The Attribution Cheat Sheet for Small Budgets

Enterprise attribution tools cost thousands per month and still get it wrong. Your job is to collect enough signal to make a better decision than you made last month. 80% confidence with free tools beats 95% confidence with software you cannot afford.

1) Where did this person first hear about me? 2) What made them decide to buy? 3) Which channel has the best ratio of effort to sales? If your system answers these three, it is good enough.

Source (where the traffic came from: newsletter, instagram, google), Medium (the channel type: email, social, cpc), and Campaign (the specific effort: may-newsletter, spring-sale). These three are non-negotiable. Content and Term are optional extras for A/B testing ads.

Go to ga-dev-tools.google.com/campaign-url-builder. Paste your destination URL, fill in source, medium, and campaign. Copy the generated link and use it everywhere you share that URL. It takes 30 seconds and gives you data forever.

Lowercase everything. No spaces (use hyphens). Pick your source names once and stick to them. If you call your newsletter 'newsletter' in January and 'email-list' in March, Google Analytics treats them as two separate sources and your data splits. Decide once: newsletter, instagram, facebook, linkedin, google-ads.

UTM links are ugly. Use bit.ly (free tier is fine) or a plugin like Pretty Links on WordPress. The underlying tracking still fires. The link looks clean.

In GA4, go to Reports, then Acquisition, then Traffic Acquisition. Filter by your UTM sources. This shows you which channel sends traffic that converts, not just traffic that bounces. Sort by conversions, not sessions.

A conversion event tells GA4 what counts as a win on your site. Common ones: thank-you page view after a purchase, form submission, booking confirmation. Without these set up, you are watching traffic numbers that mean nothing. GA4 lets you create them free inside Events, then mark them as conversions.

When a big sale comes in, go to Explore in GA4 and create a User Explorer report. Enter the customer's ID or segment by recent purchasers. You can see the exact path they took across sessions before they bought. This is the closest you get to a free customer journey map.

"How did you hear about us?" with a dropdown (Google search, Instagram, newsletter, referral from a friend, podcast, other). This is first-party data that no algorithm can distort. It is not scientific, but it is honest. If 60% of your buyers say newsletter, that is a signal worth acting on.

After someone buys or books, send a short reply: "Quick question, just so I know: what made you decide to reach out?" Not a survey. A conversation. People answer these because they feel human. The answers are worth more than most analytics dashboards.

Copy the responses monthly into a Google Sheet. Source column, month column, purchase value column. After three months you will see patterns you cannot see inside any analytics tool, because this data includes offline referrals, word of mouth, and podcast mentions that no pixel ever captures.

Set source to 'newsletter', medium to 'email', and campaign to the send date or topic (e.g. '2026-05-weekly'). Most email platforms (Kit, Mailchimp, ActiveCampaign) let you add UTMs automatically in campaign settings. Turn this on once and forget it.

Instead of sending subscribers to your main sales page, send them to a variant with a unique URL slug (e.g. /offer-newsletter). Even without UTMs, any traffic to that URL came from your email. This is belt-and-braces attribution when pixel data is missing.

Your Instagram or TikTok bio link should always be a UTM-tagged URL. Set source to 'instagram', medium to 'social'. When someone clicks through and buys, you have the data. Without this, all that traffic shows up in GA4 as 'direct' and you have no idea it came from social.

Dark social is when people share your content via WhatsApp, direct message, or text. It shows up as 'direct' traffic in analytics but it is not direct, someone recommended you. If you see a spike in direct traffic after a post goes semi-viral, that is almost certainly dark social. Note it in your spreadsheet. It counts.

Meta Ads Manager defaults to a 7-day click and 1-day view attribution window. This means if someone saw your ad and bought 6 days later, Meta claims credit. Your GA4 may show zero from that campaign because no click fired. Neither is lying. They measure different things. For a small budget, trust the one that matches your revenue: if GA4 and Meta both show a conversion in the same week, count it once.

1) Google Analytics 4 (free) for on-site behaviour. 2) UTM parameters on every link you share (free, 30 seconds per link). 3) A 'how did you hear about us' form field (free, in any form builder). 4) A Google Sheet to log monthly totals by source (free, 20 minutes a month). That is the entire system.

On the first of each month: 1) Which source drove the most conversions in GA4? 2) What did buyers say in the form field? 3) Were there any spikes in direct or referral traffic I can explain? 4) Which channel cost the most time or money relative to what it returned? Adjust the next month based on what you find. That is the whole process.

Revenue per channel, not traffic per channel. A channel that sends 500 visitors with zero sales is worse than a channel that sends 50 visitors with 10 sales. Sort everything in your Sheet by revenue contributed, not clicks or impressions. That one habit changes most small business marketing plans within 90 days.

GA4 by default credits the last touchpoint before a sale. Someone reads your newsletter, visits your site, leaves, sees a retargeting ad, clicks it and buys. GA4 credits the ad. The newsletter arguably started the whole thing. Fix: use your form field data alongside GA4 to get the assisted picture. They complement each other.

If you mention your website in a podcast, create a UTM link and a redirect (e.g. yoursite.com/podcast) that forwards to the UTM URL. Then in the episode notes, link to the redirect. Every click from that podcast episode now tracks correctly in GA4.

In GA4, 'Google / organic' includes people who searched your name AND people who found you via a keyword. These are very different signals. People searching your name already know you. People finding you via a keyword are new. Set up a custom dimension or simply note: when branded search converts, it is trust closing the deal, not SEO. When non-branded converts, SEO is working.