# What Is Analytics? Definition, Types and Examples

Analytics is the systematic use of data, statistical methods and software to find patterns and support decisions. The analytics definition most practitioners use is broader than "doing math on numbers": it covers the whole process from collecting data to acting on what the data shows. This article explains what analytics means, the four main types, and how it differs from data analysis.

## Quick Answer

- Analytics is the practice of turning raw data into decisions using statistics, software and domain knowledge.
- The four widely used types are descriptive, diagnostic, predictive and prescriptive analytics [1].
- Descriptive analytics explains what is happening. Diagnostic analytics explains why it happened. Predictive analytics estimates what is likely to happen. Prescriptive analytics suggests what to do [1].
- Analytics is the broader discipline. Data analysis is the work of examining a dataset to find patterns and insight [2].
- Business analytics applies these methods to business decisions in finance, marketing and operations [3].

## What Analytics Means

In plain terms, analytics is what you do when you take a pile of data and turn it into something a person can act on. That might be a dashboard, a forecast, a recommendation or a written finding. The output is always meant to inform a decision, not to sit in a file.

The precise definition is narrower. Analytics is the application of statistical, mathematical and computational techniques to data in order to describe past states, explain observed outcomes, estimate future states, or rank possible actions by expected value. Each of those four verbs maps to one of the four types below.

The word also names a field and a job family. A person who does this work is an analyst, and the skills involved include data handling, statistics, visualization and communication [2].

## How It Works

Analytics runs on a simple loop: measure, explain, estimate, decide. The mechanism underneath is statistical inference. You compare what you observed against what you would expect if nothing interesting were going on.

For a count of categories, the standard test is the chi-square goodness-of-fit statistic:

$$\chi^2 = \sum_{i=1}^{k} \frac{(O_i - E_i)^2}{E_i}$$

Each symbol means:

- $O_i$ is the observed count in category $i$.
- $E_i$ is the expected count in category $i$ under your assumption.
- $k$ is the number of categories.
- $\chi^2$ is the total squared deviation, scaled by the expected counts.

The degrees of freedom are $k - 1$. A large $\chi^2$ relative to the degrees of freedom means the observed counts differ from what you assumed.

## Worked Example

Here is a small dataset of six business scenarios, each classified into one of the four analytics types with a one-line justification.

| scenario | analytics_type | justification |
|---|---|---|
| Monthly sales report | Descriptive | Summarizes what happened last month |
| Churn prediction | Predictive | Estimates which customers will leave |
| A/B test on landing page | Diagnostic | Explains why variant B converts more |
| Executive KPI dashboard | Descriptive | Tracks current metrics at a glance |
| Demand forecast for Q4 | Predictive | Projects future sales volume |
| Inventory reorder recommendation | Prescriptive | Suggests the optimal order quantity |

Step 1. Count the scenarios. There are $n = 6$ rows.

Step 2. Count each type. Descriptive: 2. Diagnostic: 1. Predictive: 2. Prescriptive: 1.

Step 3. Compute the share of Descriptive. That is $2 / 6 = 0.3333$, or 33.3%.

Step 4. Set the expected count under a uniform assumption. With four equally likely types, $E = 6 / 4 = 1.5000$ for each.

Step 5. Compute the chi-square statistic:

$$\chi^2 = \frac{(2-1.50)^2}{1.50} + \frac{(1-1.50)^2}{1.50} + \frac{(2-1.50)^2}{1.50} + \frac{(1-1.50)^2}{1.50} = 0.6667$$

Step 6. Degrees of freedom: $4 - 1 = 3$.

In a spreadsheet, the count comes from a conditional count formula. `=COUNTIF(B2:B7,"Descriptive")` returns 2, and `=ROUND(33.3333,1)` returns 33.3.

The same count in Python:

```python
import pandas as pd
df = pd.DataFrame({
    "scenario": ["Monthly sales report","Churn prediction","A/B test on landing page",
                 "Executive KPI dashboard","Demand forecast for Q4",
                 "Inventory reorder recommendation"],
    "type": ["Descriptive","Predictive","Diagnostic","Descriptive","Predictive","Prescriptive"]
})
print(df["type"].value_counts())
```

Output: Descriptive: 2, Diagnostic: 1, Predictive: 2, Prescriptive: 1 (chi2 = 0.6667, df = 3).

## How to Interpret It

The chi-square value of 0.6667 with 3 degrees of freedom is small. The observed counts are close to what you would expect if the four types were equally common in this set. With only six scenarios, that is not evidence of anything beyond the sample itself.

The practical reading is about the mix, not the test. Two of six scenarios are descriptive and two are predictive, so this small portfolio leans toward reporting and forecasting. Only one scenario is prescriptive, which means the set is light on recommendations that change what someone does next.

When you classify your own work this way, the useful question is whether the mix matches your goals. A team that wants to move from reporting to decision support should see the prescriptive share grow over time.

## When to Use It (and when not to)

Use analytics when you have data that reflects a real process and a decision that depends on it. Descriptive analytics fits recurring reporting. Diagnostic analytics fits a specific "why did this change" question. Predictive analytics fits planning where the future resembles the past. Prescriptive analytics fits situations where you can act on a recommendation and measure the result [1].

Do not use analytics when the data does not measure what you think it measures, or when the sample is too small to support the claim you want to make. Do not use predictive methods on a process that has just changed in a way your historical data cannot capture. Do not use prescriptive output as an instruction without a human who understands the constraints.

