Case studies

Real projects: the problem, what we built, the result and its limits. Published only with the client's approval.

3published
3industries
3in review

[x] ptshigher repeat-purchase rate
E-commerceChurn analyticsSegmentation

An online store now spots customers going quiet before they are gone

A "lapsed" rule per product category, value segments and a weekly win-back list for the store's email tool.

Customer & revenue analyticsRead
[n] daysof reporting removed monthly
SaaS & startupsKPI frameworksBI dashboards

A SaaS company stopped reconciling MRR before every board meeting

Agreed metric definitions, one SQL data model and Power BI dashboards that replaced a hand-built board pack.

Data analytics & BIRead
[x]%of churn flagged in advance
Media & streamingChurn analyticsSegmentation

A streaming platform now sees churn coming [x] weeks earlier

One churn rule per plan, behaviour segments and a weekly at-risk list for a national subscription video platform.

Customer & revenue analyticsRead
[x]%budget moved to best segmentsComing soon
Media & streamingCampaign analytics

Ten SMS campaigns compared on what they actually sold

A campaign log joined with purchases and voucher use, and one dashboard comparing every campaign.

Marketing analyticsIn review
[n]reports consolidatedComing soon
Media & streamingPower BI consulting

A governed reporting portal replaced scattered report folders

A new folder structure by team, naming rules and one entry point for leadership reports.

Data analytics & BIIn review
[n]competitors benchmarkedComing soon
Media & streamingPricing analytics

A competitor price ladder for subscription plan decisions

Plan prices benchmarked against competitors, with a ladder the pricing team uses for every change.

Customer & revenue analyticsIn review
No case studies here yet

We haven't published work in this combination. Try another filter, or ask us about it directly.

How we write case studies

Four rules, so you can trust what you read.

1Client-approvedEvery number, screenshot and descriptor is approved in writing before publishing.
2Numbers, not adjectivesResults have a figure and a timeframe, or they don't go in.
3Limits includedEach case says what the work could not prove.
4Nothing inventedAnonymous clients get an honest descriptor, never a made-up name.

Want to be the next case study?

Start with a 30-minute call, or a two-week Data Health Check at a fixed $1,800.