TapCub is live — analytics, insights and live chat, free on one platform
Category

Analytics · Page 3

How metrics are defined, where numbers disagree, and how to read a funnel or a retention curve without fooling yourself.

This category covers the mechanics of measurement: what a visitor is, why sessions differ between tools, how to build a funnel that survives scrutiny, and how to spot bot traffic before it skews a decision. Every article uses the same sample dataset you see in the product, so you can reproduce the reasoning in your own workspace.

Articles in Analytics

304 articles · page 3 of 7

Sample data: annual churn computed as 100% minus the top-quartile 12-month retention printed in the ChartMogul SaaS Retention Report (2023); the bars are the best quarter of each segment, not the median.
AnalyticsNov 20, 2026 · 12 min read

Average SaaS Churn Rate: Benchmarks by Contract Size

There is no single average SaaS churn rate. This guide lists what Recurly, ChartMogul and SaaS Capital actually print, converts monthly and annual figures, and walks through a self-check worked example.

Read more
Sample data: of 2,000 customers, shared logins join 55.0%; probabilistic matching raises that to 81.4% (precision 95.8%) at the price of 72 wrong merges.
AnalyticsNov 20, 2026 · 12 min read

AI CDP: What AI Adds to a Customer Data Platform and Where It Stops

An AI CDP adds identity matching, predictions, audiences and assistants to a customer data platform. A capability table with limits, plus two worked examples: match precision and a churn score's break-even.

Read more
Sample data: 3,700 of 12,600 sessions (29.4%) were not human. Counting them all as visitors puts conversion at 2.12% instead of 3.00% for the 267 sign-ups.
AnalyticsNov 20, 2026 · 12 min read

Agent Analytics: How to Tell AI Agents From Human Visitors

Agent analytics separates AI agent requests from human sessions. Four classes of agent traffic, vendor-declared user agents, how to verify them, and a sample day where agents understated conversion by 29%.

Read more
Sample data: for a 4.0% baseline, detecting a 10% relative lift takes 39,475 visitors per group but a 30% lift only 4,783. Halve the lift and you need about four times the traffic.
AnalyticsNov 20, 2026 · 16 min read

A and B Testing: A Worked Test From Sample Size to Result

A and B testing compares two versions of a page with randomly split visitors. One worked test from sample size and duration to reading the confidence interval and deciding what to ship.

Read more
Sample data: a pricing page with 6,000 views and 1.4 tracked clicks per view needs 1,068 clicks for a ±3 point margin at 95% confidence. At 10% sampling that takes more than five weeks.
AnalyticsNov 20, 2026 · 11 min read

Heatmap tools: 8 criteria to compare before you choose

Compare heatmap tools with a scorable checklist: map types, clicks needed for a trustworthy sample, sampling rate, privacy defaults and how the billing meter changes your bill, with worked numbers.

Read more
Sample data: the new tool counts 11.8% more sessions, but the daily ratio stays between 1.110 and 1.134, so the two lines move together and the gap is a fixed offset rather than a broken tag.
AnalyticsNov 20, 2026 · 11 min read

Google Analytics Alternatives: Pick by the Question You Need Answered

Choose a Google Analytics alternative by the question you need answered, not by a ranked list. Documented Google limits, a kind-of-tool matrix, seven tests, and a two-week parallel run with a reconciliation you can recompute.

Read more
Sample data: cap ÷ 4.0 events per session. A 40,000-session site would use 160,000 events a month, 160% of a 100,000-event cap and 32% of a 500,000-event cap.
AnalyticsNov 20, 2026 · 13 min read

Free Website Analytics: What the Free Tiers Actually Include

Free website analytics tiers compared on five numbers: event cap, retention, sampling, bots and scope, taken from public documentation. Includes a cap-to-sessions calculation and a decision table.

Read more
Sample data: a 520 KB page (260 KB shared JavaScript, CSS and font; 260 KB unique HTML and images) viewed 5 times. With shared files cached, the first visit transfers 520 + 4 × 260 = 1,560 KB instead of 2,600 KB, and a return visit 5 × 260 = 1,300 KB.
AnalyticsNov 20, 2026 · 12 min read

Expires Header and Cache-Control: What to Set and How to Check

The Expires header gives the date after which a cached response is stale, but Cache-Control max-age wins when both are present. See the order caches follow, what to set for each file type, how to set it and a curl check.

