Guides
Analytics, explained.
How tracking really works, why your tools disagree, and what to do about it. Written by the people who build Datacenter, with sources you can check.
Why your analytics numbers never matchGA4, Shopify, Meta and Google Ads count with different rules, dates and blind spots. Where each number splits, and which one to trust for each question.UTM parameters: the complete guide, with a naming scheme that holds upWhat each UTM parameter does, the naming rules GA4 actually applies, a scheme for source, medium and campaign, and the mistakes that break your reports.Attribution models explained, with one buyer's path worked throughFirst touch, last click, linear, position-based and data-driven attribution, worked through on one buyer path. What each model is good for, and its limits.
Tracking accuracy
Why tools disagree, and which number to trust.
Why your analytics numbers never matchGA4, Shopify, Meta and Google Ads count with different rules, dates and blind spots. Where each number splits, and which one to trust for each question.Meta ads vs Shopify: why purchases don't matchMeta Ads Manager shows more (or fewer) purchases than Shopify. The six causes, how to measure your own gap, and which number to use for budget calls.GA4 not tracking conversions: a checklist that finds the causeKey events missing in GA4? Work through ten checks in order, from the event never firing to consent, filters, processing delays and Google Ads imports.Direct traffic: what (direct) / (none) really meansDirect traffic is every visit that arrived with no referrer and no tag. Where it really comes from, why it grows, and five ways to shrink it.GA4 "(not set)": every cause and how to fix itWhy GA4 shows (not set) for landing page, source, medium and campaign, how to tell which cause you have, and the fix for each one.Bot traffic in analytics: how to spot it and filter itHow GA4 and other tools filter bots, the patterns that give bots away in your reports, and what you can do about the bots that get through.Cross-domain tracking: checkouts, booking tools and iframesWhy a visitor who moves to a checkout, a booking tool or an embedded form loses their source, and how to keep the credit with the ad that brought them.Leaving GA4 without losing your historyWhat you can take out of Google Analytics 4, what you cannot, the Data API limits, and why imported days never match a new tool's own numbers.Unassigned traffic in GA4: causes and fixesGA4 labels a session Unassigned when its source and medium match no channel rule. The rules, the tag values that fall through, and how to fix a spike.
Campaign tracking
UTMs, short links, QR codes and click ids.
UTM parameters: the complete guide, with a naming scheme that holds upWhat each UTM parameter does, the naming rules GA4 actually applies, a scheme for source, medium and campaign, and the mistakes that break your reports.How to track your Instagram link in bio, through to the saleInstagram's own insights stop at link taps. How to see which bio link, story or post brought visits, signups and sales, with tags and a short link.gclid, fbclid and other click ids, explainedWhat gclid, gbraid, wbraid, fbclid, ttclid and msclkid are, what each one is for, how long they last, and why keeping them matters for ad tracking.How to track QR codes on print, packaging and signsA QR code sends no referrer, so scans show up as direct traffic. How to track scans, visits and sales per placement, and print codes that keep working.How to track email clicks and the sales they bringTag email links so every click and sale has a source, set it up in Klaviyo and Mailchimp, and spot the bot clicks that security scanners add.How to track SMS link clicks and the sales they bringTag text message links so clicks and sales have a source, keep links short and on your own domain, and tell real taps from link preview clicks.Call tracking for local businesses: which ad made the phone ringHow call tracking numbers work, static numbers versus dynamic number insertion, how many you need, recording laws, and calls as conversions.
Attribution
Which ad, post or email earned the sale.
Attribution models explained, with one buyer's path worked throughFirst touch, last click, linear, position-based and data-driven attribution, worked through on one buyer path. What each model is good for, and its limits.Server-side vs client-side tracking: what each one seesClient-side tags run in the browser and miss what they cannot load. Server events come from your server. What each sees, costs, and when to use both.Meta Conversions API explained: setup, matching and deduplicationWhat the Meta Conversions API sends, which fields Meta needs, how deduplication with the pixel works, and the mistakes that double count or lose purchases.ROAS, blended ROAS and MER: which one to run the business onFormulas for ROAS, blended ROAS, MER and break-even ROAS, a worked example, and why adding up each platform's ROAS overstates what your ads earn.Google Ads conversion tracking for lead generation sitesTrack forms, calls and the leads that become customers in Google Ads, with the right counting, values and windows so bidding learns from real sales.Offline conversions and enhanced conversions explainedHow Google Ads and Meta accept sales and leads that happen off your website, how hashed email matching works, and the rules that make uploads fail.TikTok ads attribution: pixel, Events API and why numbers differHow TikTok credits conversions, why TikTok Ads Manager reports more sales than GA4 or your store, and how the pixel and Events API work together.Lead-to-sale reporting: connect marketing sources to CRM revenueHow to carry each lead's source into HubSpot or GoHighLevel, report won revenue by campaign, and avoid the mistakes that leave deals with no source.MRR and SaaS attribution: which channels bring revenue that staysHow MRR is calculated, what moves it, how churn is measured, and how to tie subscription revenue to the campaigns that brought each customer.Klaviyo attribution windows vs your analyticsKlaviyo credits a sale to an email opened or clicked in the last 5 days. Why it runs above GA4 and Shopify, what to change, and which number to report.Break-even ROAS and target MER: the formulas, with a margin tableBreak-even ROAS is 1 ÷ margin; target MER adds fixed costs and profit. Formulas, a lookup table by margin and worked examples you can copy.Meta Ads cost per lead and signups: what Meta counts, and what to payHow Meta counts leads and signups, why Ads Manager's cost per lead differs from your CRM, and how to work out the most a lead is worth to you.
