Written for business owners and marketers who want to understand the mechanics of accurate conversion tracking, Google Ads and CRM automation without the jargon.
Clicks and conversions that never turn into real sales conversations usually trace back to intent, targeting and how a conversion is defined. Here is how to find and fix the cause.
A practical audit checklist covering campaign structure, search terms, conversion actions, bidding, targeting and landing pages so you know what to fix first.
Keywords are what you target and search terms are what you pay for. Understanding the gap between them is the fastest way to find wasted ad spend.
The actions you set as primary directly drive smart bidding. Choosing the wrong ones teaches the algorithm to buy the wrong traffic. Here is how to get it right.
Smart bidding trusts your conversion data. When tracking is wrong, every bid and targeting decision built on it is wrong too. Here is what to check first.
Both can bring local customers but they work very differently. This comparison explains how each works and when a local business might use one, the other or both.
When LSA volume drops or never builds, the cause is usually in the profile setup, categories or lead handling. Here are the most common reasons and what to check.
LSA leads only matter if you can see what happened next. This guide explains how to track calls and messages from first contact through to a booked and completed job.
A tracking audit checklist covering conversion actions, GTM, Google Ads tags, GA4 alignment, form and call tracking, enhanced conversions, duplicate events and tag testing.
Google Ads and your CRM often count different things. The gap between them points to where your funnel is leaking. Here is how to find and reconcile the difference.
A form submission is the middle of the journey, not the end. This guide explains how to connect the ad click to the closed customer using offline conversion tracking.
Sending qualified lead data back to Google Ads is straightforward once you capture the click identifier and map your stages. This guide covers the pieces and the common mistakes.
GTM is a container that lets you manage tracking code without editing your site every time. This guide explains what it is, what it does and what it does not do.
A single wrong trigger can quietly corrupt your conversion data for weeks. This diagnostic guide covers the GTM mistakes I look for when numbers do not add up.
GA4 can track almost anything, which is why many accounts track too much. This guide covers what a lead or ecommerce business actually needs to measure and what to leave alone.
It is normal for GA4 and Google Ads to show different conversion numbers. They are different platforms with different rules. This guide explains the common reasons for the gap.
The Pixel and Conversions API are two paths for the same job, not rivals. This guide explains how each works and why they are often used together.
If you rely only on the Meta Pixel, you are almost certainly undercounting conversions. This guide explains why browser only tracking misses data and how to recover it ethically.
Shopify stores need tracking across Google Ads, GA4 and Meta. This checklist covers what each platform needs and how to keep the data clean across all of them.
Duplicate purchase events inflate your counts and confuse your bidding algorithms. This guide explains why they happen and how to diagnose the source before changing anything.
Most lead loss happens after the form, in the gap between a new lead arriving and someone working it. A spreadsheet is not a pipeline and a CRM is not a fix on its own.
A pipeline your team will actually use is a simple one. This is a practical structure you can set up in an afternoon and refine as you learn.
A Google Ads click is only the start of the story. This is how to follow a lead from the click to the close inside your CRM and where the attribution gets hard.
GoHighLevel gives you contacts, opportunities, pipelines and workflows in one place. A practical structure keeps manual and automated actions organized so the team always knows what the system is doing.
Speed and persistence win lead follow-up. GoHighLevel workflows can handle both automatically, as long as they stop when a lead engages and respect consent.
When leads go missing in GoHighLevel, the cause is almost always in the pipeline setup or the workflow configuration. This is a diagnostic guide for the most common reasons leads disappear.
n8n, Zapier and Make all do the same job in different ways. The question is not which is best overall, it is which fits the kind of automation you actually need to build.
Automation does not replace the sales team, it removes the repetitive work that keeps them from selling. These ten workflows cover the moments where a small automated action saves a manual one.
A lead generation system is not one tool, it is a chain. This is how Google Ads, forms, CRM and automation connect and where the links usually break.
Automation fails for predictable reasons: wrong triggers, bad data or workflows too complicated to maintain. This is a guide to the common failure points and how to build automation that is easier to test and trust.
When a GoHighLevel workflow is active but a contact never enters it, the cause is usually the trigger, its filters or the contact data. This guide walks through how to trace a contact through each condition and find where it stops.
When GoHighLevel emails stop going out, the cause is usually in the email configuration, the contact record or a workflow condition, not the workflow itself. Here is how to diagnose it step by step.
SMS failures in GoHighLevel usually come from phone number setup, carrier filtering, DND settings or workflow conditions. This guide covers what to check before you rebuild the workflow.
Duplicate contacts break follow-up, inflate reporting and confuse automation. This guide explains why duplicates appear in GoHighLevel and how to prevent them at the source.
When opportunities stay stuck in the wrong stage, the cause is usually a workflow trigger, a filter or a manual process gap. This guide covers how to diagnose pipeline updates that do not fire.
Calendar booking problems usually come from availability rules, time zones or event settings. This guide covers how to diagnose a calendar that shows the wrong slots or rejects bookings.
When appointment reminders stop going out, the cause is usually the reminder workflow, the calendar event or contact consent. This guide covers how to trace a reminder that never fires.
When a form submission does not create a contact or an opportunity, the gap is usually in the form settings, the workflow or the field mapping. This guide covers how to trace a form that silently fails.
When Facebook Lead Ads leads stop appearing in GoHighLevel, the cause is usually the integration connection, the field mapping or the workflow trigger. This guide covers how to find where the leads drop.
