isiko
what we do packages contact →

VAVO Group.
From 37 stores to 130.

Operationally restructuring a retail holding in fast growth. From WhatsApp with 180 people in one channel to Discord per-branch, from verbal „that's how we do it here" to written SOPs. 130 branches, 300 employees, six layers built one on top of the other.

— client
VAVO Group
— scale
130 branches · 300 people
— role
Operations Director
— time
2024 (entry) → ongoing

01 · contextClient: a multi-brand retail holding in a phase of fast growth.

VAVO Group is a Polish capital group with own brands in retail. I joined VAVO in 2024 as a regional manager when the group had 37 stores. Today there are 130 of them, with around 300 employees, across several operational regions. Three promotions in twelve months — regional manager → manager → Operations Director. Today I am responsible for operations across the whole group.

Writing this case study, I'm looking at the present — a system that works. But it started with chaos that doesn't differ from what I've seen in other 100+ person companies in Poland. That's why this case is meaningful as proof: not as a consultant from outside, but as the person who lived through it and answers to the board for the result.

02 · state on entryEvery branch played by its own rules.

When I first walked into VAVO, I saw what I'd already seen in other 100+ person companies — operational chaos that the company had normalised, because nobody had ever had time to clean it up:

System stock ≠ shelf stock. Our Excel keeps drifting. Every department has its own version of the truth. — familiar sentences from the first conversations

03 · chronologySix steps, in the same order.

Operationally cleaning up a 100+ person company is not about „deploying an ERP". It's about building layers, one on top of another, in a fixed order. Skip a layer = the system doesn't work, even if all the files are in place.

01 · 2024

Operational procedures + leave — the first SOPs.

I started with what was cheapest and fastest. The first written SOPs: cash, warehouse, returns, complaints, onboarding. Leave — the process for requesting, approving, team calendar. Not perfect — sufficient. Every procedure had an author, a date, a version. The reason: without procedures there's nothing to automate.

02 · 2024

Discord — network communication.

Migration from WhatsApp (180 people in one channel) to Discord with channels per-branch, per-region, per-role. Everyone has their channel. Everyone sees only what concerns them. HQ posts once and it reaches the right branch. The rest — history, tags, search. The relief was visible in the first week.

03 · 2025

First SharePoint — single source of truth.

M365 SharePoint as the place where everything lives: procedures, reports, templates, training. Every employee knows where to look. Every manager knows where to publish. End of sending attachments by email — link, version, change history. The first attempt to bring the company into a documented rhythm.

04 · 2025

Power BI — board dashboards.

For the first time in VAVO's history, the board had a dashboard that doesn't lie and doesn't require manual compilation. KPIs per region, sales per branch, margin, operational costs — all in one place, refreshed automatically. Decisions based on numbers, not on who called louder.

05 · 2025

VRS — the first in-house system.

Vape Retail System — an in-house platform to run the retail network. Shift schedules, attendance, sales per-branch, alerts. Built around VAVO's specific processes, not abstract needs. The decision to build in-house instead of buying SaaS was deliberate — too many specific cases for SaaS to cover without compromise.

06 · 2025

VRS extended — analytics and KPIs.

After the daily operations layer was deployed, VRS expanded into the analytical layer: regional comparisons, KPIs per-branch, deviation alerting. Data flowed in by itself — nobody had to gather it once a week. Power BI stays for the board, VRS — for daily operations.

07 · 2025–2026

Warehouse — WMS, location move, cost reduction.

WMS deployed at the central warehouse. Move to a new location — larger floor space, better logistics, lower operating cost. Throughput increased (more parcels/day with the same team). Warehouse procedures written, measured, optimised. An operation that used to be a bottleneck became predictable.

04 · stackWhat sits under the layers.

— Operations layer
SharePoint (M365) OneDrive Outlook Teams
— Custom system (VRS)
Supabase → Postgres Keycloak (SSO) S3 (storage) SvelteKit (frontend) Power BI (reports)
— Communication
Discord (per-branch) Outlook (HQ) Teams (board)

The stack is not the goal — it's a tool. Each layer answers a specific need: SharePoint = documents, VRS = daily operations, Discord = team communication. Swapping one tool without rebuilding the process = swapping a piece of furniture in a kitchen where cooking still has no rhythm.

05 · effectWhat changed — measurably.

— communication
From 180 people in 1 WhatsApp channel to Discord per-branch. A message from HQ reaches its target in 1h instead of drowning in noise. An employee sees only what concerns them.
— onboarding
From 2 weeks of asking a colleague to 1 day. Written SOPs + SharePoint access + VRS access = a new hire is operationally ready on day one.
— board data
For the first time in VAVO's history: a KPI dashboard per-region per-week. Board decisions based on numbers, not on who called louder.
— operational rhythm
130 branches running in one rhythm. Shift schedules, reports, escalations — in one cadence, one template, one flow.

06 · lessonsThree things I'd do differently (and one I'd do the same).

  1. I enforced the first cleanup more weakly than I should have. The first procedures were written down, but I didn't enforce them consistently across everyone. Loose pockets remained. I should have done a hard cut-over from day one: new process in, old process out. Fewer exceptions, less straightening up later.
  2. I kept two solutions running in parallel during transitions. Old tool + new tool for several weeks — on one hand, prudent (safety, ability to roll back); on the other, later some of the team didn't want to switch. Two working solutions = natural resistance to one. Maybe next time a shorter transition window and a hard cut-over date.
  3. Communication of changes and rollout UX were weaker than the substance of the changes. The SOP, the system, the dashboard — each was OK in itself. What lagged: broad explanations of „what this means for you", better UX for the warehouse worker / cashier, short instructional videos. The best system without UX = a system nobody uses. That's what I'd do differently.

Order is everything. Procedures → communication → source of truth → analytics → system → warehouse. Skip a layer and the system doesn't work. — what I'd leave unchanged

Main lesson: AI won't solve operational chaos. Procedures, data, communication rhythm — that's what creates the structure on which AI then makes sense. Most companies of 50-300 people buy AI before they have SOPs. The result: AI in chaos = automation of chaos. Skeleton first, muscles after.

Does your company have its own „180 people in one channel"?

Operations audit · 7 days — process map, list of 5-10 losses, 3 quick wins, 30-60-90 day plan, direction recommendation (AI vs ERP vs processes). No commitment to further engagement. Pricing set after a short conversation, once I see the scale of the company.

Karol Zamarlik
isiko · Operations Director, VAVO · Fractional CTO/COO
© 2026 isiko · Karol Zamarlik · Krakow · case study #01 / VAVO Group ← back to home