Английский для анализа данных: ключевые термины и фразы

"data pipeline"

An automated series of steps that ingests, transforms, and loads data from source systems to a destination

"The engineering team built a data pipeline that pulled sales records from the CRM every hour and loaded them into the warehouse."

"granularity"

The level of detail captured in a dataset; high granularity means more detailed, row-level records

"The finance team asked us to increase the granularity of the report from monthly to daily transaction data."

"data wrangling"

The process of cleaning, restructuring, and enriching raw data to make it suitable for analysis

"She spent most of Monday on data wrangling before the dataset was clean enough to build the dashboard."

"cohort analysis"

A technique that groups users or events by a shared characteristic — such as sign-up date — to compare behaviour over time

"The cohort analysis revealed that users who signed up in Q1 had a 30% higher retention rate after 90 days."

"key performance indicator"

A quantifiable metric used to evaluate the success of an organisation, team, or individual against defined objectives

"Customer acquisition cost was the key performance indicator the marketing director tracked most closely each quarter."

"outlier"

A data point that falls significantly outside the expected range and may indicate an error or a genuinely unusual event

"An outlier in the transaction dataset turned out to be a legitimate bulk order, not a data entry mistake."

"normalisation"

The process of scaling numeric data to a common range so that no single variable dominates an analysis due to its magnitude

"Before training the model, the analyst applied normalisation to bring all features into the zero-to-one range."

"data dictionary"

A reference document that defines each field in a dataset, including its name, data type, allowed values, and business meaning

"New analysts were told to consult the data dictionary before writing any SQL queries against the production database."

"confidence interval"

A range of values that is likely to contain the true population parameter with a specified level of certainty, typically 95%

"The A/B test showed a 12% lift in conversion with a 95% confidence interval of 8% to 16%."

"statistical significance"

A result that is unlikely to have occurred by chance alone, typically indicated by a p-value below 0.05

"The difference in average order value between the two user segments reached statistical significance after two weeks."

"standard deviation"

A measure of how spread out values are around the mean; a high standard deviation indicates high variability

"The standard deviation in delivery times was 4.2 days, suggesting inconsistent performance across fulfilment centres."

"median"

The middle value in a sorted dataset, often preferred over the mean when the data contains extreme outliers

"We used the median rather than the mean revenue per user to avoid distortion from a handful of very large accounts."

"correlation coefficient"

A number between -1 and 1 that measures the strength and direction of the linear relationship between two variables

"The correlation coefficient between ad spend and revenue was 0.87, suggesting a strong positive relationship."

"p-value"

The probability of observing a result at least as extreme as the one obtained, assuming the null hypothesis is true

"With a p-value of 0.02, the analyst rejected the null hypothesis and declared the experiment result significant."

"regression to the mean"

The phenomenon where extreme values tend to be followed by more average values in subsequent observations

"The team attributed the sales rebound not to their new strategy but to regression to the mean after an unusually bad month."

"variance"

The average of the squared differences from the mean, measuring how widely data points are dispersed

"High variance in customer satisfaction scores indicated that the experience differed greatly by region."

"JOIN"

A SQL operation that combines rows from two or more tables based on a related column

"She wrote an INNER JOIN to combine the orders table with the customers table on the customer_id field."

"aggregate function"

A SQL function that performs a calculation on a set of values and returns a single result, such as SUM, COUNT, or AVG

"The analyst used the COUNT aggregate function to tally the number of transactions per product category."

"window function"

A SQL function that performs calculations across a set of rows related to the current row without collapsing them into a single result

"He used a window function with PARTITION BY to calculate each salesperson's running total within their region."

"query optimisation"

The process of rewriting or restructuring SQL queries to reduce execution time and resource consumption

"Query optimisation reduced the dashboard load time from 45 seconds to under 3 seconds on the production database."

"primary key"

A column or set of columns that uniquely identifies each row in a database table

"The orders table uses order_id as its primary key, ensuring no two transactions share the same identifier."

"subquery"

A SQL query nested inside another query, used to break complex problems into logical steps

"The analyst wrote a subquery to first filter active users before joining them to the subscription history table."

"index"

A database structure that speeds up data retrieval by creating a sorted reference to one or more columns

"Adding an index on the created_at column reduced the time for date-range queries from minutes to milliseconds."

"CTE"

Common Table Expression — a temporary named result set defined with the WITH clause to simplify complex queries

"The senior analyst refactored the 200-line SQL query into three clean CTEs, making the logic much easier to review."