Losing Google Ads lead source data in GoHighLevel usually comes from missing GCLID capture, broken field mapping or no attribution fields. This guide covers how to keep campaign data intact.
When lead source data disappears in GoHighLevel, the cause is usually in how UTM parameters travel from the landing page to the form to the contact. This guide covers how to keep source data intact.
A webhook failure can be inside GoHighLevel or at the receiving endpoint. This guide covers how to use the workflow history and the HTTP response to find which side is failing.
Duplicate messages almost always come from duplicate triggers, not from a single workflow sending twice. This guide covers how to find the real source before changing anything.
When leads are not assigned to the right user, the cause is usually the assignment action, the team setup or the trigger conditions. This guide covers contact owner versus opportunity owner too.
A messy CRM builds up slowly through extra stages, duplicate pipelines and overlapping workflows. This guide covers a controlled cleanup that validates and backs up before removing anything.
When a contact field does not update, the first step is to find out which system is supposed to update it. This guide covers field mapping, duplicate contacts and empty value overwrites.
When one contact does not enter a workflow that works for everyone else, the problem is that contact's eligibility. This guide covers missing data, prior enrollment, DND and pipeline conditions.
A workflow that starts but stops before finishing leaves a clear trail in the history. This guide covers wait steps, condition exits, failed actions and missing data for later steps.
When if/else branches send contacts the wrong way, the conditions are usually the cause. This guide covers wrong operators, empty fields, AND or OR logic and data type mismatches.
A workflow that looks stuck at a wait step is usually waiting as configured. This guide covers duration, specific time, time zone, business hours and event-based waits.
A workflow that does not run for the contacts you expect usually comes from confusing the trigger with the filter. This guide explains the difference and how to diagnose which one is wrong.
Re-entry lets a contact run through a workflow more than once. It is useful for repeating events but can create duplicate messages when a trigger fires repeatedly. Here is how to control it.
Workflow history is the fastest way to diagnose an automation problem. This guide shows how to read it to find whether a contact entered, where they stopped and whether an action failed.
If/else branches that look correct in the builder can still route contacts the wrong way. This guide covers operator choice, empty fields, AND or OR logic and how to test both paths.
Continuing automated follow-up after a lead replies feels careless. This guide covers how to detect a reply, stop the sequence and hand off to a human without duplicate messages.
Emails landing in spam rarely have one cause. This guide covers domain authentication, sender reputation, content, list quality and engagement so you can find which area is weakest.
Deliverability is not one problem. An email can fail to send, bounce, land in spam or be delivered but not found. This guide helps you find which failure point you have before you change anything.
SPF, DKIM and DMARC prove your email is genuine. This guide explains what each record does, the common mistakes and how to set them up without breaking your sending.
A delivered status means the receiving server accepted the message, not that it appeared in the inbox. This guide covers spam, filtered folders, recipient rules and stale addresses.
An email sent at the wrong time usually comes from a time zone mismatch, a wait set to a specific time or a scheduled action in an unexpected time zone. Here is how to diagnose it.
An SMS sent at the wrong time usually comes from a time zone mismatch, a wait set to a specific time, a scheduled action or a business hours restriction. Here is how to find the cause.
A missing SMS reply is usually attached to an unexpected contact or conversation, not lost. Search the inbound number across all contacts to find where it landed.
Continuing automated SMS after a lead replies feels careless. Detect the reply inside the workflow, stop the sequence before the next wait completes and hand off to a human.
Phone number problems usually come from assignment, capability, routing or workflow configuration. Confirm the number is active, what it supports and how it routes inbound activity.
A missed call text back is a chain from the missed call to the contact to the automation to the message. Trace the contact through each step to find where it stops.
A calendar showing no available times usually has an availability rule blocking every slot. Check working hours, blocked dates, minimum scheduling notice and the booking window.
Double booking usually comes from how availability is calculated. Identify whether the overlap is two bookings or a booking and a busy event before changing availability.
A wrong appointment time is usually a mismatch between the business, calendar, team member and client time zones. The appointment is often correct and the display is what disagrees.
Google Calendar sync has four layers that can fail independently: the connection, the calendar selection, availability sync and event sync. Find which layer is broken before reconnecting.
A Google Meet link is created on the Google Calendar event, not by GoHighLevel directly. Confirm the booking became an event and conferencing is configured to add Meet.
A missing Zoom link usually means the appointment was created but the Zoom connection or calendar configuration is incomplete. Here is how to find the cause.
New, Confirmed, Cancelled and No Show each describe a distinct state of a booking and can trigger different workflow automation. Here is what each one means.
An appointment workflow that does not trigger usually means the booking landed on the wrong calendar, the filters exclude the contact or the workflow is unpublished.
Reschedule and cancellation problems usually come from the appointment status not updating, the workflow trigger not reading the change or the connected calendar not syncing.
A no-show follow-up workflow triggers on the No Show status, sends timed messages with a rebooking opportunity and stops when the contact replies or books again.
A disappeared opportunity is usually hidden by a filter, moved to another stage, changed status or reassigned. Clear the filters and check all stages before recreating the record.
Duplicate opportunities usually come from multiple workflows, triggers, forms or integrations creating opportunities for the same contact. Find the cause before deleting them.
Good pipeline stages represent real changes in the sales process and stay few enough to manage. Each stage should map to a clear action or outcome.
A custom field that does not update usually means the action that should write it never ran, wrote the wrong value or mapped to the wrong field. Trace the update before changing the field.
