Hej! Let's talk.

Product Design Lead in Copenhagen. Happy to chat about work, design or a good coffee spot.

Stefan Cioroiu's resume: Product Design Lead with over 10 years of experience, based in Copenhagen. Senior Product Designer at Veo Technologies, earlier Senior Product Designer at Novo Nordisk, Senior then Lead Product Designer at Planday, and Product Designer at Tradeshift, Founders and Zigna, starting as a UI and graphic designer at Me-Mover. Studied at KEA, Copenhagen School of Design and Technology.

4 min read B2B · Research · Data · Iteration

Business Metrics: Scheduling to Real Demand

Two large customers asked for two tools to staff against demand. We built one feature that brings business data and formulas into the schedule.

Context
Planday • Web / Analytics
Role
Senior Product Designer
Scope
Discovery, data model, reports
Status
Shipped 2020

Skip to the solution

The brief

Two requests, one underlying need

Randstad asked for “demand planning”. Meanwhile, another team scoped “business metrics” for Scandic, a major hotel deal.

Both landed with me. Same problem: managers wanted to staff by numbers, not gut feeling.

What I’d do differently: mocking up Randstad’s solution anchored us too early. Today I question the brief first.

An early mockup of the demand planning feature Randstad asked for
The first mockups, built around the customer’s proposed solution

Research

Bigger than two customers

Surveys and interviews showed a broad need, strongest in the UK, split between two kinds of users.

Most of our customers

Busy managers

  • Small and medium businesses
  • Scheduling is one of many jobs, alongside supervising and running the place
  • Need it simple, quick and easy to learn

Our largest accounts

Data specialists

  • Larger businesses like Scandic and Randstad
  • Hired to work with data, some of them developers
  • Need flexibility above everything else

What else we learned

  • Businesses plan in fixed periods that rarely change.
  • Comparing periods matters, for example this Christmas against last.
  • Customers wanted Planday data in their calculations: revenue, costs, overtime, absence.
It would be great if we could see the number of bookings per hour and compare it with the number of employees on shift.
Hospitality customer, interview

Designing the model

One model for two very different customers

The Scandic deal had a fixed deadline. For weeks, every discipline worked the flows at one whiteboard.

Whiteboard sketches of the creation flow and overview screens
Whiteboard flow diagram for creating, inputting and displaying a metric
Whiteboard notes on formulas, data and display options

Fitting Planday’s structure

Randstad needed daily demand per position, Scandic hourly data per section, both in the existing structure.

So a metric got five properties: data or formula, hourly or daily, scope, unit and visibility. Both customers’ stories tested the model.

Diagram of a schedule with departments split into sections and positions
Four metric cards: Number of rooms, Hours budget, Distribution and Comparison, each with type, time stamp, location, unit and display
Scandic housekeeping: rooms × 15 minutes gives a daily budget, spread across the day and compared to the schedule.

InteractiveDrag the sliders and watch where the schedule falls short.

Hours budget = rooms × minutes ÷ 6030.0 hours needed · 24 scheduled

Hourly distribution vs scheduled Needed Scheduled Short

8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00

The first release

From a number to a better schedule

Demand shows right where shifts are planned.

The week schedule with an hours prognosis and comparison row above each housekeeping section

Demand next to the shifts

Daily totals and hourly bars in each section header.

The day schedule with an hourly comparison row showing negative and positive numbers

Hour by hour

The day view shows staffing gaps hour by hour.

A spreadsheet-style grid of shifts needed per position and day

Enter the numbers like a spreadsheet

Values go into a familiar spreadsheet-like grid.

Step 3 of 3 of the Create business metric flow: input per day or hour, assign to positions, display in Schedule

Set up a metric in three steps

Name it, pick data or formula, resolution and placement.

Testing and launch

Shipped on time, with known gaps

We tested the prototype, but the deadline pushed most findings to the backlog. We launched in spring 2020 and let Mixpanel set priorities.

A rainbow sheet of usability test observations, colour-coded by participant
Usability test observations, one colour per participant
A Mixpanel dashboard tracking business metric usage
The Mixpanel dashboard we used to track adoption

After launch

Iterating on real usage

Usage data drove the next improvements:

A line chart of planned versus needed hours above the schedule

Space versus numbers

We tested lighter charts, but users kept the plain numbers next to their shifts.

The week schedule with small hourly bar charts above each day

Hourly peaks in the week view

Hourly charts in the week view, where managers actually work.

A NaN value in the schedule with a tooltip explaining the formula divides by zero

Honest empty values

Broken formulas showed nothing. Now they show zero, as managers expected.

The formula field with an autocomplete list of available metrics

A formula field you can type in

Autocomplete and keyboard support for formulas.

A dialog for managing time blocks for a metric

Time blocks

Time blocks match the fixed periods customers plan in.

The redesigned Data Center edit screen with general settings and time grouping

A simpler Data Center

Simpler setup, better accessibility, and in-product guidance.

Looking back

Powerful, but formulas are still the barrier

It grew into Planday’s Data Center. But formulas still need near-developer skills, which leaves out most managers. A colleague took on a visual formula builder next.

Next time: question the brief on day one and agree on success metrics up front.

Thanks for reading!