Read more
Sample data: 12 raw rows become 11 after removing the duplicate E003 and 9 clean rows after rejecting E007 (month 13) and E008 (no user). Rejects are kept in their own table with a reason.
AnalyticsNov 20, 2026 · 13 min read

ETL with SQL: A Runnable Extract-Transform-Load Example

ETL with SQL means extracting with a SELECT, transforming with expressions and window functions, and loading with INSERT ... SELECT and upserts. A complete example you can run, with row counts and money reconciled and a safe rerun.

Read more
Sample data: adding 3,000 one-day installs to 2,000 core users who are active 15 days a month moves the ratio from 50.0% to 22.0%, with no change in core usage.
AnalyticsNov 20, 2026 · 12 min read

DAU/MAU Ratio: How to Calculate Stickiness and What Is Good

DAU/MAU is average daily active users divided by monthly active users, and it equals the average days per month each user is active divided by 30. Recomputed examples, real filings and a ceiling rule for what a good ratio is.

Read more
Sample data: ten session durations (12, 18, 25, 31, 40, 44, 58, 75, 240 and 900 seconds). Two long sessions pull the mean (1,443 ÷ 10 = 144.3 s) far above the median (42 s); eight of the ten sessions are shorter than the mean.
AnalyticsNov 20, 2026 · 11 min read

Data Examples: Types of Data With Real Analytics Rows and What You Can Do With Each

One analytics event row classified into eight data types, a table of what each type allows and what to avoid, and a worked example where the mean (144.3 s) and median (42 s) of the same sessions tell different stories.

Read more
Sample data: five criteria scored 1 to 5 and weighted for a small team (25/20/20/20/15) and a data team (15/10/25/10/40). All-in-one wins for the small team (4.10 vs 3.30); warehouse plus BI wins for the data team (4.10 vs 3.60).
AnalyticsNov 20, 2026 · 11 min read

Data Analytics Platform: Layers, Shapes and How to Choose One

Six layers from collection to governance, three common shapes, and a weighted scorecard whose winner flips with team context, plus the export, retention and volume limits to check before you commit.

Read more
Sample data: sentiment = 100 − 1.5 × error rate − 1.5 × frustration rate − 0.5 × exit rate, each in percent. Pay scores 58.0 (100 − 9 − 10.5 − 22.5). The score is a behavioural proxy from sessions, not a survey.
AnalyticsNov 20, 2026 · 12 min read

Customer Experience vs Customer Journey: How They Differ and Connect

The journey is the path; the experience is how it felt. A stage-to-metric table and a worked five-stage example show how the two views point at different problems and how to use them together.

Read more
Sample data: |r| for the worked examples. One broken-landing-page week (30,000 emails, 20 sign-ups) drops the email correlation from 0.971 to 0.023; the U-shape is an exact relationship with r = 0.000. Anscombe's values are the published 1973 data.
AnalyticsNov 20, 2026 · 11 min read

Correlation Examples: Positive, Negative and Zero, With r Computed

Three small datasets with the raw numbers and r recomputed, plus Anscombe's quartet, an outlier that erases a correlation and a zero r that hides a strong pattern.

Read more
Sample data: 60,000 product-page visitors, 6,000 carts (10.0%), 3,000 checkouts (50.0% of carts) and 1,800 purchases (60.0% of checkouts), 3.0% overall; 70% of carts are abandoned, close to Baymard's 70.22% average.
AnalyticsNov 20, 2026 · 12 min read

CRO (Conversion Rate Optimization): A Step-by-Step Cycle With a Worked Test Plan

Conversion rate optimization as a six-step loop: find the biggest leak in your funnel, rank test ideas by expected extra purchases, compute each test's sample and weeks, and decide with a rule written beforehand.