Custom fields store per-contact information on a record. Custom values store reusable account-level information you reference in many places. Here is how to tell them apart.
When Google Ads spends your budget but produces no leads, the break is somewhere between the search, the ad, the landing page, the form and the tracking. Follow the lead journey to find where it stops.
A campaign that barely spends is usually limited by bid, search volume, targeting or approval, not by budget. Check whether the campaign is even eligible to spend before raising the budget.
Impressions mean eligibility, not interest. When ads show but do not get clicked the cause is usually ad relevance, weak headlines, poor positioning or a mismatch between the search and the offer.
Clicks without form submissions mean the ad worked but the landing page did not. Check the landing page, the form, the mobile experience and the traffic relevance to find why visitors leave.
Low quality leads usually come from the search intent, the targeting or the offer, not from the campaign being broken. Change what the campaign asks for to get better leads.
A high cost per lead is a symptom of expensive clicks, a low conversion rate, poor lead quality or inaccurate tracking. Find which factor drives the cost up before touching bids.
A high cost per click comes from auction competition, keyword intent, match types and Quality Score. Focus on what you can change: relevance, match types and negative keywords.
A low click through rate means your ad shows but does not earn clicks. The cause is usually keyword relevance, ad copy, search intent or competition. Fixing CTR improves performance and cost.
A low conversion rate means clicks arrive but visitors do not act. Check whether the right people are arriving and whether the page actually converts before changing the campaign.
Budget waste is rarely one big leak. It is usually several small ones across search terms, match types, targeting and tracking. This 10 point audit walks through where budget disappears.
Irrelevant search terms mean your ads show for searches that do not match your service. The cause is almost always broad match with too few negative keywords. Read the report and tighten match types.
Negative keywords stop your ads from showing on unwanted searches. Used well they cut waste. Used too aggressively they block the traffic you want. The skill is adding enough without overblocking.
Broad match is not good or bad. It is a tool that works with strong conversion signals and hurts you without them. Whether it helps depends on your tracking, your bid strategy and your account maturity.
Phrase match reaches more people with close relevance. Exact match reaches fewer people with precise relevance. The choice depends on how much control you need and how much reach you want.
Your keywords do not bid against each other. Google only enters one ad per advertiser per auction. The real issue is usually overlapping structure and confused reporting, not true competition.
A low Quality Score is a diagnostic signal from three components: expected CTR, ad relevance and landing page experience. Find the weak one and fix the real mismatch instead of chasing the score.
Ad Strength rates how well your responsive search ad uses its assets. It is not a measure of how well your ad performs. Treat it as a setup suggestion, not a performance verdict.
Responsive search ads are only as good as their assets. When they underperform the problem is usually weak assets, poor keyword relevance or a landing page that does not convert.
Assets show based on eligibility, Ad Rank and the search. The usual reasons they do not show are eligibility, approval, Ad Rank or scheduling. Maximize eligibility rather than expecting display on every impression.
Call tracking problems fall into two types: the call happens but is not tracked, or the call is tracked but not counted as the right conversion. Check the call asset, the conversion action and the CRM capture.
When a form conversion does not record, the break is between the form succeeding and the conversion action receiving the data. Trace the event from submission to Google Ads before reinstalling tags.
A duplicate conversion means more than one tracking path fires for the same action. Find every firing path, keep one and disable the rest before removing any tags.
A wrong conversion value comes from the tag setup, the data layer, currency, a GA4 import or CRM data. Find how your value is set before changing the number.
Enhanced conversions need user data, correct field mapping, proper hashing setup and consent. Confirm each piece is present and connected before re-enabling the feature.
The GCLID has to survive from the ad click to the CRM. Check auto-tagging, redirects, the hidden form field, the CRM mapping and cross-domain setup to find where it drops.
Offline imports fail on the identifier, the conversion action, the conversion time or the format. Check each piece and review the upload results to find rejected rows.
Qualified lead feedback needs a clear CRM stage, a stored identifier, a dedicated conversion action and fresh uploads. Complete the chain before expecting improvement.
Google Ads and your CRM often count different things. Match the definitions, remove duplicates, account for spam and reconcile attribution before assuming one system is wrong.
Out-of-area impressions often come from the presence vs interest setting, ambiguous location signals or a too-wide radius. Check the location report before assuming the targeting is broken.
For a local service business, tight location targeting concentrates budget in the area you serve. Define your area, choose the right targeting type and exclude what you do not serve.
An ad schedule only gates an eligible campaign. Check the account time zone, campaign eligibility and budget exhaustion before assuming the schedule is broken.
Device differences are often normal behavior. Test the mobile experience, check lead quality and confirm the sample size before adjusting device bids or excluding a device.
Structure follows intent and budget. Group by theme, separate budgets into campaigns, keep ad groups tight and use clear names so the account stays manageable.
Too many campaigns fragment budget and starve data. Consolidate what shares intent and budget, but check business and geographic needs before merging anything that needs separate control.
A bidding strategy only works if the data, budget and goals support it. Check tracking, conversion volume and primary conversions before switching strategies.
Fast spend is often Maximize Conversions working as intended. The question is whether it produces conversions and quality leads. Check tracking, search terms and targeting first.
Target CPA underperforms when the target is unrealistic, data is thin, tracking is wrong or recent changes triggered learning. Ground the target in real history before judging the strategy.
Learning is normal after a significant change. Confirm the change was correct, avoid more changes during learning and investigate if performance does not recover rather than waiting indefinitely.
