I spent the last week building something I've wanted since day one of the DTC Stack: a way to feed live analytics data into every skill automatically.

The trigger was a post from David Dokes, co-founder of Polar Analytics, about deploying 10 AI agents that pull live ecommerce data every Monday morning. His system uses Polar's MCP server to connect Claude to real store metrics - revenue, ad spend, ROAS, email performance - instead of having agents guess or hallucinate numbers.

That post validated something I'd been thinking about. The DTC Stack already has 20 skills that know your brand. What if they also knew your numbers?

So I built the integration. Here's how to connect Polar Analytics to the DTC Stack's Customer Intelligence Engine, what data flows through, and what it actually changes about your skill output.

What This Integration Does

The DTC Stack has two intelligence layers. Your Brand Brain captures who you are - voice, positioning, personas, objections, guardrails. Your Customer Intelligence file captures what's happening - revenue trends, email performance, product velocity, customer feedback.

Until now, the Customer Intelligence layer was mostly manual. You'd paste metrics, export CSVs, or connect individual platforms one by one through Composio. It worked, but it was friction.

Polar Analytics changes that. One MCP connection and the Customer Intelligence Engine pulls store performance, email/SMS data, product intelligence, and trend analysis directly from Polar's aggregated data layer. Polar already connects to your Shopify store, Meta ads, Google Analytics, Klaviyo, Amazon, and 40+ other sources. So instead of wiring up each platform individually, you wire up Polar once and get all of them.

The result: when you run any DTC Stack skill, it reads from live data instead of assumptions or stale context.

What Changes in Practice

Here's the before and after.

Before Polar: You run the CRO Audit skill. It asks for your conversion rate, AOV, top products, and traffic sources. You either paste them from your Shopify dashboard or type them from memory. The audit is only as good as what you remembered to include.

After Polar: You run the CRO Audit skill. It reads your Customer Intelligence file, which was refreshed this morning from Polar. It already knows your conversion rate dropped 12% last week, your AOV shifted from $67 to $61, and your top product changed from the Starter Kit to the Refill Pack. The audit starts from what's actually happening, not what you remember.

Same thing happens with Ad Creative, Klaviyo Flows, Product Pages, Retention Optimizer - every skill that reads Customer Intelligence gets sharper because the data is real and current.

Setup: 5 Minutes, 4 Steps

Step 1: Make Sure Polar Is Ready

You need a Polar Analytics account with at least one data source connected (Shopify at minimum). If you're new to Polar, sign up at polaranalytics.com - there's a free 7-day trial. The initial data sync takes up to 24 hours, so do this first.

Step 2: Add the MCP Server

Open your .mcp.json file in the DTC Stack project root and add the Polar entry:

{
  "mcpServers": {
    "polar": {
      "url": "https://api.polaranalytics.com/mcp"
    }
  }
}

If you already have other MCP servers configured (keywords-everywhere, composio), just add the "polar" block alongside them. Restart Claude Code after saving.

That's it for config. Polar's MCP is a remote server - no npm package to install, no local dependencies.

Step 3: Authenticate

Run the Customer Intelligence Engine skill: /customer-intelligence

When it connects to Polar's MCP server for the first time, you'll get a browser-based OAuth prompt. Log in with your Polar account and grant access. This only happens once.

Step 4: Pull Your Data

Say: "Pull my customer intelligence"

The agent will detect Polar as the highest-priority data source, query it for everything it covers, and write the results to shared/customer-intelligence/CUSTOMER_INTELLIGENCE.md. Each section gets a source tag showing where the data came from:

## Store Performance Snapshot
> Source: Polar Analytics | Refreshed: 2026-04-08

You'll also see which sections Polar doesn't cover - more on that below.