Read more
Sample data: removing 9,100 known-bot requests (17.5%) leaves 42,900 human-looking requests; the tag counted 38,900, a 4,000 gap (9.3%) caused by blocked scripts, declined consent, cached pages and unidentified bots.
AnalyticsNov 20, 2026 · 11 min read

Website Traffic Analysis Tools: What Each Type Answers

Website traffic analysis tools come in four types: on-site analytics, server logs, search console data and panel estimates. A question-by-type matrix, a recomputed one-month reconciliation and a rule for which number to trust.

Read more
Sample data: a mobile product page with P75 LCP 4.1 s is 1.6 s over a 2.5 s target. The image download and render wait cause 66% of it (1.05 s); the server, often blamed first, causes 19% (0.30 s).
AnalyticsNov 20, 2026 · 13 min read

Website Performance Optimization: Measure First, Then Fix in This Order

Website performance optimization starts with real-user P75 data, splits LCP into four subparts and fixes the largest overshoot first. Worked example, prioritised fix table, INP and CLS rules, and how to verify a fix.

Read more
Sample data: with conversion rates fixed at 4.80% and 5.40%, z crosses the 1.96 line between 10,000 and 14,000 visitors per version. That does not license stopping early; the sample size has to be set in advance.
AnalyticsNov 20, 2026 · 13 min read

Statistical Significance Example: A Full A/B Test Calculation

A complete statistical significance example for an A/B test: counts, pooled rate, z statistic, p-value, confidence interval, sample size and a ship decision, all recomputed.

Read more
Sample data: Social starts the $90 order and never closes one. Judged on last touch it earns $0; judged on first touch it earns $90. Total revenue is $270 under every model.
AnalyticsNov 20, 2026 · 11 min read

Revenue attribution: how five models credit the same three orders

Revenue attribution splits each order's value across the touches before it. Three orders worth $270 credited under first, last, linear, position and time-decay models show which channels start and which close, and how to decide budget from the range.

Read more
Sample data: 1,152 orders at $60.00 = $69,120 gross; 15% returns leave $58,752 net; 48,000 sessions give 2.40% conversion and $1.22 per session.
AnalyticsNov 20, 2026 · 11 min read

Retail Analytics Dashboard: A 10-Metric Layout With Formulas

A retail analytics dashboard layout in three bands with ten metrics, their formulas and one recomputed sample week, plus a way to decompose a bad week into sessions, conversion and order value.

Read more
Sample data with made-up prices ($0.0001 per event after 1M free; $0.02 per user after 10k free): autocapture multiplies the event bill by 3.6 and leaves the user bill flat; a campaign multiplies the user bill by 2.3.
AnalyticsNov 20, 2026 · 12 min read

Product Analytics Tools: What to Compare Before You Pick One

How to compare product analytics tools: a five-row matrix (identity, retention, funnels, session replay, pricing basis), a trial test for each row, a worked funnel-counting example and a recomputed events-versus-users pricing table.

Read more
Sample data: break-even ROAS is 2.5 at a 40% gross margin. The blended 2.83 looks healthy, but three of the four campaigns are below break-even; brand search earns $3,960 and the others lose $2,496 between them.
AnalyticsNov 20, 2026 · 11 min read

PPC Dashboard: The Metrics, Formulas and Layout That Drive Decisions

What a PPC dashboard should show, with Google Ads definitions for CPC, conversion rate and ROAS, a four-campaign worked example, break-even ROAS rules and a layout you can copy.

Read more
Sample data: 40 customers, $100,000 start MRR. NRR is 100.5%; leaving expansion out, GRR is 91.5%. Six new customers added $15,000 and sit outside both metrics.
AnalyticsNov 20, 2026 · 11 min read

NRR vs GRR: net and gross revenue retention on one worked cohort

NRR counts expansion, GRR does not. One 40-customer cohort recomputed under both formulas shows how a 100.5% NRR can hide a 91.5% GRR, why one account can swing it, and why monthly figures must be compounded.

Read more
Sample data for an invoice-reminder idea. Thresholds were written before launch; the small counts are directional, not proof.
AnalyticsNov 20, 2026 · 12 min read

MVP Test: How to Validate an MVP With Real Users, Step by Step

How to run an MVP test with real users: pick the riskiest assumption, choose a test type, write a plan with hypothesis, metric, threshold and decision, and read small samples with confidence intervals.

