Marketing Analytics Flashcards

Question 1

What does the discipline of marketing analytics involve for a typical business?

Accepted Answer

Marketing analytics is the practice of collecting and analyzing data about customers, campaigns, and channels to understand performance and guide decisions. It combines measurement, statistics, and business judgment so that teams can see what is working, explain why, and choose what to do next. Its purpose is to replace guesswork with evidence.

Question 2

How does marketing analytics differ from simple marketing reporting?

Accepted Answer

Analytics differs from reporting because reporting summarizes what happened, such as visits or sales last month. Analytics goes further by asking why it happened, what is likely to happen next, and what action should follow. Reporting tends to be descriptive and recurring, while analytics is investigative and oriented toward decisions.

Question 3

What are the four common types of analytics that marketers use to answer business questions?

Accepted Answer

The four common types are descriptive, diagnostic, predictive, and prescriptive analytics. Descriptive analytics asks what happened, diagnostic asks why it happened, predictive estimates what will happen, and prescriptive recommends which action to take. Marketers usually move through them in roughly that order, because each stage relies on the one before it.

Question 4

Why do marketers need a clear business question before they start analyzing data?

Accepted Answer

Marketers need a clear question because it defines which data matters and which measures will show a useful answer. Without one, analysts often collect large amounts of information and produce charts that look impressive but change no decisions. Starting with a question such as whether a discount increased repeat purchases keeps the work focused.

Question 5

What is a key performance indicator, and how does it differ from an ordinary metric?

Accepted Answer

A key performance indicator is a metric tied directly to a specific business goal and monitored because it signals whether that goal is being met. An ordinary metric is any quantity that can be measured, such as page views or email opens. Every KPI is a metric, but not every metric is a KPI.

Question 6

What makes a good marketing metric, according to common analytics practice?

Accepted Answer

A good metric is relevant to a goal, understandable to the people who use it, and measurable with consistent data over time. It should be actionable, meaning a change in its value suggests a specific response. Vanity metrics, such as raw follower counts, fail these tests because they look encouraging while offering little guidance.

Question 7

What is the difference between a leading indicator and a lagging indicator?

Accepted Answer

A leading indicator is a measure that tends to move before an outcome changes, such as trial sign-ups that often precede purchases. A lagging indicator reflects results that have already happened, such as quarterly revenue or customer churn. Marketers use leading indicators for early adjustments and lagging indicators to confirm results.

Question 8

How should a marketer decide how often to review a dashboard of metrics?

Accepted Answer

A review cadence should match how quickly a metric changes and how quickly the team can act on it. Fast-moving paid media spending may need daily checks, while customer lifetime value shifts slowly and suits monthly or quarterly review. Reviewing too often invites overreaction to random noise, and reviewing too rarely lets problems grow unnoticed.

Question 9

What does conversion rate measure, and how is it usually calculated?

Accepted Answer

Conversion rate measures the share of visitors or prospects who complete a desired action, such as a purchase or form submission. It is calculated by dividing conversions by total visitors, then multiplying by one hundred. For example, 500 purchases from 10,000 visits gives a 5 percent conversion rate.

Question 10

What is customer acquisition cost, and why do marketers track it?

Accepted Answer

Customer acquisition cost is the total spending on marketing and sales needed to win one new customer over a given period. It is calculated by dividing acquisition spending by the number of new customers gained in that period. Marketers track it to check whether new customers are worth acquiring.

Question 11

What is customer lifetime value, and how does it relate to acquisition cost?

Accepted Answer

Customer lifetime value is the net profit a business expects from a typical customer over the whole relationship. Comparing it with acquisition cost shows whether acquisition spending is sustainable. When lifetime value is well above acquisition cost, a company can invest in growth. When it falls below, each new customer loses money.

Question 12

What does return on ad spend tell a marketer, and how is it calculated?

Accepted Answer

Return on ad spend shows how much revenue a campaign generates for each dollar spent on advertising. It is calculated by dividing revenue attributed to the campaign by the cost of the campaign. A result of 4 means that every dollar spent produced four dollars of revenue, although this figure does not reveal profit.

Question 13

Why is email open rate considered a weaker measure than click-through rate?

Accepted Answer

Open rate is considered weaker because it depends on whether the email client loads tracking images, and privacy features block or preload them, distorting the count. Click-through rate requires a recipient to take a deliberate action, so it reflects engagement more directly. Clicks and downstream conversions are therefore more reliable signals.

Question 14

What is a marketing funnel, and why do analysts measure stage-by-stage drop-off?