Landing page experience is more than a Quality Score component. Match the page to the ad, fix the mobile experience, improve speed and make the form usable so clicks have a chance to convert.
Leads not becoming customers is usually a downstream problem. Check lead quality, response time, the CRM pipeline and feedback to the ads before blaming the campaign.
When LSA runs but produces few leads, the cause is usually visibility, targeting, profile setup or lead handling. Work through what controls whether your business appears before changing budget.
Low quality LSA leads usually come from broad categories, a wide service area or a lead handling gap. Separate irrelevant inquiries from unqualified prospects before changing anything.
Fast LSA spend is not automatically a problem. The real question is whether those charges are turning into real leads and booked jobs. Check lead quality before changing budget.
An LSA budget not spending usually means the profile is not showing, not that the budget is too low. Confirm visibility and eligibility before raising the budget.
An LSA ranking drop can come from reviews, responsiveness, profile changes, competition or market demand. No single setting guarantees recovery. Check what changed first.
An LSA profile not showing usually means eligibility, category, location or hours are blocking the impression. Searching for your own ad is not a reliable visibility test.
LSA calls not coming through is either a no-call-leads problem or a call-handling problem. Check the lead records first to find which one you have.
Missing LSA leads in the dashboard are usually a view problem: date range, profile selection or a filter. Check the view settings before assuming leads were lost.
LSA showing outside your service area is often a broad area setting or a location intent mismatch. Review the actual leads before assuming the targeting is broken.
LSA service area problems come from an area that is too broad or too narrow. Match the area to where you actually travel for work rather than expanding it to chase volume.
Wrong LSA business hours reduce lead flow and increase missed calls. Match the hours to when you can actually respond and plan for after-hours inquiries.
LSA profile approval depends on category, market, verification and screening. Requirements vary by industry and location, so follow the instructions in your account.
LSA reviews not updating usually trace back to the source profile, the connection or a synchronization delay. Confirm the source before assuming reviews are lost.
An LSA lead dispute not working as expected is often a category or evidence issue. A bad sales outcome is not the same as a credit-eligible lead. Follow the current process.
Being charged for a bad LSA lead is not the same as being wrongly charged. Separate invalid leads from unqualified prospects and review only the ones that meet the current criteria.
LSA leads not booking jobs is usually a response, qualification, scheduling or follow-up problem. LSA generates the lead, the rest of the process turns it into a job.
Response time and missed calls directly affect whether LSA leads convert. Answer fast, handle after-hours leads with a follow-up process and build follow-up into your CRM.
Wrong LSA categories and job types pull in irrelevant inquiries. Select only the work you actually want to do and review the leads to confirm the change improved relevance.
Multiple LSA locations need separate profiles, areas, numbers and routing. Do not duplicate profiles to gain visibility. Track each location separately in your CRM.
LSA lead count is not the business outcome. Track the full journey from lead to booked job in your CRM so you can see what is actually working. Do not invent integrations that do not exist.
A nurture sequence that stops after one or two messages assumes silence means no. Most leads need more touches, with reply conditions and a defined end before the sequence gives up.
When a lead replies and the CRM keeps sending automated messages, the automation has no exit condition. Add a reply check before every send and hand off to a human.
Leads owned by unavailable reps sit unworked. Plan reassignment for both existing and new leads, route to a backup or pool and notify the new owner so leads do not sit twice.
Uneven round robin is usually uneven setup, not a broken rotation. Check active users, existing ownership and segment-based routing before blaming the feature.
An appointment that does not move the pipeline is a broken link between the calendar and the CRM. Confirm the booking trigger, record links and status-to-stage mapping.
Bad won and lost data makes every pipeline report unreliable. Close every deal with a status, a reason and a date, and make the close-out process consistent across the team.
Clean old leads with tags and status, not deletion. Segment by status and last activity, handle duplicates first and back up before any bulk change.
Organize the list before the outreach. Segment by source and past status, remove anyone without consent, separate closed-lost from quiet leads and clean the data first.
An inaccurate pipeline report is almost always inaccurate data underneath it. Fix stage movement, close out stale deals, remove duplicates and check the filter.
A full audit from new lead to closed customer. Walk the same journey a lead takes through your CRM and check each link to find exactly where the process breaks.
When the Meta Pixel is not firing, no browser event reaches Meta. Trace the event flow from page load to Meta receiving the event to find the exact step that breaks.
A local fire only proves the code ran. Use Test Events against the correct dataset to confirm Meta actually received the event before you start changing tags.
Double firing is usually two installations, not one broken tag. Find both sources using the network initiator column and remove one before changing trigger logic.
A missing Purchase event is usually a trigger pointing at the wrong page or purchase data that is not available when the tag fires. Complete a test order and check the confirmation page first.
AddToCart fails most often because the trigger type does not match the cart behavior. Find out whether your cart reloads or updates in place before choosing a trigger.
InitiateCheckout fails when the trigger does not match where checkout actually begins in your flow. Define that point for your specific platform before building the trigger.
A wrong purchase value is a definition and mapping problem, not a tag bug. Decide what the value should represent, confirm the tag pulls that exact figure and make the currency match.
Wrong tracking comes from trigger conditions that are too broad or pointed at the wrong signal. Scope each trigger to the exact page, click or event that represents the action.
A diagnostic describes the state of your data, not a command to rebuild your setup. Match the warning to the right part of your implementation before you change anything.
