Custom Marketing Dashboard

Problem

A sporting goods e-commerce company was paying $2,800 a year for Whatagraph, an off-the-shelf marketing reporting tool. The numbers were often a day or two old, sometimes wrong, and the team was stuck with whatever layout the vendor decided to show. The bigger problem: the things leadership actually wanted to see (how the month was tracking against plan, period-over-period comparisons that handled the company’s specific calendar, Amazon-side performance) weren’t even possible in the off-the-shelf product.

Approach

Built a custom marketing dashboard from scratch. A Python pipeline pulls daily from 9 platforms: Shopify, Meta Ads, Google Ads, Microsoft Ads, Reddit Ads, Amazon Ads, Klaviyo, GA4, and Amazon Seller Central. Each source is verified against the platform UI. Two GitHub Actions jobs refresh everything every morning, and the finished page is served from Cloudflare Pages behind a password.

Some metrics (like Meta reach or Klaviyo comparisons) can’t just be added up day by day. For those, small cloud functions query each platform’s API live for the exact date range picked, so the number is always right.

The big upgrade over the old SaaS: this is fully custom. Anything leadership wants to see can be added. The Trends tab has 10 charts with a Trellis-style plan overlay so you can see at a glance how the month is tracking against budget. Period comparisons (year-over-year, prior period, custom range) work on every metric. The Top 10 Amazon products block updates monthly. New metrics or new views are quick to add.

Result

• $2,800 a year saved in software cost, with a running cost under $5 a month
• Over 17,000 daily metric values tracked across 9 platforms, refreshed every morning
• Plan-vs-actual overlay lets leadership see at a glance whether the month is on track
• Full control: new metrics or new views take one commit
• Leadership uses the numbers in weekly marketing reviews and trusts them