Security stream of a CEE telecoms group separation

Context

A CEE mobile operator group had just changed hands. The new owner wanted to separate network infrastructure from retail, the same carve-out it had done in another market five years earlier. The infrastructure side became a large regional neutral-host tower and fibre operator.

Several operating companies had to be separated at the same time. Every corporate function had to land on one side of the new boundary, or on both under a transitional arrangement.

Mandate

Design and deliver the target operating model for the regional security organization through the separation. Multiple countries, multiple legal entities per country. Every capability had to be allocated: identity, network, SOC, incident response, GRC, physical. The transition had to be smooth, and both sides had to stand on their own on day one.

Four dimensions, designed and executed in parallel: the technical split, the financial model, the organization setup, and the legal inputs into the inter-company contracts.

Role

I was Regional Director at the group's regional shared-services entity, reporting to its CEO. That entity was restructured at the end of the separation, and I continued afterwards as Regional Security Director for CEE inside the new infrastructure group.

I owned the security stream end to end. I designed the target operating model myself, across all four dimensions, and I drove it to go-live: resourcing it, running the workstream, and carrying the regional security function as a day-to-day service while everything underneath it was being rebuilt.

The remit spanned several countries. In each country, multiple legal entities: the local NetCo, the local ComCo, and the regional shared-services entity above them. The remit didn't own the commercial negotiation between the retail and infrastructure sides, but every decision with a security angle fed into it.

Approach

The programme had a hard go-live date. My method was to get to a clean, defensible split first and accept some transitional inefficiency, then schedule optimization for the year after. The first test for any decision was "does this work on day one." The second was "does this scale."

I designed each dimension and drove it to delivery at the same time. The four dimensions ran in parallel and had to reconcile with each other every week.

On the technical side, I went capability by capability: identity, network segmentation, security monitoring, endpoint, vulnerability management, cryptography and key material, physical, OT, network management. For each one I decided whether it duplicated, migrated, or stayed shared under a transitional arrangement. Each one got a target state and a transition path.

Financially, I priced the steady-state cost of every capability on each side, the cost of duplicating it, and the price of any transitional service the shared-services entity would keep delivering. The model had to line up with the group's financial case for the whole separation.

Organization design came next. Target security organizations on both sides, down to roles, headcount, seniority mix, reporting lines, critical skills, and hiring sequence. For every role I decided whether an existing person could fill it and on which side they should land, or whether we had to hire.

Legal pulled it all into contract language. Security schedules, SLAs, liability, data-sharing, incident-cooperation clauses in the inter-company contracts and transitional services agreements. Of the four tracks, legal was the one that forced every decision to become explicit and signed.

Each market had its own operating company, its own regulator, and its own legacy. I designed one target model centrally and localized it per market with a variance register, rather than running per-market redesigns.

Then there was the entity structure. Each country had its own local NetCo and its own local ComCo, each a separate legal entity. A regional security function had to sit across all of them. The operating model had to carry through several contract layers: inter-NetCo agreements across the region, because the security team stayed regional while the NetCos were country-level, and commercial contracts between each NetCo and its local ComCo.

Regional security had to get paid under this structure. The cost allocation followed a provided-versus-consumed logic that had to reconcile across every pair of legal entities and sign off consistently against the local commercial contracts. That part was as much engineering as negotiation.

Security also had two roles inside the new architecture: securing the services delivered under the inter-company contracts, and providing security services directly to the retail side for the business applications that stayed with them. Both roles had to be priced, contracted, and run.

All of this had to land, not just look good on paper. Go-live was one date, across every market, with the security function still running.

Deliverables

The TOM itself. One document tying the technical, financial, organizational, and legal work into an end state and a transition path.
A capability-level security separation blueprint, with a target state and transition path for every in-scope domain.
A regional security cost-allocation model that reconciled across every legal-entity pair.
Target organization designs for both sides of the split, down to role definitions, headcount, seniority mix, and hiring sequence.
Security schedules and annexes for the inter-NetCo and NetCo-to-ComCo contracts.
A programme risk register for the security-specific risks through to go-live.

What made it hard

The commercial track moved faster than the technical facts. The boundary between retail and infrastructure kept shifting, and every shift reopened decisions already documented, modelled, and drafted into contracts. Every artefact had to be a living document, with one source of truth that the legal, financial, and technical tracks all pulled from.

The legal-entity structure multiplied the number of documents. A regional security function sitting across several country-level NetCos, each with its own ComCo and its own commercial agreements, meant every document had to be consistent with every other. Getting alignment was an engineering job on top of a negotiation: the cost allocation, the service scope, and the security schedules had to close at the same time.

The stakeholder map was wide. Every legal entity brought its own people, each of them with their own interests, their own culture, their own expectations. The upstream programme execution sat above the security stream with its own tempo and its own agenda. Alongside, my own team and my own line management needed a different kind of conversation again. Different interests, different cultures, different communication approach for each. Carrying the TOM through all of them, and then delivering against it, was as much a communications job as a design job.

Running in parallel with business as usual meant nothing could be interrupted for the separation. Changes to identity, monitoring, or network controls had to work for the pre-separation group and the post-separation entities at once.

Multi-country regulatory variance, inside and outside the EU, meant "one target model, several localizations" was the only affordable path. Per-market redesigns would not have finished on time.

What I took from it

Running both the design and the delivery of a transformation this broad is a rare experience. Business, technical, organizational, financial, all at once. A lot of how I think about this kind of work now comes from having been in the middle of it.

Three things stuck.

One: designing and executing in the same head is where the learning is. Reality corrects your design in time to fix it, and you learn things about the model no hand-off could teach you. I don't take strategy seriously when it's separated from delivery anymore.

Two: a transformation at this shape is as much a communications job as a design job. Every legal entity, every level of line management, the upstream programme execution, and my own team each needed a different kind of conversation. The TOM moves only as fast as the slowest conversation, and the conversations are not interchangeable. I budget communication effort on that basis now.

Three: the four dimensions only work if they reconcile, and the reconciliation is the hard part. Technical decisions break the financial model. Financial decisions break the organization design. Legal forces every technical and organizational decision to become specific. Holding all four at the same time, not sequentially, was what required the hardest thinking, the most flexibility, and the fastest switching between contexts.

And the residue. I came out of it sharper at the practical habits a programme of that density demands. Focusing hard while changing context constantly. Telling important from less important. Deciding fast. Carrying the stress without it leaking into the work.

Sources (public record on the transaction and the resulting infrastructure group):