Free GTM workflows for every stage of the lifecycle.

If your content lives in git, a workflow file catches broken links and banned phrases before a human review ever starts.

Default retention curves flatten everything into one line, and custom brackets let you ask about days 15 to 30 specifically.

Most broken HubSpot automation is a re-enrollment setting nobody read, and this is the article that explains it.

The fastest way to kill the ad hoc DM: a form in Slack that collects what you need and posts it where it gets picked up.

The only way a sale that closes six weeks later gets attributed to the campaign that produced it.

Five steps to a report you can send to a client or an exec, with no license cost and no BI team involved.

Free, privacy-first traffic numbers that survive ad blockers, which makes it a useful second opinion on whatever your analytics says.

The model your analyst built stops being a dashboard and becomes a field your reps can filter on.

The difference between a style guide in a document and a style guide that blocks the output, which is the only kind that holds.

A researcher, a writer, and a fact-checker each with its own prompt and permissions, instead of one prompt trying to be all three.

Configuration, then Custom, then Custom Extraction, and a crawl turns into a spreadsheet of whatever element you pointed it at.

No code at all: create the product, switch the portal on, brand it, and the billing support queue mostly disappears.

Six emails you configure once: failed payment, unpaid invoice, trial ending, renewal, payment confirmation, and expiring card.

Most analytics installs stop at the snippet, and this is the checklist that turns a dead install into something a team reads weekly.

First Contentful Paint, Dom Interactive, and Page Loaded per session, attached to the recording of the person waiting.

Churn stops being one monthly percentage and becomes a list of named users you can watch leave, then a saved list you can message.

Fewer third-party scripts on the page, one place to control what data leaves, and the same Tag Manager interface you already use.

TEAM PICK

A SKILL.md plus any scripts and reference files, loaded through progressive disclosure so a long procedure costs nothing until it's needed.

The free version covers 500 URLs, which is enough to audit most sites, and the inlinks tab names the exact page you need to edit.

Involuntary churn is customers who wanted to stay and whose card declined, and this is the setting that gets most of them back.

The document that stops your analytics rotting: one agreed list of event names, with violations reported when code sends something else.

A free protocol you call from your CMS on publish, so a new page gets discovered in minutes and a deleted one drops out.

Session recordings and heatmaps at no cost on any traffic volume, which is the cheapest way to find out why a page fails to convert.

TEAM PICK

Retention per feature, side by side, so you learn which parts of the product hold people and which ones nobody returns to.

TEAM PICK

Human review is counted per article at 8-12 minutes, against about 3 hours before the system existed.

It saves 15 minutes of manual data entry per brief, which is dull work that never stops arriving.

Toshiya Minami's template batches responses before analysis, so the themes hold up instead of drifting answer by answer.

Five parallel checks scored out of 20 apiece, which finishes in 8 to 15 minutes and produces a number you can put in front of a board.

TEAM PICK

It took six hours a report and now takes about 30 minutes of drafting plus review, and the four prompts that got it there are published.

Six sources in one daily sweep, with the output landing in a Google Sheets calendar instead of a dashboard nobody opens.

Product feedback arrives in five places and gets read in none, so this consolidates the channels and ranks requests by how often they recur.

The calendar stays in a spreadsheet anyone on the team can edit, so planning a month never needs automation access.

TEAM PICK

Daniel Shashko's template only surfaces mentions scoring above 60 on engagement quality, so the daily digest stays short enough to read.

Yassin Zehar's template treats churn MRR and feature adoption like uptime: a baseline, an anomaly, an incident record, and a named owner.

Rahul Joshi's template catches the response, sorts it into a category, and tells somebody, so a score never just sits in a database.

It goes past the alert, generating a Stripe discount code and drafting the email that carries the offer to the customer.

Rahul Joshi's template turns the satisfaction rating nobody reviews into a daily churn signal with a ticket link attached.

TEAM PICK

PollupAI's template builds a health score out of three inputs nobody looks at together: deal age, ticket sentiment, and usage trend.

Alexander Schnabl's template runs entirely on n8n nodes, so a feedback widget stops being a reason to buy another subscription.

Twelve months of monthly tests lifted visitor-to-paid by 2.3x while customer acquisition cost stayed flat.

Adem Tasin's template adds the enrichment step before scoring, so you score on company size instead of guessing at fit.