A working test only proves the tag fires in ideal conditions. Real visitors have ad blockers, stricter browsers and different consent choices. Reproduce those to find missing events.
CAPI failures are silent in the browser. Log the server response, confirm the authentication and dataset are correct and verify the required payload fields to find where the server event stops.
Pixel and CAPI sending the same event is expected. The goal is deduplication through a matching event name and event ID, not removing one of the delivery paths.
Failed deduplication is an identifier mismatch. Confirm the event name is identical and the event ID is the same value on both paths for the same action before looking anywhere else.
Event Match Quality reflects the data you send. Improve it by sending more matching data you are permitted to collect, correctly formatted, not by toggling a setting or bypassing consent.
Browser and server counts will not always be identical because the paths face different restrictions. Find unexpected differences, confirm event names and deduplication, then investigate.
A missing CAPI Purchase is a server-side flow problem. Confirm the order triggers the server logic, confirm the payload carries the order data and check the CAPI response.
Duplicate CAPI Purchase events come from repeated server requests or multiple integrations. Use a unique event ID like the order ID and make sure only one path sends Purchase per order.
A value mismatch means the browser and server pull from different sources or calculate differently. Pick one value definition, apply it to both paths and confirm the currency matches.
The event ID is the link between the browser and server events. Use a stable identifier like the order ID, generate it once and confirm the same value is present on both paths.
Missing user data you are permitted to collect is a fixable gap. Send the data you lawfully can, correctly formatted. Never send data without appropriate consent or fabricate unavailable data.
Server-side tracking has a longer flow than a browser tag. Trace the event from the user action through the web tracking, server endpoint and container to the destination to find the exact step that breaks.
Confirm whether the event reaches the server container before you debug the Meta tag. A tag that never fires because no event arrived is an upstream problem, not a Meta configuration problem.
A server-side GA4 setup sends measurement protocol requests from the server container. Trace the path from website to server container to GA4 to find where the event stops.
Web and server counts do not always need to be identical. Find the differences you cannot explain, then trace whether they come from routing, filtering, consent or browser restrictions.
Duplicate conversions come from multiple delivery paths or repeated server requests. Map every path before you remove anything, because some paths are intentional and use deduplication.
UTMs have to survive the journey from landing page through the browser and server to analytics or your CRM. Find the step where the parameters drop to recover attribution.
Click IDs link a lead back to the ad click. They have to survive from the landing URL through storage and the server request to your CRM or ads platform. Find where the identifier drops.
Server-side tracking does not bypass consent. The web layer still has to send the request, and that depends on consent. Align the consent configuration with the server flow rather than working around consent.
Preview proves configuration works under controlled conditions. Compare preview and production to find what changes between them, from unpublished versions to endpoint and environment differences.
A server-side audit tests the full path from the browser event to the destination platform. Test each link in order and fix the earliest failing step first, because later steps depend on it.
A missing GA4 event usually has a specific cause in the flow from user action to reporting. Diagnose the flow before you rebuild anything, because most missing events come from a single point of failure.
DebugView proves GA4 received the event. If it appears there but not in reports, the cause is usually processing delay, a report filter or a date range issue, not a tracking failure.
Duplicate GA4 events usually come from two installations or two triggers, not one broken tag. Find both sources using the network initiator before you remove anything.
A key event depends on the underlying event. Validate the event first in DebugView. Only when the event reliably reaches GA4 should you check the key event configuration.
GA4 and Google Ads are different platforms with different rules. They will not always report identical numbers. Find unexpected gaps and explain them against attribution, counting and timing differences.
A GA4 lead is an event and a CRM lead is a contact record. They are different things. Find where the journey breaks between the website action and the CRM record before treating a gap as a tracking error.
Wrong source and medium usually come from missing UTMs, redirects stripping parameters or self-referrals. Find where the wrong value originates before changing attribution settings.
Some direct traffic is real, but a large direct number often means lost attribution. Check missing UTMs, redirects, referrer loss and untagged links before assuming the direct number is accurate.
UTMs in the URL is not the same as UTMs in GA4. The parameters have to be captured and processed into the right acquisition dimension. Confirm the URL, the capture and the dimension scope.
Cross-domain tracking maintains session continuity and attribution across domains. Confirm both domains use the same property, the linker decorates the links and the client ID persists across the boundary.
Do not exclude a referral until you understand its role in the customer journey. Exclude only the ones that distort attribution, like payment processors in your own flow.
Separate actual traffic decline from measurement decline before you fix anything. A real decline is a marketing problem. A measurement decline is a tracking problem. Check other data sources to tell them apart.
A sudden change comes from a tracking change, a real traffic change or unwanted traffic. Identify which source applies before you react, because each needs a different response.
A tag that does not fire has nothing to send. Use GTM preview to see whether it is fired, not fired or skipped. The status tells you whether to fix the trigger, the consent condition or the container version.
Coverage gaps come from page templates, landing pages, subdomains and single-page routes that the tag setup does not cover. Audit every page type and confirm the pageview tag fires on each.
A missing GA4 purchase event usually breaks somewhere between order completion and the GA4 request. Trace the flow from order data to the purchase event to find where the data stops.
GA4 revenue and store revenue will not always be identical. Reconcile the two by separating missing purchases, duplicate purchases and value definition differences before treating a gap as a tracking error.
Duplicate GA4 purchases usually come from multiple tracking implementations or a confirmation page that fires the event more than once. Map every purchase path before you remove anything.
transaction_id is what makes a purchase deduplicatable and reconcilable. Trace the order ID from the confirmation page through the data layer to the GA4 tag to find where the identifier drops.