Read more
Sample data: cost per link stayed at $300 in both months, but cost per goal conversion fell 16% because the visits converted better (2.35% to 2.97%).
AnalyticsNov 20, 2026 · 11 min read

Link Building Report: Metrics to Include and a Template With Sample Numbers

What a link building report should contain: a filled-in monthly template (links won, referring domains, referral sessions, conversions, cost), the limits of Google's Links report and how to label paid links.

Read more
Sample data: a model with traffic proportional to 1 ÷ rank. One million domains, 0.34% of the 291,965,818 in Netcraft's December 2025 survey, hold about 72% of visits; the median domain gets about 548 visits a month on an assumed 80 billion for rank 1.
AnalyticsNov 20, 2026 · 11 min read

How Many Visitors Does a Website Get? A Rank-to-Traffic Model

Website traffic follows a steep curve: 96.55% of pages get no Google traffic while a few sites get billions of visits. A rank-to-traffic model, the noise at each size, and how to count your own visitors.

Read more
Sample data: 2,000 visitors. Pooled, viewers convert 10.6 points lower (150/1,000 vs 256/1,000); inside each visitor type they convert higher (+2.0 and +5.0 points) because viewers are mostly new visitors. Standardized to the same mix: 22.5% vs 19.0%.
AnalyticsNov 20, 2026 · 12 min read

Confounding vs Lurking Variable: The Difference, With Worked Examples

A confounding variable is a measured third factor; a lurking variable is the same thing nobody measured. Two sample-data tables and the 1973 Berkeley admissions counts show how each one reverses or shrinks a headline gap.

Read more
Sample data: 10,000 users; 3,200 tried the feature. Retention is 47.5% for users and 25.6% for non-users, but engaged users both adopt and stay. Within each group the gap is 2 points, so the raw gap overstates the effect 10.9 times.
AnalyticsNov 20, 2026 · 11 min read

Causation Examples: What Evidence Separates a Cause From a Coincidence

Examples of causation from a randomized hormone trial, Bing's headline experiment and Facebook's ad studies, plus a worked product-analytics case where a 21.9-point retention gap is really 2 points, and a checklist for what you can claim.

Read more
Sample data: control 500/10,000 against five variants of 10,000 visitors each. Bonferroni requires p below 0.01, so only A passes; Holm also keeps B (0.0123 is below 0.0125). Bars are scaled to 1 ÷ p for readability.
AnalyticsNov 20, 2026 · 11 min read

Bonferroni Correction Example: Five Variants and Adjusted Alpha

A worked Bonferroni correction example with a control and five variants: raw versus adjusted p-values, the 0.01 threshold, the extra traffic it costs, and how Holm keeps one more winner.

Read more
Sample data: percentages show the rate versus the previous step. The whole funnel converts 0.03% of visits to customers, and 1.5% of leads.
AnalyticsNov 20, 2026 · 11 min read

B2B Marketing Analytics: Metrics by Funnel Stage, With Formulas and Sample Numbers

B2B marketing analytics from visit to customer: one funnel table with formulas and sample numbers, CAC and payback, sales velocity, account-level metrics, attribution across a long cycle and the cohort trap in monthly reports.

Read more
Sample data: the correct total for seven customers is $24,138.40. Counting multi-year contract value as annual adds $18,000, or 74.6%; all five mistakes together give $47,376.00.
AnalyticsNov 20, 2026 · 11 min read

ARR Means Annual Recurring Revenue: How to Calculate It With Examples

ARR is the yearly value of the subscription revenue you expect to repeat. A seven-customer worked example handles discounts, monthly versus annual plans, multi-year deals and churn, and shows how each wrong choice inflates the number.

Read more
Sample data: 12,000 sessions and 244 conversions over 30 days. Social sends 25.0% of sessions but 6.1% of conversions, so its quality index is 0.50% ÷ 2.03% = 0.25.
AnalyticsNov 19, 2026 · 12 min read

Website Traffic Analysis: A 3-Pass Method With a Worked Example

How to analyse website traffic in three passes: split visits by source, judge each source by engagement and conversion, then compare periods to see what moved. One sample you can recompute, a weekly routine and common mistakes.