Accepted Answer

A marketing funnel is a model of the path prospects take from first awareness through consideration, purchase, and repeat business. Measuring drop-off between stages reveals where the largest losses occur, which shows where improvement matters most. A strong overall conversion rate can hide a weak stage, so each step deserves its own measure.

Question 15

How do churn rate and retention rate differ, and why does each matter?

Accepted Answer

Churn rate is the share of customers who stop buying or cancel during a period, while retention rate is the share who remain. Retention is roughly one minus churn for the same period. Both matter because keeping customers is usually cheaper than acquiring new ones.

Question 16

How can a marketer measure the average order value, and why is it useful?

Accepted Answer

Average order value is calculated by dividing total revenue by the number of orders over a given period. It helps marketers judge whether promotions, bundles, or recommendations are encouraging larger purchases. Because even a small increase across thousands of orders can add meaningful revenue, it is a useful lever to monitor with conversion rate.

Question 17

What is a tracking tag, and how does it help collect website behavior data?

Accepted Answer

A tracking tag is a snippet of code on web pages that sends data to an analytics system when a visitor acts. Tags can record page views, clicks, form submissions, and purchases. Because they depend on code being installed correctly, teams should test them regularly.

Question 18

What is the difference between first-party data and third-party data?

Accepted Answer

First-party data is information a business collects directly from its own customers and visitors, such as purchase records. Third-party data is gathered by outside organizations and purchased or shared with a business, often covering broader audiences. First-party data is usually more accurate and easier to govern, so many marketers prioritize it.

Question 19

Why does a consistent naming convention for campaigns and tracking parameters matter?

Accepted Answer

Consistent names let analysts group results by channel, audience, or offer without manually cleaning spreadsheets. When one campaign is labeled three different ways, reports split its results and understate its performance. A shared convention, documented and followed by everyone, keeps comparisons accurate across teams and time periods.

Question 20

What are UTM parameters, and what information do they usually carry?

Accepted Answer

UTM parameters are short labels added to the end of a web address to identify where a visit came from. They commonly include source, medium, campaign name, and sometimes content or term, so that analytics tools can attribute traffic to specific efforts. Inconsistent typing fragments reports, so labels must match exactly.

Question 21

What is a data sampling problem, and how can it distort marketing conclusions?

Accepted Answer

A sampling problem occurs when the data analyzed does not represent the full population of interest, such as surveying only customers who responded to an email. Conclusions drawn from an unrepresentative group can overstate or understate true preferences and results. Analysts reduce this risk by checking who is missing from the data.

Question 22

Why is data quality important, and what are common signs of poor data?

Accepted Answer

Data quality determines whether analysis reflects reality, because flawed inputs produce misleading outputs regardless of how sophisticated the methods are. Common warning signs include duplicate customer records, missing values, sudden unexplained jumps in traffic, and metrics that change definition without notice. Routine validation checks help catch these problems before decisions depend on them.

Question 23

What is a survey, and what makes survey results more trustworthy?

Accepted Answer

A survey asks a group of respondents structured questions to collect opinions, attitudes, or reported behavior. Results are more trustworthy when the sample is representative, the questions are neutral and clearly worded, and response rates are high enough to limit bias. Surveys capture what people report, which may differ from what they do.

Question 24

How do privacy rules and browser restrictions affect what marketers can track?

Accepted Answer

Privacy rules and browser restrictions limit cookie-based tracking, which reduces the data marketers can collect. Cookies are small files that let websites recognize a browser across visits, so consent requirements and blocked third-party cookies shrink what is visible. Analysts respond with first-party data, aggregated measurement, and modeling.

Question 25

What is attribution in marketing analytics, and why is it difficult?

Accepted Answer

Attribution is the process of assigning credit for a conversion to the marketing touchpoints that contributed to it. It is difficult because customers often interact with several channels across many days, while devices, privacy limits, and offline sales hide parts of the journey. No method reveals the full picture, so results are estimates.

Question 26

What is the difference between last-click and first-click attribution?

Accepted Answer

Last-click attribution gives all credit for a conversion to the final touchpoint before the sale, while first-click attribution gives all credit to the initial touchpoint that introduced the customer. Each is simple and popular, but each ignores the middle of the journey. Both can exaggerate certain channels, so they work best as comparisons.

Question 27

What is multi-touch attribution, and what kinds of models does it include?

Accepted Answer

Multi-touch attribution spreads credit for a conversion across several touchpoints rather than assigning it to just one. Common rule-based versions include linear, which splits credit evenly, time-decay, which favors recent interactions, and position-based, which emphasizes the first and last touches. Data-driven models instead estimate each channel's contribution statistically from historical paths.