"drill-down"

The ability to navigate from a high-level summary in a chart or dashboard to more detailed underlying data

"The executive clicked the regional revenue bar and used drill-down to see which product lines were underperforming."

"heat map"

A data visualisation that uses colour intensity to represent the magnitude of values across a two-dimensional grid

"The heat map of website clicks showed that users rarely scrolled below the fold on the product landing page."

"scatter plot"

A graph that displays the relationship between two numerical variables using dots plotted on an x-y axis

"The scatter plot of marketing spend versus revenue made the positive correlation immediately visible to the leadership team."

"KPI dashboard"

An interactive visual display that consolidates the most important metrics for monitoring business performance

"The weekly KPI dashboard gave the operations team a single view of order volume, fulfilment rate, and customer satisfaction."

"data storytelling"

The practice of combining data, visuals, and narrative to communicate insights in a way that motivates decisions or action

"The analyst used data storytelling techniques to turn a dry spreadsheet of churn figures into a compelling board presentation."

"chart junk"

Unnecessary visual elements in a chart that add complexity without conveying additional information

"The design review removed 3D bars and decorative shadows that were identified as chart junk cluttering the report."

"time series"

A sequence of data points recorded at successive, equally spaced intervals of time, used to identify trends and seasonality

"The time series chart of daily active users revealed a consistent weekly pattern with peaks on Tuesdays and Thursdays."

"stacked bar chart"

A bar chart in which each bar is divided into segments representing different sub-categories of the total value

"A stacked bar chart showed revenue broken down by product line for each quarter, making contribution trends easy to spot."

"data warehouse"

A centralised repository that stores large volumes of structured, historical data optimised for analytical queries

"The company migrated from spreadsheets to a cloud data warehouse to handle the growing volume of transaction records."

"ETL"

Extract, Transform, Load — the process of moving data from source systems, cleaning it, and loading it into a target database

"The ETL job ran every night at 2 a.m. to refresh the reporting tables with the latest customer activity."

"data mart"

A subset of a data warehouse focused on a specific business function, such as sales, finance, or marketing

"The finance team had its own data mart that contained only the cost and revenue tables relevant to their reporting."

"OLAP cube"

An Online Analytical Processing structure that pre-aggregates data across multiple dimensions for fast multidimensional analysis

"The OLAP cube allowed the analyst to pivot revenue by region, product, and time period in milliseconds."

"self-service analytics"

An approach that empowers business users to create their own reports and queries without relying on a data team

"After rolling out self-service analytics in Tableau, the marketing team reduced their data requests to the BI team by 60%."

"data governance"

The set of policies, standards, and processes that ensure data is accurate, consistent, secure, and used appropriately across an organisation

"The data governance committee defined who could access customer PII and how long it could be retained for analysis."

"star schema"

A database design pattern where a central fact table is surrounded by dimension tables, resembling a star shape

"The analyst modelled the data warehouse using a star schema with a sales fact table linked to date, product, and customer dimensions."

"ad hoc analysis"

An unplanned, one-off analysis performed in response to a specific business question that is not covered by existing reports

"The CEO requested an ad hoc analysis of refund rates by supplier after a spike in customer complaints."

"actionable insight"

A finding from data analysis that is specific and clear enough to directly inform a business decision or action

"The analyst presented three actionable insights from the churn data, each with a recommended intervention and projected impact."

"data-driven decision"

A decision made primarily on the basis of quantitative evidence rather than intuition or anecdotal experience

"The product team committed to a data-driven decision process by requiring A/B test results before any feature launch."

"baseline"

A reference measurement taken before a change is made, used to evaluate the impact of an intervention

"They established a baseline conversion rate of 2.4% before running the experiment on the new checkout flow."

"root cause analysis"

A structured investigation method used to identify the underlying reason for a problem, not just its symptoms

"After the sales drop, the team conducted a root cause analysis and traced the issue to a broken promotional email link."

"data caveat"

A qualification or limitation noted alongside an analytical result to flag potential issues with the data or methodology

"The report included a data caveat explaining that mobile app data was missing for the first three days of the period."

"north star metric"

A single, primary metric that best captures the core value a product delivers to users and aligns the whole team

"The analytics lead declared weekly active users as the north star metric and restructured all reporting around it."

"benchmark"

A standard or reference point against which performance can be compared, such as an industry average or a prior period

"The 4.2-minute average session duration compared favourably to the industry benchmark of 2.8 minutes."

"hypothesis"

A specific, testable statement about an expected relationship or outcome that guides an analysis or experiment

"The analyst framed the investigation as a hypothesis: users who complete onboarding are 3× more likely to convert to paid."