Read more
Sample data: five audit findings. Extra conversions = affected sessions × baseline conversion × expected lift, divided by days of effort. The first two fixes give 45.3 of 69.7 monthly conversions for 1.5 of 16.5 days. The lifts are assumptions.
AnalyticsNov 19, 2026 · 11 min read

How to Audit Your Website: Technical, Content and Conversion Checks

How to audit your website in three layers, technical, content and conversion, then rank findings by expected extra conversions per day of work. Scope rules, a check table for each layer and a worked prioritisation example.

Read more
Sample data: the share of usability problems found as participants increase, using L = 0.31 (typical value reported by Nielsen Norman Group). Going from five to fifteen adds about 15 points at three times the participants.
AnalyticsNov 19, 2026 · 12 min read

UX Testing: Which Method to Use, With How Many People

UX testing means watching real people attempt real tasks. Learn the three axes of methods, how many participants find problems versus measure them, and how to combine tests with behaviour data.

Read more
Sample data: stage rates 25.0%, 40.0% and 20.0% multiply to a 2.0% lead-to-win rate; at $6,000 per deal, 20 wins are $120,000, or 80% of a $150,000 quota.
AnalyticsNov 19, 2026 · 11 min read

Sales Dashboard: The Four Metric Groups That Drive Action

What to put on a sales dashboard: four groups of metrics (leads, conversion, cycle, amount), formulas, a recomputable monthly example, a pipeline coverage calculation and views for rep, manager and executive.

Read more
Sample data: 4.0% of visitors sign up, 40.0% of sign-ups activate and 20.0% of the activated pay. New MRR is 64 × $50 = $3,200, yet net revenue retention is 98.5%, so the existing base is shrinking.
AnalyticsNov 19, 2026 · 11 min read

SaaS Analytics: Read Acquisition, Activation, Retention and Revenue in Order

SaaS analytics measures a subscription business in four stages with a few numbers each. One recomputable month of data, a cohort retention curve, net and gross revenue retention, and how to join product and billing data.

Read more
Sample data: at week 8 of 12, 66.7% of the time has passed but only 55.6% of the tasks are done, while 47.5% of the budget is used. At 1.875 tasks per week the project finishes in week 14.4.
AnalyticsNov 19, 2026 · 12 min read

Progress Report: How to Write One People Read in a Minute

A progress report is a short update on work in progress. Put the status, decision and risk first, then the numbers. See a worked example at week 8 of 12 and a template you can copy.

Read more
Sample data: labour at $75 an hour. Without labour the order reverses: self-hosted $14,400, open core $38,700, commercial $40,500.
AnalyticsNov 19, 2026 · 12 min read

Open Source BI: The Hidden Costs of Sources, Permissions and Operations

Open source BI fits teams that have someone to maintain it. A ten-capability checklist, the hidden costs of data sources, permissions and operations, and a three-year cost that flips depending on whose hours are counted.

Read more
Sample data: self-hosting costs $11,100 (setup $1,200 plus $3,300 a year in server, storage and 3 maintenance hours a month) against a $150-a-month hosted plan. Break-even is (150 − 50) ÷ 75 = 1.3 hours a month.
AnalyticsNov 19, 2026 · 11 min read

Open Source Analytics: Licence, Hosting and Data Protection Costs

Open source analytics saves licence fees but moves deployment, upgrades and data protection to you. A recomputable self-hosted vs hosted cost with a break-even, licence checks, GDPR duties and a five-step project check.

Read more
Sample data: revenue rose 5.6% (1.20 × 0.88 × 1.00), but the conversion rate fell and the error rate more than doubled. At a held 2.5% conversion, the same traffic would have made $72,000.
AnalyticsNov 19, 2026 · 10 min read

Metrics Dashboard: Arrange Metrics in Three Layers, Not One Wall

How to design a metrics dashboard: one north star, three to five input metrics that add up to it, and guardrails with thresholds. A recomputable promotion-week example, layout rules and common mistakes.

