Case study · Retail & E-commerce · Product Engineering + BI

A deals webshop rebuilt — and taught to read its own numbers.

For Hópkaup, Iceland's group-buying webshop, Moberg rebuilt the shop end to end, then built the reporting layer that tells the business what each deal actually earns — before the campaign runs.

The client & the project

A deals business lives or dies on its margins.

Hópkaup, part of Kolbeinn Kafteinn, is an Icelandic deals and group-buying webshop — a model comparable to daily-deals platforms, where campaigns launch, sell and expire on short cycles and the margin on each deal decides whether the model works. It is retail with the tempo turned up.

Moberg's work came in two engagements: first, redesign workshops and full webshop development — backend, every connection and synchronisation, and the frontend; second, a Power BI reporting implementation, delivered by three FTEs over six months, that pulls the business's scattered data into one layer.

The solution

Rebuild the shop, then wire up its instruments.

The webshop rebuild started with redesign workshops and went all the way down: a .NET backend, all the connections and synchronisations that keep a deals shop consistent with the systems around it, and a Next.js frontend. The platform is now in maintenance — the phase where the quality of the build shows, or doesn't.

The second engagement made the business legible. The Power BI implementation combines three sources — SQL databases, Google Analytics and Zendesk — into one integrated reporting layer, with storytelling dashboards across sales, customer behaviour and support. Commerce, traffic and complaints stop being three separate stories told by three separate tools.

The part external partners notice: vendor reports with row-level security and automated email distribution. Each partner sees exactly their own deals and nothing else, delivered to their inbox without anyone at Hópkaup exporting a spreadsheet. And because the reporting layer ties it together, deal-level margin is known before a campaign runs — which, in a group-buying business, is the difference between pricing and guessing.

The challenges

The honest part.

Integration

Synchronisations everywhere

A deals webshop is only as reliable as its connections — inventory, orders and campaign state have to stay consistent across every system it talks to, on deal-cycle timing.

Data

Three sources, one truth

SQL databases, Google Analytics and Zendesk count different things in different ways. Reconciling them into dashboards people trust is the unglamorous majority of the BI work.

Security

Partners must see only their own numbers

Vendor reporting to external partners makes row-level security load-bearing: one misconfigured filter shows one vendor another vendor's business. That class of bug is not allowed to exist.

The value

What it bought Hópkaup.

Hópkaup runs on a webshop built for its actual model — deals, campaigns, partner inventory — rather than a generic e-commerce platform bent into shape. With the build now in maintenance, the engineering effort goes into keeping it steady and cheap to change. Launch day is the easiest day; the next ten years are the product.

The reporting layer changed the quality of decisions. Deal-level margin known before a campaign runs means pricing conversations happen up front, with numbers, instead of in the post-mortem. Sales, customer behaviour and support live in one set of dashboards, and vendor partners get their own reports automatically — scoped by row-level security, delivered by email, no manual work in the loop.

KPIs & numbers

The shape of this one.

End to end
full webshop rebuilt — backend, integrations, frontend
3
data sources in one reporting layer — SQL, Google Analytics, Zendesk
RLS
vendor reporting with row-level security, auto-distributed
Under the hood

The stack on this one.

Backend

.NET

Webshop core, plus the connections and synchronisations around it.

Frontend

Next.js

The customer-facing webshop.

Reporting

Power BI

Storytelling dashboards across sales, customer behaviour and support.

Sources

Google Analytics & Zendesk

Traffic and support data integrated alongside the SQL databases.

Security

Row-level security

Scopes each vendor's reports to their own deals, with automated email distribution.

← All stories

Get in Touch