Question 28

What is customer segmentation, and what are common ways to divide customers into segments?

Accepted Answer

Customer segmentation divides a customer base into groups that share characteristics, so that marketing can be tailored to each group's needs. Common bases include demographics, geography, purchase history, engagement level, and stated interests. Effective segments are large enough to matter, distinct from one another, and different enough in behavior to justify separate messages.

Question 29

What is cohort analysis, and what question does it help answer?

Accepted Answer

Cohort analysis groups customers by a shared starting point, such as the month they first purchased, and then tracks their behavior over time. It helps answer whether newer customers behave differently from older ones, and whether retention is improving or declining. Comparing cohorts reduces distortion from looking only at overall averages.

Question 30

What is a marketing mix model, and what kind of decision does it inform?

Accepted Answer

A marketing mix model is a statistical analysis that estimates how spending across channels, pricing, and seasonality relate to sales over time. It helps leaders divide a budget among channels without tracking individual customers. Relying on aggregate historical data, it suits broad budget decisions more than daily tuning.

Question 31

What is a customer journey map, and what does it reveal?

Accepted Answer

A customer journey map is a structured outline of the steps a customer takes from first awareness through purchase and beyond. It reveals where customers get stuck, which touchpoints shape decisions, and where data is missing. Teams use it to align analytics with real experiences rather than assumed ones.

Question 32

How does RFM analysis rank customers, and what do its three letters stand for?

Accepted Answer

RFM analysis scores customers on recency, frequency, and monetary value, meaning time since last purchase, number of purchases, and total spending. Each dimension is typically scored from one to five, and combined scores help group customers, such as loyal or dormant. It is a simple way to prioritize retention efforts.

Question 33

Why does correlation not establish causation, and why does this matter for marketing decisions?

Accepted Answer

Correlation does not establish causation because two measures can move together without one producing the other. Ice cream sales and sunburn cases rise together because both follow warm weather, not because one causes the other. Marketers who confuse the two may cut working promotions or fund ineffective ones.

Question 34

What is an A/B test, and what makes it a controlled experiment?

Accepted Answer

An A/B test compares two versions of a marketing element, such as an email subject line, by randomly showing each to different people. It is a controlled experiment because the only systematic difference between the groups is the version they saw. Random assignment lets any outcome difference be attributed to the change.

Question 35

What is a holdout group, and why do marketers use one when measuring campaign impact?

Accepted Answer

A holdout group is a portion of the eligible audience withheld from a campaign for comparison. Without a holdout, marketers cannot tell whether purchases would have happened anyway, which overstates the campaign's true effect. Comparing exposed and holdout groups estimates incremental lift, the additional outcome the campaign actually caused.

Question 36

What does statistical significance indicate in an experiment, and what does it not prove?

Accepted Answer

Statistical significance indicates that an observed difference would be unlikely to arise from random variation alone if there were truly no effect. It does not prove that the effect is large, important, or worth the cost of implementing. Sample size matters too, so tiny differences can look significant in very large tests.

Question 37

How should a marketer decide how long to run an experiment before judging its results?

Accepted Answer

A test should run until it reaches its pre-set sample size and covers a full cycle of customer behavior. Stopping early when a result first looks significant inflates false positives. Setting the sample size and stopping rule before launch guards against wishful interpretation.

Question 38

What is a confidence interval, and how does it help interpret a measured lift?

Accepted Answer

A confidence interval is a range likely to contain the true effect, given the data's variability. A 3 percent lift with an interval from negative 1 to 7 percent could reflect a small, zero, or moderately positive effect. Reporting the interval, not only the point estimate, keeps decisions honest about uncertainty.

Question 39

What is diminishing returns in marketing spend, and why does it matter for budgeting?

Accepted Answer

Diminishing returns means each extra dollar spent on a channel tends to produce less result than the one before. It matters for budgeting because the most efficient spend level is usually lower than the largest affordable one. Analysts estimate response curves or test several spend levels to find where extra investment stops paying off.

Question 40

Why should marketers be cautious about acting on results from a single test?

Accepted Answer

A single test can mislead because it is one sample from a process that varies, so its result may not repeat when conditions change. Seasonality, audience differences, and chance can all produce a result that looks convincing but does not generalize. Confirming important findings through replication or holdout checks builds more reliable evidence before major budget decisions.

Marketing Analytics Flashcards

1Foundations of marketing analytics

2Core metrics and performance indicators

3Data collection and tracking

4Attribution and customer analysis

5Experiments and applied decisions

Learning Resources