Read more
Sample data: the pricing page's P75 LCP of 4.1 s splits into 0.9 + 0.8 + 1.7 + 0.7 seconds. Fixing the three highlighted parts (0.2 + 0.9 + 0.4) brings it to 2.4 s without touching the server.
AnalyticsNov 19, 2026 · 11 min read

How to Optimize Website Performance: Measure, Find the Slow Part, Fix It

How to optimize website performance in a fixed order: measure real users at P75, rank pages, split LCP into four parts, fix only the part that is out of proportion and verify. Recomputed example included.

Read more
Sample data: 300 people made 420 add-to-cart entries and 90 ordered, 60 within a day. The rate ranges from 14.3% to 30.0% depending on counting people or attempts and on the window length.
AnalyticsNov 19, 2026 · 11 min read

Funnel Analytics Tools: Criteria That Change Your Funnel Numbers

How to choose funnel analytics tools by three capabilities: flexible step definitions, group splits and drill-down. A criteria table, two recomputed examples and a 14-day trial plan.

Read more
Sample data: revenue fell from $60,000 to $58,880. Changing one factor at a time in the order traffic, conversion, AOV gives +$9,000, −$13,800 and +$3,680, which sum to −$1,120.
AnalyticsNov 19, 2026 · 12 min read

Ecommerce Data Analytics: Four Metric Families That Explain Revenue

Ecommerce data analytics comes down to traffic, conversion, basket size and repeat purchase. See how they multiply into revenue, with a sample revenue drop split by factor and channel.

Read more
Sample data: score = times per year × value per decision × data readiness, e.g. 52 × $2,000 × 0.8 = $83,200. The value per decision is an estimate; the score is a heuristic for ranking, not a measurement.
AnalyticsNov 19, 2026 · 11 min read

Data Analytics Strategy: A One-Page Plan Built Around Decisions

A data analytics strategy answers which decisions analytics serves, who owns each number and how you know it works. See a scored decision list, owner roles and a decision log you can copy.

Read more
Sample data: three anonymous options scored 0 to 5 on five criteria. With the first weights C wins (3.95); with the second B wins (4.05), so agree the weights before you see the scores.
AnalyticsNov 19, 2026 · 12 min read

Customer Analytics Software: What to Check Before You Choose

Customer analytics software works on people, not page views. Check identity merge, segmentation and drill-down to one person, and see how weights can flip a scoring result.

Read more
Sample data: Post A has 12,000 sessions and 96 signups; Post B has 3,000 sessions and 90 signups. Which is better depends on each post's purpose, so judge each against its own primary metric.
AnalyticsNov 19, 2026 · 11 min read

Content Metrics: Which Numbers to Use for Which Purpose

Content metrics fall into four purposes: reach, engagement, conversion and retention. Learn to pick one primary metric per piece, with a worked example of two posts that each win a different purpose.

Read more
Sample data: licence, setup, upkeep and infrastructure over 36 months at $75 an hour. The option with no licence fee costs $59,400 because upkeep is 73% of its total.
AnalyticsNov 19, 2026 · 11 min read

Business Analytics Software: How to Choose With 4 Criteria and a Scorecard

How to choose business analytics software on four criteria: data connection, modeling, collaboration and three-year cost. A recomputable cost table, a weighted scorecard that flips with the weights, and a two-week trial.

Read more
Sample data: 10,000 sessions, 1,800 of them bots, with 164 conversions all from people. Counting bots, the conversion rate is 1.64% instead of 2.00%, and the bounce rate is 59.7% instead of 52.0%.
AnalyticsNov 19, 2026 · 12 min read

Website Traffic Statistics: How to Read the Numbers Behind Visitors, Sessions and Bots

Website traffic statistics depend on definitions: what counts as a visitor, how a session is cut, and whether bots are in. Worked examples on daily sums, bot inflation, session timeouts and noise, plus a test for published statistics.

Read more

See your data clearly. Find your growth.

Every click, backed by data. Install one line of code and see your first numbers in a minute.

No credit card · Free plans for analytics and chat · Cookieless analytics