A complete picture of hardware and ad performance
Objective
Extend Carista’s analytics capability to encompass hardware sales and paid advertising performance.
Obstacle
No visibility into unit sales by market, no ad performance tracking, and no way to connect hardware revenue to the subscription.
Outcome
Delivered a fully integrated Amazon sales and ads dashboard within four weeks, covering hardware units sold, ad spend, PPC performance, and exchange-rate-adjusted revenue across all markets.
Background
Carista is a leading provider of aftermarket car services, empowering drivers with advanced diagnostics, maintenance tracking and personalised vehicle insights. With an ambition to double its active subscriber base within a year, Carista needed stronger data foundations to understand customer behaviour and optimise subscription performance.
Carista’s subscription business is fundamentally downstream of hardware sales. A customer buys an adapter, connects a vehicle, and then converts to a paid subscription. Without Amazon sales data, the analytics infrastructure had a structural blind spot at the very top of the funnel. Revenue was being reported, but only the software half of it.
Challenges
No Hardware Visibility in a Hardware-Led Business: Carista’s subscription business is fundamentally downstream of hardware sales. A customer buys an adapter, connects a vehicle, and then converts to a paid subscription. Without Amazon sales data, the analytics infrastructure had a structural blind spot at the very top of the funnel. Revenue was being reported, but only the software half of it.
Fragmented Ad Spend with No Performance Measurement: Carista was running Amazon Sponsored Products, Sponsored Brands, and Sponsored Display campaigns across multiple markets but had no consolidated view of spend, clicks, or attributed sales. Campaign data lived only inside Amazon’s native reporting, making it impossible to compare efficiency across markets or ad formats, and impossible to connect ad spend to downstream subscription revenue.
Multi-Market Complexity and Currency Risk: Operating across UK, US and European markets meant revenue was denominated in GBP, USD, and EUR. Without exchange rate normalisation, any cross-market revenue comparison was unreliable. Additionally, Amazon’s marketplace IDs required mapping to recognisable country names, and the attribution windows for PPC reporting needed to be standardised, decisions with a direct impact on reported performance.
Solution
Fivetran-Powered Data Ingestion: We introduced Fivetran as the extraction and loading layer for Amazon data, connecting both the Amazon Selling Partner (SP) API and Amazon Ads API. We also ingested daily exchange rate data via Open Exchange Rates through Fivetran, combined with a historical rates backfill via Google Finance functions, enabling accurate currency conversion across all markets.
Sales and Ads Data Modelling: With raw data flowing into BigQuery, we built dbt models to clean, transform, and structure both datasets. For Amazon SP, this included units sold by adapter type and market, promotion and discount handling, and revenue calculation as item price minus promotional discount. For Amazon Ads, we modelled spend, clicks, impressions, and PPC sales across all three ad types, aligning on a 14-day attribution window after reviewing Carista’s existing reporting conventions.
Impact
Sales and Ads Data Modelling: With raw data flowing into BigQuery, we built dbt models to clean, transform, and structure both datasets. For Amazon SP, this included units sold by adapter type and market, promotion and discount handling, and revenue calculation as item price minus promotional discount. For Amazon Ads, we modelled spend, clicks, impressions, and PPC sales across all three ad types, aligning on a 14-day attribution window after reviewing Carista’s existing reporting conventions.
Ad Performance Tracking Across All Markets: Carista could now see spend, ROAS, click-through rates, and attributed PPC sales for every ad type across all markets in one place. Comparisons between the UK and US, or between Sponsored Products and Sponsored Brands, became straightforward, replacing the manual process of pulling individual Amazon reports.
Standardised, Trustworthy Revenue Numbers: With daily exchange rate conversion applied consistently across all markets, cross-currency revenue comparisons became reliable. Known discrepancies between dashboard figures and Amazon’s native reports, including attribution window differences and data retention limits on Amazon Ads were surfaced transparently, building trust in the numbers.
Creating Value For Carista...
From zero to live hardware & ads dashboard in just 4 weeks ,
Tracked data in 4 markets, across UK, US, DE, ES & IT,
And delivered a complete picture of how hardware sales influenced software subscriptions.
Get In Touch
Our friendly team are always on hand to answer questions, troubleshoot problems and point you in the right direction.