What Polar Covers (and What It Doesn't)

Polar is strong on quantitative metrics. It handles four of the six Customer Intelligence sections:

Section	Polar?	What Flows In
Store Performance	Yes	Revenue, AOV, conversion rate, top products, MER
Email/SMS Intelligence	Yes	Flow performance, campaign stats, list metrics (if Klaviyo is connected in Polar)
Product Intelligence	Yes	Best sellers, trending products, inventory velocity
Trends & Signals	Yes	Cross-channel trend analysis, seasonal patterns
Voice of Customer	No	Reviews, support tickets, survey responses
Retention Data	Partial	Some metrics; Kleio is deeper for LTV and cohorts

The gap is qualitative data. Polar doesn't read your product reviews or support tickets. For Voice of Customer - what customers love, complain about, and ask before buying - you still need a direct connection to your review platform (Junip, Yotpo, Judge.me) or CX tool (Gorgias, Zendesk).

This is actually a clean separation. Polar handles the numbers. Your review platform handles the words. Both feed into the same Customer Intelligence file, and every skill reads from both.

How It Affects Each Skill

Every DTC Stack skill checks for Customer Intelligence before it runs. With Polar connected, here's what changes:

CRO Audit gets real conversion rates, actual AOV trends, and current top products. Revenue impact estimates go from guesses to math.

Ad Creative sees which products are actually selling, what your current ROAS looks like, and where your ad spend is going. Creative briefs start from performance data, not assumptions.

Klaviyo Flow Architect reads your actual email metrics - which flows drive revenue, what your open rates look like, where engagement is dropping. It doesn't suggest building flows you already have.

Product Page Engine knows your best sellers and rising products. It prioritizes the pages that will move revenue.

Retention Optimizer sees churn signals in the data before you notice them in the dashboard.

The pattern is the same across all 20 skills: less guessing, more context, better output.

The Source Attribution System

One thing I built specifically for this integration: per-section source tracking. Every section in your Customer Intelligence file now shows exactly where its data came from and when it was last refreshed.

## Email/SMS Intelligence
> Source: Polar Analytics | Refreshed: 2026-04-08

## Voice of Customer
> Source: Yotpo via REST API | Refreshed: 2026-04-01

Skills check freshness per section, not per file. So if your Polar data is from this morning but your review data is a week old, the skill only warns you about the stale reviews. It doesn't block on data that's perfectly fresh.

What If I Don't Have Polar?

The DTC Stack works the same way it always has. Every skill runs without live data - Polar is an enhancement, not a gate.

Your options without Polar:

• Composio MCP - connect Shopify, Klaviyo, and other platforms individually
• REST APIs - provide API keys for direct platform access
• Manual paste - export data and paste it when prompted

And one more thing: even without any analytics connection, you can still make every skill smarter by logging your winning ads, emails, and product page copy in brand/proven_winners.md. Skills reference it automatically. Live data makes output better, but documented wins compound too.

FAQ

Do I need Polar Analytics to use the DTC Stack?

No. Every skill works without Polar. You can paste data manually or connect individual platforms like Shopify and Klaviyo via Composio. Polar is an enhancement that makes setup faster and data richer.

What does Polar Analytics cost?

Polar is a separate subscription (not included in the DTC Stack). They offer a free 7-day trial. Check polaranalytics.com/pricing for current plans.

Can I use Kleio instead of Polar?

Yes. The Customer Intelligence Engine supports both. Kleio is a Shopify app ($29/mo) that's stronger on unit economics - P&L, product margins, cohort LTV, CAC. Polar is broader, covering 45+ data sources across channels. Pick the one that matches how you think about your data. See the Kleio setup guide for details.

How often should I refresh?

Weekly is the sweet spot for most brands. Polar syncs daily, so your data is always current on their end. Run /customer-intelligence every Monday morning and your skills have fresh context all week. You can also set up automatic weekly refreshes with Claude Code's /schedule feature.

Does Polar replace my Shopify and Klaviyo connections?

For the Customer Intelligence Engine, yes - Polar covers what those individual connections would provide. You might still want direct Shopify access via Composio for skills that write to your store (like Shopify Page Builder), but for reading analytics data, Polar is all you need.

Get Started

If you already own the DTC Stack and use Polar Analytics, this is a 5-minute setup that makes every skill meaningfully better. Add the MCP server, authenticate, pull your data, and run any skill.

If you don't use Polar yet, the free trial is enough to test the integration with your real store data.

And if you don't own the DTC Stack yet - it's 20 AI marketing skills that know your brand. Polar makes them smarter. But they're already pretty sharp without it.