Adrian Kendall's template keeps the counter in a Data Table, so the rotation carries on after a restart instead of starting over.

Comparison pages convert at roughly 15x a standard blog post because they catch someone mid-decision, so the build rules matter more than the copy.

TEAM PICK

Node by node, with the stage boundaries marked, and a batch mode that turns 20 keywords into 20 briefs in 20 to 25 minutes.

TEAM PICK

22 nodes across four models, processing a 2,000-word article in under 3 minutes, against 45 to 60 minutes for the same job by hand.

8-12 links on a long post, 2-6 words per anchor, and a published argument for pointing at /pricing over another blog post.

The one Slate agent whose full method is published, and it ranks link targets by page tier so product pages win over another blog post.

Negative keywords get skipped because scanning the search terms report by hand is dull, so this hands the scan to Claude and keeps the judgment with you.

Seven published prompts across three stages. Frequency on its own is a poor fatigue signal, so the reading happens on weekly trends.

TEAM PICK

Set it up through Zapier with no code, or run a direct MCP server, then use the published prompts for search term mining, budget pacing, and account audits.

Yaron Been's template treats a job posting as intent: if a company is hiring for the role you sell to, the budget conversation already happened.

All nine name a change you can make to a page today, starting with adding citations and injecting statistics an engine cannot generate itself.

Short enough to run against a single draft, and it puts a number on where the answer has to sit on the page.

Documented node by node: Ahrefs MCP for matching terms, related terms, and current rankings, then one Claude node that clusters and classifies intent.

TEAM PICK

Blended AI visibility hides the gaps, so this builds a platform-by-category matrix first, then scores content against three named signals.

Work through it line by line when you want one complete pass over a site, from AI bot crawlability through to citation signals.

Incrementors built this to test one site across ChatGPT, Claude, DeepSeek, and Perplexity in a single run, with the competitors named.

TEAM PICK

Shiyam Sunder's write-up of the Skills behind our own work, including a content brief Skill that runs on 25 nodes for about $0.10.

Avkash Kakdiya's template handles the leads a scoring model quietly buries: an owner alert, a CRM task, and a re-engagement email.

Krishna Sharma's template does the dedupe step most form-to-CRM automations skip, then sorts every lead into one of three tiers.

Friedemann Schuetz's template does the year-over-year comparison nobody does by hand, then writes the summary and sends it to email and Telegram.

Manoj Palanikumar's framework starts by defining a query set of 50-100 questions, and most GEO reporting skips that step entirely.

The event-level data behind every GA4 report, in a table you can join to pipeline, with a documented 1 million event daily ceiling.

David Olusola's template covers the first week after signup, with Wait nodes handling timing and Slack telling the team at each step.

Avkash Kakdiya's template reads trial start dates from MongoDB and routes each user to whichever of four emails matches their day.

How a Q&A block, a comparison table, and a listicle item each have to be shaped before a model will lift the answer out.

Mahesh Sundararaman's checklist ends on a 7/30/90 day read, so you know when to judge a refresh and when to leave it alone.

Joe Kurian's brief carries the two fields most templates skip: a refresh trigger, and a 40 to 60 word answer-first summary.

Available on every Cloudflare plan including the free one, so deciding which AI crawlers may read your content stops being a guess.

Custom Workflows AI wired a crawler to Search Console, so the report tells you which broken pages are also the ones earning clicks.

Google's own daily export into BigQuery, which gives you query-level history you own and can join to anything else in the warehouse.

Shiyam Sunder's audit runs from crawlability through to AI citation readiness, and the output is a ranked action list with an owner per row.

TEAM PICK

Research the questions, build an entity-mapped brief, draft answer-first, add schema, retrofit old pages, then measure citations.

One daily run across Meta Ads and Google Ads, with the underperformers pushed to Slack, WhatsApp, and Gmail and logged to a Sheet.

Yesterday's cost, CPC, click-through rate, and impression share in your inbox before standup, with every day logged to a spreadsheet.

TEAM PICK

Google's own script, and it catches the broken landing page before your budget spends another day pointing at a dead one.

A job change is the cleanest buying signal in B2B, and this catches them in batches of five with the role already classified.

Uche Madu's template runs Apollo's organization and people search off a Google Sheet, then parks each contact in a verification tab.

TEAM PICK

The whole BDR loop in one template: source, enrich, qualify, verify the address, then push a written sequence into Instantly.ai.