Full Marketing Analytics Stack for Under $50/Month

Most small businesses are flying blind. They have Google Analytics installed, they check the sessions number once a week, and they call it "data-driven."

Most agencies have the opposite problem: they're paying $2,000/month for tools they use 10% of.

There's a middle ground. Here's what it looks like, what it costs, and exactly how to build it.

The Stack

Total: $0/month for most small businesses, with Supabase Pro at $25/month if you're storing a serious volume of events.

The money-saving insight: you don't need Hotjar Pro, Mixpanel, or Segment to answer the three questions that actually matter:

1. Where are leads coming from?
2. Where are they dropping off?
3. What does a converting session look like vs a non-converting one?

Setting It Up

Step 1: GA4 with server-side conversion events

Browser-based GA4 tracking misses roughly 20-40% of conversions in 2026 due to ad blockers and ITP (Intelligent Tracking Prevention) in Safari.

Fix this with a server-side event for form submissions:

// app/api/lead/route.ts
import { track } from "@/lib/analytics"; // wrapper around GA4 Measurement Protocol

export async function POST(req: Request) {
  const data = await req.json();

  // Insert lead to DB
  await supabase.from("leads").insert(data);

  // Fire server-side conversion event
  await track({
    client_id: data.ga_client_id, // passed from browser
    name: "generate_lead",
    params: {
      source: data.utm_source,
      medium: data.utm_medium,
      campaign: data.utm_campaign,
      value: 1,
      currency: "AED",
    },
  });

  return Response.json({ success: true });
}

This fires a generate_lead event directly from your server to GA4's Measurement Protocol, bypassing the browser entirely. It won't get blocked.

Step 2: Capture GA4 client ID in your form

The server needs the GA4 client_id to attribute the conversion correctly:

// Get GA4 client ID from browser
function getGa4ClientId(): Promise<string> {
  return new Promise((resolve) => {
    gtag("get", "G-XXXXXXXXXX", "client_id", resolve);
  });
}

// Add to form submission
const clientId = await getGa4ClientId();
formData.ga_client_id = clientId;

Step 3: Microsoft Clarity for behaviour data

Install Clarity's script (one line of code, or via GTM) and you immediately get:

• Session recordings for every visitor
• Heatmaps by page
• Rage clicks and dead clicks flagged automatically
• "Smart Events" that detect frustration patterns

It's free, GDPR-compliant, and honestly better than Hotjar's free tier.

For each session with a conversion, Clarity lets you tag it and replay exactly what that user did. This is how you find out what your converting users have in common.

Step 4: Custom Supabase events table

For anything that GA4 can't capture cleanly — CRM stage changes, email opens, WhatsApp replies — I push custom events to a Supabase table:

create table custom_events (
  id uuid default gen_random_uuid() primary key,
  event_name text not null,
  properties jsonb,
  session_id text,
  created_at timestamptz default now()
);

This becomes a queryable event log you fully own. No sampling, no 90-day retention limits, no $300/month bill.

You can see this exact events table used in a live lead routing system in Real-Time Lead Routing CRM with Next.js & Supabase.

Step 5: Looker Studio dashboard

Connect Looker Studio to:

• GA4 (official connector)
• Search Console (official connector)
• Supabase (via the PostgreSQL connector or a simple API endpoint)

Build one dashboard with four sections:

1. Traffic — sessions, source/medium breakdown, top landing pages
2. Conversions — goal completions, conversion rate by source, cost per lead (if you import ad spend)
3. Behaviour — top pages by engagement, scroll depth, bounce rate
4. SEO — clicks, impressions, average position, CTR from Search Console

Schedule it to email to your client every Monday morning. They will think you're a genius.

The One Metric That Changes Decisions

Everything above is infrastructure. But the single metric I watch most closely for any lead-gen business is:

Cost per qualified lead by source

Not total leads. Not traffic. Qualified leads — ones that turn into conversations — segmented by where they came from.

A Facebook lead at AED 50 that converts at 2% is more expensive than a Google lead at AED 150 that converts at 12%.

Most clients have never seen this number calculated. When you show it to them, it instantly clarifies where to put budget.

What This Stack Can't Do

To be honest: if you're running a complex SaaS product with multiple pricing tiers, user roles, and a 90-day sales cycle, you'll outgrow this. You'll want Segment, Amplitude, or PostHog.

But for a service business, agency, or property developer doing under $5M/year in revenue? This stack answers every question you'll actually ask.

---

If you're using this stack to measure paid ad performance, the full landing page and UTM setup that feeds it is in The Landing Page That Fixed Our Google Ads ROI.

Need a proper analytics setup built for your business? Let's talk.