A wrong purchase value usually comes from tax, shipping or discounts being included or excluded differently at some step. Define what the value should represent, then check each step in the flow.
Add to cart often happens without a page reload. Confirm whether your store uses AJAX add to cart and use a data layer event trigger so the add_to_cart event fires on the actual action.
Checkout can begin with a button click, a page load or an express button. Determine the real point at which checkout begins before you configure the begin_checkout event trigger.
An event firing is not the same as the product data being correct. Confirm the data layer holds the right product, including variants, before you check the GA4 tag variables.
The items array connects an ecommerce action to the products involved in it. Confirm it is present in the data layer with the correct structure and count before you check the GA4 tag mapping.
When the items array exists but specific values are wrong, build a field-by-field audit. Compare each field against the page and trace the wrong value to its source in the data layer or tag variable.
The data layer is the foundation under every GA4 ecommerce event. If the push is missing, mistimed or malformed, no tag configuration can recover the data. Find where the data stops.
Shopify stores can send the purchase event through native integrations, GTM, apps and custom code. Using more than one method creates duplicates. Confirm which methods your store actually uses.
Some orders exist in Shopify but never appear in GA4. Find the pattern in the missing orders, because the condition that causes the failure tells you which orders are affected and why.
Funnel stages do not always form a linear sequence. Customers skip steps and use direct checkout. Focus on unexpected inconsistencies rather than expecting every stage to decrease neatly.
A complete ecommerce audit walks the full journey from product view to purchase. Test each stage in order and fix the earliest failing step first, because later stages depend on the ones before them.
A workflow that never starts usually breaks between the business event and the first action. Trace the flow from trigger to conditions to entry to find where the workflow fails to start.
A workflow that stalls ran successfully up to a point then failed or waited. Trace the workflow step by step through the history instead of rebuilding everything.
Duplicate execution has two causes that look the same. One workflow running twice needs a different fix than two workflows doing the same action. Identify which you have first.
Duplicate messages come from one action running twice or multiple paths sending the same content. Map every message path before removing any workflow.
A workflow that keeps running after a reply has a broken detection or stop condition. Map the flow from message sent to reply detected to automation stopped.
When contacts take the wrong branch, the condition evaluated against unexpected data. Check the actual field values before changing the condition rules.
Wait steps fail in the duration, the condition, the time zone or the scheduling logic. Time zone handling varies by platform, so confirm which one applies before debugging.
A message at the wrong time is either a scheduling problem or a delivery delay. Check the send timestamp in the platform first to distinguish the two.
Trigger filters control who enters a workflow. If the wrong contacts enter, the filters are too broad, too strict or checking fields set after the trigger fires.
Re-entry may be legitimate or unwanted. Check the enrollment count, the re-entry setting and the trigger count before deciding whether to block repeated entry.
Automated assignment fails in the rules, the eligible users or the trigger. Check the rule order and conditions before assuming the CRM assignment system is broken.
Round robin skews when the eligible list includes inactive users, existing owner checks skip leads or a second workflow runs independently. Confirm what your platform promises first.
Duplicates come from create versus update logic, not the CRM database. Fix the matching in the workflow before you deduplicate the records.
Duplicate opportunities come from trigger repetition, re-entry or a missing existing opportunity check. Confirm the trigger fired once before treating it as a duplication bug.
A field that does not update usually means the value never reached the field. Trace the value from source through the mapping to the CRM record.
Tag actions fail in the trigger, the conditions or conflicting workflows. Check the workflow history for the contact before assuming the tag system is broken.
Overwrites are caused by the last process to run. Trace the field history to find the last writer before changing any individual workflow.
The wrong record updates when the match finds a duplicate or the lookup returns multiple results. Reliable identification has to happen before the update action runs.
A pipeline move needs a trigger, an opportunity to update and a valid destination stage. Map the flow from lead action to opportunity update to find where the move fails.
A notification either was never generated or was generated but not delivered. Check the workflow history first to find which stage failed.
A webhook that does not fire either never reached the action or created no request. Determine whether the request was created before debugging the endpoint.
A webhook that fires but delivers no data failed at one of three points. Check both the sending and receiving sides to find where the request failed.
Wrong field mapping happens when the key does not match the payload or the data type does not match the destination. Capture the actual payload before changing the mapping.
Duplicate webhooks come from repeated triggers, re-entry, duplicate source events or retries. Trace the source before assuming retries caused the duplicate.
A test that passes proves the logic works. The production failure is an environment difference, so compare the endpoint, authentication, payload and trigger between the two.
A webhook error means the request was sent and a response came back. Read the status code first, because a 4xx tells you to fix the request and a 5xx tells you to look at the receiving system.
Most auth failures are a single expired or rotated credential. Confirm the auth method, the expiry and the permissions before reconnecting, because reconnecting can mask the real cause.
Missing data can disappear at the source, the request, the response, the mapping or the destination. Find the stage where the value exists on one side and is missing on the other.
Duplicates come from a repeated trigger or a missing lookup before create. Trace the repeated request before changing the destination, because the fix is in the trigger or the create logic.
Rate limits reject valid requests because there are too many in a window. Confirm the actual limit for your API, then batch, space or queue requests to stay under it.
A Zap that never runs is usually a status, connection or trigger type problem. Test the trigger to see if Zapier receives data before debugging the action steps.
