and companies forget to run on themselves.
10. Which recurring contracts and subscriptions expire in the next 60 days? A renewal negotiated at the last minute is an expensive renewal.

How do you turn these questions into a routine?

Thirty questions are not a Monday-morning checklist. In practice it works like this: each area picks five to eight questions tied to the quarter's goals, defines who answers them, at what cadence, and where the answer gets recorded. The rest of the list is on-demand repertoire for specific decisions.

And every answer should end by producing one of three things: a decision, a sharper follow-up question, or an explicit confirmation that nothing needs to change. Analysis that ends in none of the three has become a hobby.

Frequently asked questions

What questions should I ask in a data analysis?

It depends on the decision you need to make. Start from the decision (cut a cost, reallocate budget, hire) and build the question with a metric, a segment, a period and a benchmark. The 30 questions in this guide cover the most common decisions in marketing, sales and finance.

How do I know if a data question is a good one?

Quick test: does it contain the four elements (metric, segment, period, benchmark), and would each possible answer lead to a different action? If any answer would leave everything as is, it's curiosity, not analysis.

How many questions should each area track?

Five to eight standing questions per area is usually the healthy ceiling. More than that dilutes attention and nobody digs deep into anything. The rest of the list works as repertoire for one-off decisions.

Do I need advanced tools to answer these questions?

Not to get started: a CRM, a spreadsheet and the native reports of ad platforms answer a good share. The bottleneck shows up when a question crosses sources, like CAC by channel against CRM revenue. That's where data integration or AI tools that do the crossing for you come in.

The right question beats the perfect dataset

Many companies postpone analysis waiting for the day their data will be complete, clean and integrated. That day doesn't come. Teams that decide well go the other way: they start from concrete questions and let those questions reveal which data is missing. Start this week with three questions from the list.

And if joining spreadsheets, CRM and ad platforms is what's blocking your answers, Sherlok handles that step: connect your sources once and ask these questions in plain language, getting the analysis back in seconds.