Privacy and consent
Cookies, consent and what the law asks.
Cookieless analytics and GDPR: what needs consent, and what usually does notDo you need a cookie banner for analytics? How ePrivacy, the GDPR, the CNIL exemption and the UK change of 2026 apply to cookieless and cookie tracking.iOS and Safari tracking changes, version by versionWhat Apple's Link Tracking Protection, ITP, App Tracking Transparency and Safari 26 remove, where each applies, and what still works for ad tracking.First-party data for small businesses: what to collect and whyThe first-party data worth collecting, from orders and emails to how did you hear, what each is for, what to leave out, and the consent rules that apply.
Metrics explained
What each number means, and how it is counted.
Bounce rate in GA4: what it means now, and when to worryGA4 defines bounce rate as the share of sessions that were not engaged. The rule, the formula, how it differs from the old one, and when it matters.Visitors, sessions and pageviews: how each tool counts themWhat a user, a session, a visit and a pageview are in GA4, Shopify and privacy-first tools, why the counts differ, and which one to report.Average session duration and time on page: how each tool measures itWhy GA4, Universal Analytics and privacy tools report different session durations and time on page, the method each uses, and which to trust.How to set up goals and conversions without a developerFour ways to count sales, leads, bookings and signups without writing code, how to do each in GA4, and the checks that keep a goal honest.Funnel analysis: find the step where people leaveHow a conversion funnel is counted, open versus closed funnels, time windows, which drop-offs are normal, and how to read one without fooling yourself.Cohorts and retention: do the people you win come back?How a cohort table is built, how to read it by row, column and diagonal, three ways to count retention, and how to compare sources by customer.Ask your analytics in plain words: MCP and AI assistantsWhat the Model Context Protocol is, how an AI assistant reads your analytics through it, what it answers well, and the safety rules for keys.
Dashboards
What to put on a board, by business.
Ecommerce dashboard template: the cards that matter, and who counts eachAn ecommerce dashboard template card by card: revenue, orders, conversion rate and MER first, which tool should count each number, and a worked month.Lead generation dashboard: what to show, and where each lead is countedA lead gen dashboard template: leads, lead rate, cost per lead and qualified leads by source, which tool counts each, and how to read a month.SaaS metrics dashboard: signups, MRR and the source of each customerA SaaS metrics dashboard template: signups, cost per signup, MRR and its moves, lifetime value by source, plus MRR vs ARR and churn vs retention.Marketing dashboards for agencies: one for you, one for each clientAn agency needs two dashboards: a portfolio view that ranks clients by what needs attention, and a client board with three numbers and a target.Year over year comparison in analytics: same dates or same weekdaysYoY growth formula, when to compare 364 days back so weekdays line up, partial months, moving holidays, and what breaks a comparison with last year.
Agency reporting
Reports clients read, and trust.
Client reporting for agencies: a template clients actually readWhat to put in a monthly client report, how often to send it, which numbers to lead with for stores and lead-gen clients, and how to explain gaps.Uptime monitoring for agency client sitesWhat to monitor on every client site, how intervals and retries trade speed for false alarms, SSL and domain expiry, and reporting uptime.AI summaries in client reports, without made-up numbersHow to use AI to write the summary at the top of a client report: what it should say, why models invent numbers, and a checking method that catches it.
How we write these guides
- From the people who build Datacenter. We write tracking code for a living, so we explain how it works, not just what to click.
- Sourced. Platform rules link to the platform's own help pages, with the date we checked them.
- No made-up numbers. Examples with numbers say they are examples. We never invent studies, quotes or customers.
- Fair to other tools. Where Google, Meta or Shopify does something well, we say so.
- Kept current. Each guide shows the date its words last changed. Found a mistake? Email hello@datacenter.so.