Duplicates come from a repeated trigger, a missing search step or multiple Zaps. Map every path that creates the destination record before removing anything.
Wrong mapping happens when the output is mapped to the wrong CRM field or the data type does not match. Confirm the source value is correct, then check the mapping step.
A test that passes proves the logic works with clean data. A live failure means real data exposed a gap. Find the failing real lead and compare it with the sample.
A multi-step Zap that stops early failed at a filter, a path or a step. Read the task history to find the last successful step and the first missing one before rebuilding.
A scenario that never runs is usually a status, scheduling or connection problem. Confirm it is active, then test the trigger to see if Make receives data.
Wrong mapping happens when the output is mapped to the wrong field or bundles are handled incorrectly. Inspect the output bundle, then check the destination module.
A scenario that stops halfway failed at a module, a filter or a router. Read the execution history to find the last successful module and the first failing one.
Duplicates come from duplicate input or duplicate processing. If the input is duplicated, fix the source. If the processing is duplicated, add a search before create.
The payload arrived but the scenario stalled after the trigger. Check the filters, the router and the data structure to find where the bundle stopped.
A workflow that works manually but not in production is almost always not activated. Confirm the active toggle, then check the trigger node type and credentials.
The test URL and the production URL differ. Confirm the workflow is activated and the source sends to the production URL, not the test URL.
Duplicates come from repeated triggers, a missing lookup node or multiple workflows. Trace the repeated execution before changing the destination system.
A node that stops tells you what it needed and did not get. Read the error message, check the input items and confirm required fields are present.
The source data is correct but the next node receives the wrong values. Find where a correct source value becomes an incorrect value passed to the next node.
An API request fails when the endpoint, method, auth, headers or body does not match what the API expects. Read the status code and compare with the current documentation.
Conflict happens when the boundary between manual and automated actions is unclear. Define ownership at each stage and add stop conditions so automation steps aside when a rep takes over.
Failed workflows go unnoticed when there is no error path, notification or fallback. Make failures visible, cap retries and route unrecoverable failures to a manual review queue.
A test that passes proves the logic works with clean data. Trace one real lead through the full flow and compare it with the sample to find where the data breaks.
A complete audit from source event to final action. Most automation problems are in the handoff between stages, so work through the checklist in order with a real record.
When form automation never starts, the break is usually before the workflow. Confirm the form recorded the submission, the contact was created and the trigger is on the correct form before debugging the workflow.
Missing form data means a value existed at one stage and was absent at the next. Confirm the field was captured, then follow it through the payload and mapping to find where it drops.
A created contact proves the trigger and data work. The missing opportunity points to the opportunity action, its conditions, the pipeline configuration or duplicate prevention logic. Isolate that step.
The opportunity exists but landed in the wrong place. The creation worked, so the problem is the action configuration, the routing conditions or the default fallback. Find where the wrong choice was made.
A sales action only updates the status if it generates an event the automation can trigger on. Confirm the trigger fired before debugging the status update logic.
Appointment automation starts with the booking record. If the appointment is not in the calendar or not linked to a contact, the workflow has nothing to trigger on. Confirm the booking before debugging the workflow.
First determine whether the appointment itself updated correctly, because an automation cannot react to a change that was not recorded. Confirm the appointment state before debugging the trigger.
Nurture continues after a booking because the workflow never learned the booking happened or the stop condition checks the wrong signal. Confirm the booking was detected before removing follow-up steps.
A customer stays in prospect automation when the lifecycle transition is not connected to the exit logic. Confirm the conversion was detected and the exit condition checks a value the conversion sets.
Distinguish a task that was never created from a task that exists but is invisible. The first is a workflow action problem. The second is an assignment or visibility problem.
A delayed sync is not a broken sync. The data arrives, which means the path is correct. Find the step that introduces the wait, whether it is polling, a wait step, batching or rate limits.
Missing data between apps is a path problem, not a delay. Trace the value from the source record through the trigger, payload, mapping and destination to find where it drops.
Two-way sync problems come from design, not a single failed request. Define a source of truth, use reliable record matching, prevent update loops and compare timestamps.
Wrong date and time fields are conversion problems, not capture problems. The value was right at the source. Find where the time zone, UTC interpretation, format or daylight saving handling changes the value.
Phone and email formatting problems are normalization problems. The value was captured but stored or sent in a format the destination rejects. Normalize whitespace, capitalization and country codes before the destination.
An empty field is not the error. The error is the workflow not accounting for the possibility that the field would be empty. Handle missing data with explicit branches, meaningful defaults or skipped steps.
Filters that exclude valid leads are too restrictive or check the wrong value. Test the filter against the actual excluded lead to find the failing condition. Loosen the filter carefully to avoid letting the wrong leads in.
Old leads re-enter because a change made them eligible again, not because of a trigger bug. Find what changed, confirm whether the changed field is a trigger and check the re-entry settings.
Too many workflows create conflicts, duplication and overwrites. Audit the whole set before deleting anything. Group by trigger, identify overlapping actions and field updates, consolidate genuine duplicates and document ownership.
An automation system is a chain. A break at any stage breaks everything downstream. Work through this checklist in order from lead capture to sales handoff, fix the first failure you find and retest the full journey.
When a Looker Studio report fails to load, the cause sits at the report, data source, connector, permissions or browser layer. Isolate which layer is failing before rebuilding anything, because the fix depends on where the break is.
A data source that will not connect usually fails on authentication, account access, the connector or the underlying resource. Confirm where the connection breaks before reconnecting everything, because each cause needs a different fix.
