Building a tablet operations system and multi-store analytics site for a beauty-salon chain

What changed

• Input time on store devices
  Shortened versus paper operation
• Multi-store metric visibility
  Checkable daily at the head office
• Stores rolled out
  Expanded through phased rollout

Challenge

In a multi-store salon chain, workflows differ subtly from store to store, while the head office wants to aggregate cross-store metrics to inform management decisions. On top of that, the tablets are used by on-site estheticians — many of them not comfortable with IT — so unless the device feels as natural as their paper routine, it won't take hold. We needed a design that reconciled what is normally a trade-off: "tolerate store-by-store variation while keeping the head office's aggregation axes consistent."

Approach

1. 01 Through store interviews, separated what could be unified in the workflow from what could not
2. 02 Implemented the tablet operations app with C#.net + WPF, designing an input UI optimized for the front line
3. 03 Built the multi-store operations analytics site with CakePHP + MySQL
4. 04 Designed the link between store devices and the head-office database, structurally cutting head-office aggregation effort
5. 05 Agreed with the head office on the visualization axes for management metrics (sales, treatment time, repeat rate, etc.)
6. 06 Brought stores online in sequence through phased rollout, confirming adoption before expanding to the next store

Key success factors

Judging what can be unified

We separated what differs by store from what should be unified, and made it explicit as a design premise.

A tablet-optimized UI

We committed fully to a tap-driven input UI so it would take hold even at sites unaccustomed to IT.

Separating store and head-office responsibilities

We split responsibilities between the on-site app and the head-office analytics site so neither side's constraints distorted the other.

Agreement on management-metric axes

We agreed on the aggregation axes with management first, keeping the metrics from drifting as features were added.

Solution

On the tablet side we built a UI optimized for the front line with C#.net + WPF, and on the head-office side we implemented multi-store analytics with CakePHP. By linking data from store devices to the head-office database, we switched the head office's aggregation work from manual to automatic. We achieved both a parameter design that tolerates store-by-store variation and unified aggregation axes for the head office.