## Analytics vs Data Analysis

The two terms overlap heavily, and many people use them interchangeably. The distinction is scope. Data analysis is the act of examining a dataset to find patterns and insight, in any domain [2]. Analytics is the wider system: the data pipeline, the methods, the tools, the reporting layer and the decisions that follow.

| Aspect | Analytics | Data Analysis |
|---|---|---|
| Scope | End-to-end process and decision support | Examination of a dataset |
| Typical output | Dashboards, forecasts, recommendations | Findings, patterns, insight |
| Orientation | Ongoing function inside an organization | A task or a project |
| Related field | Business analytics, data analytics [2] | Statistics, data science |

Business analytics is a related term that sits closer to strategy and operations, while data analytics can extend into research, engineering and science [2]. Business analytics is often described as the combination of skills, technologies, applications and processes used to gain data-driven insights for decision making across finance, marketing and operations [3].

If you want the narrower concept first, start with what data analysis is and then come back to the broader picture.

## Common Mistakes

- Treating a dashboard as the whole of analytics. A dashboard is descriptive output. The fix is to add a diagnostic question and a decision the dashboard is meant to trigger.
- Skipping the diagnostic step. Teams jump from "sales fell" to a prediction without asking why. The fix is to test candidate explanations before forecasting.
- Using predictive models on a changed process. If pricing, policy or the market shifted, history may not apply. The fix is to check for structural breaks before trusting a forecast.
- Reporting a chi-square or p-value without the counts. A statistic alone hides how thin the data is. The fix is to show observed and expected counts next to the test result.
- Confusing correlation with a cause in diagnostic work. The fix is to name the mechanism you believe is operating and look for evidence that distinguishes it from alternatives.
- Building prescriptive recommendations with no feedback loop. The fix is to record what was recommended, what was done, and what happened.

## Limitations

Analytics cannot tell you what you should value. It can rank options by an objective you specify, but choosing the objective is a human decision. A model that maximizes short-term revenue may damage retention, and no statistic in the dataset will flag that unless you put retention in the objective.

Analytics also inherits every flaw in the data. Missing values, biased sampling, inconsistent definitions and measurement error all pass straight through to the output. Small samples make results unstable, and the chi-square example above shows the pattern: with six observations, a statistic of 0.6667 on 3 degrees of freedom carries almost no weight. Treat analytics as evidence, not as proof.

## Frequently Asked Questions

### What is the simplest analytics definition?

Analytics is the use of data and statistical methods to describe what happened, explain why, predict what comes next, and recommend what to do. It covers the tools and processes around those four activities, not just the calculations.

### What are the four types of analytics?

Descriptive, diagnostic, predictive and prescriptive [1]. Descriptive explains what is happening, often through a dashboard. Diagnostic determines the cause of something that has happened. Predictive uses available data to estimate what is likely to happen. Prescriptive suggests how to optimize practices for different possible outcomes [1].

### Is analytics the same as data analysis?

No, though they overlap. Data analysis is analyzing data for patterns and insight in any domain. Analytics is the broader function that includes data pipelines, methods, tools and the decisions the analysis supports [2]. In everyday use, many people treat the two as synonyms.

### What is the difference between data analytics and business analytics?

Data analytics can extend into research, engineering and science, while business analytics sits closer to the strategy and operations side of an organization [2]. Business analytics degrees typically combine business core courses with analytics coursework in visualization, business intelligence and applied statistics [2].

### Do I need programming to do analytics?

No. Spreadsheets handle counting, summarizing and simple tests, and tools like Tableau and Power BI cover reporting and dashboards [2]. Programming becomes useful when datasets grow, when you need repeatable pipelines, or when you want methods that spreadsheet functions do not provide. If you are starting from the data side, what a dataset is is a good next read, and data analytics methods covers which technique fits which question. For a domain view, see healthcare analytics, and for the storage layer underneath, see what a database is.

## References

1. [What Is Big Data Analytics & How It Works - Bay Atlantic University - Washington, D.C.](https://bau.edu/blog/what-is-big-data-analytics/)
2. [Data Analytics vs. Business Analytics: Key Differences](https://online.siue.edu/degrees/undergraduate/bsba/gen-program/business-analyst-career-guide/)
3. [Business Analytics | Bachelor's Degrees at Manor College](https://manor.edu/academics/bachelordegree/business-administration/business-analytics/)

## Further Reading

- [Wilson G, Bryan J, Cranston K et al. (2017). Good enough practices in scientific computing. PLOS Computational Biology](https://doi.org/10.1371/journal.pcbi.1005510)
- [Wilkinson MD, Dumontier M, Aalbersberg IJ et al. (2016). The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data](https://doi.org/10.1038/sdata.2016.18)
- [NIST/SEMATECH e-Handbook of Statistical Methods](https://www.itl.nist.gov/div898/handbook/index.htm)

## Related Articles

- [What Is Healthcare Analytics? Definition and Examples](/blog/data-analysis/what-is-healthcare-analytics)
- [What Is Data Analysis? Definition, Steps and Examples](/blog/data-analysis/what-is-data-analysis-definition)
- [Data Analytics Methods: Types and When to Use Each](/blog/data-analysis/data-analytics-methods)
- [What Is Data Mining? Definition, Meaning and Examples](/blog/data-analysis/what-is-data-mining)
- [What Is a Dataset? Definition, Types and Examples](/blog/data-analysis/what-is-a-dataset)