Collecting data is the easy half. I’ve watched teams hoard terabytes and still argue about basic questions in meetings. Because collecting and UNDERSTANDING are different skills.
This category covers the understanding half. The analytics disciplines that turn stored records into answers, and the big-data concepts that kick in when the volume gets serious. Term by term 👇
30-Second Summary
Analytics is the practice of extracting answers from data, and big data is what happens when the volume, speed, or variety outgrows ordinary tools. These terms cover both: the craft of analysis and the scale game around it.
What This Category Covers
- The core craft: data analytics, data mining, and data exploration
- The scale game: big data and big data analytics
- The business lens: business intelligence and contextual information delivery
📌 Quick take: Analytics maturity isn't about tool count. It's about how short the path is from question to defensible answer.
The Terms in This Category, Walked Through
Begin with data analytics. It’s the umbrella practice of turning stored records into answered questions. Four types ladder up: descriptive, diagnostic, predictive, prescriptive. The page’s workflow arc runs frame → assemble → analyze → communicate → close the loop. And its trap list (selection effects, correlation theater, Simpson’s paradox) applies to every other term in this folder.
Before any analysis deserves belief, the data deserves a look. Data exploration covers that first-look discipline. Profiling, distribution walks, oddity hunting. It comes with a working checklist and the classic horror stories (sentinel dates, meaning shifts, duplicate eras) that make the case better than any argument. Its industrial cousin data mining turns discovery into method. Clustering, association, anomaly detection. Plus the validation discipline that separates found patterns from hallucinated ones.
Then the scale pair. Big data defines the threshold where volume, velocity, or variety break ordinary tools. You get the history of how distributed machinery became boring infrastructure. And the myth-retirement section every executive conversation needs. Big data analytics covers what analysis becomes at that scale. Samples optional. Significance useless. Exploration industrialized, and cost part of the method.
Business intelligence is analytics domesticated, the recurring questions industrialized into governed dashboards and metrics. The page maps the stack from source to Monday meeting. Top billing goes to the semantic layer. It’s the highest-payoff investment. And it catalogs the failure patterns (dashboard graveyards, definitional civil wars, trust spirals) with their organizational repairs.
And contextual information delivery is where the whole category is heading. Insight embedded where decisions actually happen. Not waiting in reports nobody visits. The patterns (embedded panels, event pushes, in-flow enrichment) and the build-out sequence (one decision moment at a time, measured by decisions changed) close the folder.
How to Read This Category
The dependency order is real. Exploration before analysis. Analysis before scale. Governed BI before conversational interfaces. Skip a step and you automate your confusion at whatever level you skipped to.
And every page repeats one dependency deliberately: analytical quality is capped by record quality underneath. The analysis of half-empty fields is astrology with SQL. At every scale, in every tool.
Questions This Category Answers
“Can we trust this number?” Trace it backward. The BI page’s stack map shows where dashboard numbers are actually made, and the exploration checklist finds the input problems that aggregates hide.
“What should we look into?” When nobody knows the question yet, mining earns its keep. Clusters, associations, and anomalies surface the questions worth asking. And its validation discipline keeps found patterns honest.
“Why do two teams report different revenue?” That’s the BI page’s definitional-civil-war section. Metrics live in tools and heads. Not in one governed semantic layer. The repair is organizational, and the page names it.
“Do we need big data tools?” The big data page’s practical threshold (“big” begins where YOUR tools stop coping) plus the scale analytics page’s honest prerequisites answer it without the vendor gravity.
“Why does our analysis keep being wrong?” Usually one of the analytics page’s five traps. Selection effects, correlation theater, metric drift, Simpson’s paradox, or dashboards mistaken for analysis. Pin the trap list above your desk.
Read as a set, these seven terms form the literacy layer of a data practice. How answers get made, at any scale, and how they go wrong. Everything downstream (models, decisions, strategy) inherits whatever this layer gets right.
How This Category Connects to the Rest of the Wiki
Analytics consumes what every other category produces. The records it examines were shaped by enrichment and quality work. They’re delivered by pipelines at freshness the flow folder’s tiers define, from systems the architecture pages map, on platforms the cloud folder prices. So when an analysis disappoints, the cause usually lives in one of those upstream folders. Which is exactly why the exploration and BI pages here keep pointing at them.
Downstream, this folder feeds two neighbors directly. Advanced Analytics & ML takes analysis into prediction and presentation. Business Metrics applies the craft to the specific numbers revenue teams live by. Read those folders as this one’s specializations. Same discipline, narrower aim.
The dependency chain also gives you a debugging order for any “the numbers look wrong” moment. Check the flow first. Quality second. Definitions third. And only then the analysis itself. Decades of collective experience compress into that ordering. The analysis is the last suspect, not the first.
Start Here If You’re New
New to analytics vocabulary? Here’s your reading order. Data analytics first, the umbrella and its four types. Then data exploration, the habit that keeps everything honest. Then business intelligence, which shows how answers become infrastructure. Save the big-data pair for when scale questions actually arrive, and contextual delivery for when your dashboards work but nobody visits them.
And a promise the whole folder stands behind: none of these pages requires a statistics degree. They require curiosity, skepticism toward your own conclusions, and the patience to look at the data before believing anything about it. Those three travel further in analytics than any credential.
One closing calibration. The analytics field ships a new must-have tool roughly quarterly, and the terms in this folder outlive every one of them. So concepts before platforms is the durable investment. A team fluent in exploration, honest analysis, and governed definitions can adopt any tool in weeks. But no tool has ever installed those fluencies in a team that lacked them.
Frequently Asked Questions
What is data analytics?
Data analytics is the practice of turning stored records into answered questions. It ladders through four types (descriptive, diagnostic, predictive, prescriptive), each answering a deeper question than the last. The craft is the same at every scale; only the tools change.
What makes data “big data”?
Data becomes “big” when its volume, velocity, or variety breaks the tools you already have. There’s no fixed terabyte line. The practical threshold is where YOUR machinery stops coping, and that’s when distributed storage and parallel processing earn their complexity.
What’s the difference between business intelligence and data analytics?
Analytics is the craft of answering questions; business intelligence is that craft industrialized into governed dashboards and metrics. BI takes the questions a business asks every week and turns them into infrastructure. Analytics handles the new questions BI hasn’t automated yet.
What is data mining used for?
Data mining finds patterns you didn’t know to look for: clusters, associations, and anomalies that surface the questions worth asking. It shines when nobody knows the question yet. The validation discipline matters as much as the discovery, because found patterns can be hallucinated ones.