No data means either the source has no matching rows or the chart filters everything out. These need opposite fixes, so confirm which one applies before changing the report configuration.
Looker Studio and GA4 can show different values because of date range alignment, time zones, the connector, filters, attribution and data freshness. The goal is to explain the difference, not to force identical numbers.
Google Ads data in Looker Studio can differ from the interface because of account scope, date range, conversion attribution, metric definitions and data freshness. Match the scope and definition before comparing totals.
When the date range control does not work, charts ignore the range or stay stuck on a default. The cause is usually the control link, the chart date dimension, the default range or the date field type.
Filters that show the wrong data usually fail on scope, include or exclude logic, AND or OR conditions or the filtered dimension. Separate a filter problem from a source data problem before changing anything.
A calculated field that fails usually breaks on a missing field, a type mismatch, null values, aggregation or syntax. Simplify the formula and test one operation at a time to isolate the failing part.
A wrong total can come from the source data, aggregation, blending, a calculation or a filter. Trace the value through each layer to find which one introduced the wrong number before changing the chart.
Blended data mismatches usually come from join keys, granularity, aggregation or null values in the join dimension. Test each source separately before blaming the blend.
A slow report is a configuration problem. Large date ranges, many charts, blended data and calculated fields all increase load time. Reduce the work the report does on each load.
Duplicated metrics usually come from blended joins, duplicate source rows, aggregation or multiple sources pulling the same metric. This is a reporting problem, not a duplicate conversion tracking problem.
Wrong conversion data usually comes from the conversion definition, attribution, date range, filters or a blended join. Validate the conversion definition before comparing any totals.
A CRM dashboard that differs from the CRM usually fails on the connector, the date field, filters, duplicate records or data freshness. Trace the value from the CRM record to the dashboard metric.
A complete audit checks data sources, authentication, date ranges, dimensions, metrics, filters, calculations, blended data and the final totals against the source platforms. Work through each layer in order.
When Hotjar tracking does not work, the cause is the tracking code, page coverage, consent, a website change or browser behavior. Confirm the code is present and loading before debugging anything deeper.
Recordings can be missing even when tracking is installed, because of consent, sampling, page targeting or browser restrictions. Distinguish a tracking problem from a capture problem first.
A heatmap that stays empty usually fails on the page URL target, traffic volume, dynamic page behavior or consent. Confirm the URL matches the target and the page gets enough matching sessions.
Wrong page tracking comes from URL targeting rules, query parameters, dynamic URLs or SPA behavior. Compare the rules with the actual URLs the site uses to find the mismatch.
Missing form interactions usually come from dynamic forms, iframes, privacy masking, consent or page behavior. Do not disable privacy protections to see more data.
Hotjar counts visual clicks and GA4 counts events sent through tracking. They measure different things under different consent and coverage rules, so a difference is usually expected.
A rage click signals frustration, but the cause varies: a broken button, slow response, misleading design, a mobile issue or a form problem. Watch the recording to diagnose the actual cause.
A dead click means a visitor clicked an element that did nothing. The cause is either an element that looks clickable but is not, or a broken link or button. Identify which one before changing the page.
Form drop-off can come from visibility, field length, required fields, errors, mobile usability, trust or page speed. Use recordings and heatmaps to find where visitors leave.
Map the journey from landing to conversion, find the step with the largest drop and use Hotjar behavior alongside GA4 and CRM data rather than relying on Hotjar alone.
No recordings can mean the tracking is missing, firing but blocked or firing but not surfacing sessions. Isolate which failure state you are in before reinstalling anything.
Empty heatmaps usually come from low traffic, wrong URL targeting or a page that changed since the heatmap was set up. Confirm the cause before recreating the heatmap.
Sessions on unexpected URLs usually come from query parameters, redirects, subdomains or dynamic routing. Understand your URL structure before assuming the tracking is broken.
A rage click is a signal of frustration, not a diagnosis. Watch the recording, reproduce the click and confirm whether the element is broken, slow or misleading.
A dead click can reveal a broken link or a misleading element. Confirm whether the element is meant to be interactive before choosing a fix.
Excessive scrolling is not always bad. Use heatmaps and recordings to distinguish visitors who read from visitors who hunt for missing information.
A quick back often signals an intent mismatch, slow load or content that does not match the link promise. Segment by traffic source before concluding the page is the problem.
Clarity connects JavaScript errors to real visitor sessions. Use the recordings to find the scenario, then debug with browser developer tools rather than relying on Clarity alone.
Clarity and GA4 define sessions differently and apply different consent and coverage rules. A difference is normal. Investigate only sudden or unexplained changes.
Empty UTM filters usually come from redirects that strip parameters, wrong landing pages or filter values that do not match the recorded UTM exactly.
Three different measurement models produce three different numbers. Focus on identifying unexpected differences rather than forcing identical totals across platforms.
A lead present in analytics but missing from the CRM failed somewhere between the conversion event and the CRM record. Trace the data journey to find where it drops.
More traffic does not automatically create more leads. Segment the increase by source and use behavior data to find where new visitors arrive and where they leave without converting.
Clicks without conversions mean visitors engage but never complete the goal. Combine GA4 events, heatmaps, rage and dead click signals and recordings to find where the journey breaks.
A full analytics audit from tracking installation to final reporting across GA4, GTM, Looker Studio, Hotjar, Clarity and CRM. Find gaps, duplicates and mismatches across the whole stack.
