Blog

Notes on finding what pays.

38 practical guides. No invented numbers, no borrowed proof.

AI traffic

• Answer engine optimization checklist for small sitesA 12-point answer engine optimization checklist: page rules, brand facts, off-site mentions and a free weekly check. No citation can be promised.6 min read
• Do AI-referred visitors convert better? How to checkDo visitors from ChatGPT or Perplexity buy more than others? Nobody can say for your site until you measure revenue per visitor. Here is how.6 min read
• Generative engine optimization basics for small sitesWhat generative engine optimization (GEO) means for a small site, which steps are cheap and sensible, and what nobody can promise about AI citations.7 min read
• How to track ChatGPT referrals in your analyticsChatGPT links carry utm_source=chatgpt.com and usually a chatgpt.com referrer. Here is how to find them, what gets lost, and how to check sales.6 min read
• llms.txt: what it is and whether it mattersllms.txt is a proposed Markdown file that points AI tools to your key pages. Its effect on search is unproven. Here is when it is still worth adding.5 min read
• Perplexity traffic: what to expect and how to track itPerplexity cites sources with links, so clicks usually arrive with a perplexity.ai referrer. Here is what to expect, what to count, and what to ignore.6 min read

Attribution

• Bot and referrer spam: how to filter your analyticsSpot fake visits in your analytics with five quick checks, then keep them out of the numbers you use to make decisions about traffic and sales.6 min read
• Conversion funnel for small SaaS: three steps that matterA small SaaS needs only three funnel steps to find where buyers drop off: a pricing view, a checkout start and a paid sale. Here is how to set them up.6 min read
• Cookieless analytics and GDPR: what to knowCookieless analytics reduces what a tracker stores, but it does not settle GDPR on its own. General information on what to check and who to ask.6 min read
• Direct traffic explained, and what dark social hidesDirect traffic is not just people typing your URL. Learn what hides inside it, how dark social inflates it, and how to split it into readable parts.6 min read
• First touch vs last touch attribution for solo foundersFirst touch credits what introduced a buyer, last touch credits what closed the sale. Here is which to use, with a worked example and pitfalls.6 min read
• How long do buyers take to purchase? Measure your ownTime to purchase varies by product and source. Learn how to measure your own median, why averages mislead, and what it means for attribution.6 min read
• How to find which post got you paidTo connect a sale to the post behind it you need the referrer, the landing page and the payment. Here is how the three fit together.5 min read
• How to know which blog post made a salePage views don't show which article earned money. Here is how to tie each sale to the post a buyer read first, using links, ids and a simple checklist.6 min read
• How to track conversions without cookiesYou can count conversions without cookies, but sales link to visits only from the same day. Here is what that costs you and how to work around it.6 min read
• Revenue per visitor, the metric that beats trafficRevenue per visitor shows which sources actually pay. Learn the formula, a worked example, and when the number misleads you.6 min read
• UTM parameters, a complete guide for foundersWhat each UTM parameter means, how to name them so reports stay clean, and how to keep them alive through checkout. Includes a naming checklist.7 min read
• Why your payment dashboard cannot show where buyers came fromStripe-style dashboards record who paid, not the visit before it. See what they miss and the small change that adds the traffic source to each sale.6 min read

Growth

• A weekly analytics review routine for solo foundersA 20-minute weekly routine: six numbers to check, what each should make you decide, and when to ignore the week entirely.6 min read
• Cold email tracking from send to sale, with UTM tagsTag every cold email link with utm_source, utm_medium and utm_campaign so each angle is measurable from click to revenue.6 min read
• How to measure a Product Hunt launch beyond upvotesTag your links, mark launch day, and track sales and revenue per visitor so you know whether a Product Hunt launch paid off.6 min read
• Measure a Twitter thread launch from click to saleTag each tweet in a launch thread, then follow the click to the sale so you know which post earned money, not just which got likes.6 min read
• Traffic up but sales flat? What to check, in orderWhen visitors rise and revenue does not, check source mix, landing pages, pricing page behavior, time to buy and attribution gaps, in that order.6 min read
• What to measure when you build in publicFollowers and likes feel like progress. Here are the five numbers that show whether building in public is actually bringing in buyers.6 min read

Payments

• Creem sales attribution explained, step by stepHow Creem sales, subscriptions and refunds get credited to the post, search or referrer that brought the buyer, and what to check when one goes unattributed.6 min read
• Dodo Payments analytics for indie foundersConnect Dodo Payments with one API key and see which traffic source made each sale, with payment links tagged and checkout sessions supported.5 min read
• Polar sales by traffic source: how to see what workedConnect Polar with a token and see which post, search or referrer brought each order, renewal and refund, with no extra tagging on buy links.6 min read
• Refunds and attribution: why net revenue mattersA source that sends buyers who ask for refunds looks better than it is. Here is how to attribute refunds to the original source and read net revenue.6 min read
• Stripe Checkout UTM attribution walkthroughHow to carry a visitor from landing page to Stripe Checkout so every payment, renewal and refund traces back to the post or link that sold it.6 min read
• Subscriptions vs one-time sales attributionSubscriptions need attribution that follows renewals back to the first visit. Here is how credit works for one-time sales, trials, upgrades and renewals.6 min read

Self-hosting

• Ad blockers and analytics: using a first-party proxyAd blockers can hide visitors from your analytics. A first-party proxy serves the tracker from your own domain. Here is how it works and its limits.6 min read
• AI agents reading your analytics: MCP explainedWhat MCP is, how an AI agent reads your analytics through a read-only key, what it can and cannot see, and how to decide if you want it.6 min read
• Cloudflare free plan and analytics: what fitsWhat counts toward Cloudflare's free-plan limits when you self-host analytics, a rough traffic ceiling, and the signs it is time for the $5 plan.6 min read
• Connect Claude Code to your analytics in one commandAdd your analytics to Claude Code with one claude mcp add command, then ask which post made money. Setup, scopes, first prompts and common fixes.6 min read
• Do you need your own analytics server? Probably notYou do not need a server to own your analytics data. A checklist for deciding between self-hosted on Cloudflare, hosted, or neither.6 min read
• One-time vs monthly analytics cost over three yearsA simple three-year cost model for a one-time analytics license versus a monthly plan, with break-even months you can recompute for any price.6 min read
• Pay once or pay monthly: the arithmeticA $9 a month tool against a $49 one-time price. When do they cross?3 min read
• Self-hosted analytics on Cloudflare Workers, explainedHow self-hosted analytics runs on a Cloudflare Worker and a D1 database in your own account, what you manage, and what you never have to touch.6 min read