Blinko
All case studies
Case study · Fractal · United States

How Fractal tracks $100M+ in yearly transactions and catches unhealthy clients early

We built a health score model that separates healthy clients from unhealthy ones and alerts account managers and sales reps, inside one live system for every role.

in yearly transactions tracked in the system we built, not revenue generated or saved
$100M+
in yearly transactions tracked in the system we built, not revenue generated or saved
What changed

Before and after

  • Before: Numbers came from spreadsheets and days-old exports

    After: Every role sees live numbers

  • Before: Client problems showed up late, in reports

    After: A health score flags unhealthy clients and alerts account managers and sales reps

  • Before: Forecasting was guesswork

    After: Forecasts are part of daily decisions

The problem

Leadership could not see revenue, sales activity, or problems as they happened. Teams worked from old spreadsheets and waited days for reports, so decisions came late.

What we built

A health score model that separates healthy clients from unhealthy ones, inside one live system across sales, payments, and operations. When a client turns unhealthy, account managers and sales reps get a notification, and AI explains what changed and where. Executives, sales reps, and operations managers each have a view built for their role.

  • Client health scores
  • Alerts to account managers and sales reps
  • Connected data
  • Live views by role
  • AI alert explanations
Who used it

Executives, account managers, sales reps, operations managers and field reps, each with a view built for their role.

Illustration with sample figures, not Fractal's data: live revenue and sales activity in one place.
Illustration with sample figures, not Fractal's data: live revenue and sales activity in one place.
How it works

How a health alert reaches the right person

AI handles 1 of 5 steps. People make the final call.

  1. 1
    Automation

    Connect

    Sales, payments, activity, and revenue data flow into one place.

  2. 2
    Automation

    Agree

    Everyone works from the same agreed numbers, updated live.

  3. 3
    Scoring

    Score

    The health score model separates healthy clients from unhealthy ones and spots unusual changes as they happen.

  4. 4
    AI (LLM)

    Explain

    AI explains each alert: what changed and where.

  5. 5
    Your team

    Act

    Account managers and sales reps get the alert and deal with it.

The results

The bottom line

The system tracks more than $100M in transactions a year and has run the business every day for over a year. Account managers and sales reps get an alert when a client turns unhealthy, instead of finding out from a report.

  • $100M+ is the transaction volume the system tracks each year. It is not revenue we generated or saved.
  • These results describe Fractal's own use of the system. They are not a forecast for another business.

Blinko delivered exceptional work on a highly complex problem. Deep expertise in data science, machine learning, and business domain knowledge – exceeded all expectations.

Steve Conn · Analytics Director, Fractal & Sky Systemz
Illustration with sample figures, not Fractal's data: tailored views for executives, sales reps, and operations.
Illustration with sample figures, not Fractal's data: tailored views for executives, sales reps, and operations.

Next case study

Consumer fintech

Solventis doubled its assessment capacity without hiring

Read it

Want results like these?

In a 30-minute call we look at the process you want to fix and tell you honestly what it would take.