Thirty Per Cent Out, Forty Per Cent Up — Operational Transformation Across EMEA’s Telecoms Operators
Incentive design is not a negotiating tactic; it is an operating-model decision.
Executive Summary
Between 2008 and 2011, a transformation practice delivered against the same contractual commitment — thirty per cent OPEX reduction, forty per cent improvement in service-delivery cycle times — across six telecoms operators in five EMEA countries. The practice was built on three interlocking disciplines: data-driven process baselining that replaced assumption with measurement, systematic channel migration from assisted to self-service, and industrialised delivery that treated operational processes as manufacturing lines rather than craft activities. What distinguished this programme from the dozens of cost-reduction exercises running concurrently across the industry was not the method itself but the commercial model behind it: the transformation partner held full P&L accountability for the delivered outcome, sharing both the upside and the downside with the operator. That single structural decision changed every behaviour that followed. This case examines the mechanism in detail, follows its replication across Southern, Central, and Eastern Mediterranean operators, and is candid about the engagement where the model did not land and the structural reasons it could not.
The Baseline Nobody Wanted to See
The first engagement began in the second quarter of 2008, with a Southern European incumbent whose fixed-line revenues had been declining for three consecutive years. Mobile was growing, but not fast enough to compensate for the structural erosion in voice and broadband margins. The operator’s board had approved a cost-transformation programme eighteen months earlier; what it had produced so far was a consultant’s report recommending headcount reduction and a procurement renegotiation that had saved four per cent on network equipment contracts. Neither had touched the operational cost base in any meaningful way.
The transformation partner’s opening move was not a strategy document. It was a baselining exercise that took eleven weeks and mapped forty-seven end-to-end processes across the operator’s service-delivery chain — from fault report to resolution, from order capture to activation, from provisioning request to live service. The team that ran it was drawn from the partner’s own practice, not the operator’s staff, and they measured three things for every process: cycle time, cost per transaction, and first-time-right rate.
The resulting data set was not flattering. Average fault-resolution cycle time was four times the operator’s published service-level agreement. Provisioning for a standard broadband installation involved seventeen handoffs between five departments, with a first-time-right rate of sixty-one per cent — meaning that nearly four in ten orders required rework, re-dispatch, or customer callback. The cost per transaction on the assisted service desk was fourteen euros; the equivalent self-service transaction, where it existed, cost under two euros, but self-service penetration sat at eleven per cent of the base.
What made the baseline politically explosive was not any single number but the pattern across all forty-seven processes. The operator’s leadership team had believed — because their dashboards told them so — that service delivery was essentially sound, with problems concentrated in a few legacy systems. The baseline revealed something different: the inefficiency was systemic, embedded in handoff structures and approval chains that had accumulated over a decade of incremental organisational change. No single system was the bottleneck. The architecture of the work itself was the bottleneck.
The Commercial Structure That Changed Everything
The conventional advisory engagement bills for time and expertise. The consultant recommends; the client implements; the risk sits with the client. This programme was structured differently. The transformation partner contracted against a defined outcome — a percentage reduction in operational expenditure, measured against the validated baseline, delivered within a fixed timeframe — and was compensated through a share of the realised savings. If the savings did not materialise, the partner absorbed the delivery cost. If they exceeded the target, the upside was shared.
This was not an outsourcing contract, though it borrowed elements from managed-services arrangements then common in the industry. The operator retained its staff, its systems, and its customer relationships. What changed was that the partner now had a direct financial interest in the speed and depth of operational improvement — not in the volume of advice delivered, but in the measurable result.
The behavioural consequences were immediate and specific. When the baselining exercise produced its forty-seven-process map, the partner’s team did not present it as a set of recommendations and step back. They challenged every process owner’s assumptions about what “good” looked like, because every inflated baseline would become a cost the partner would later have to absorb. When the operator’s middle management resisted changes to established workflows — and they did, in every engagement — the partner’s programme director had a P&L reason to escalate, not merely a project-management reason. The commercial model converted what would normally be advisory detachment into operational urgency.
A specific example illustrates the difference. In the Southern European engagement, the redesign of the fault-management process required the operator’s network-operations centre to change its shift-handover protocol — a relatively minor procedural change that would eliminate a four-hour information gap between shifts and reduce the average fault-resolution time by roughly twenty per cent. Under a conventional advisory contract, this recommendation would have appeared in a report, been discussed in a steering committee, and most likely been deferred because the head of network operations had other priorities. Under the P&L model, the transformation partner’s programme director escalated the refusal within forty-eight hours, because every week of delay cost the partner money. The protocol changed within a fortnight.
The Three Levers
The transformation method rested on three disciplines, applied in sequence across every engagement.
The first was process baselining and redesign. Every process that touched the customer — ordering, provisioning, fault management, billing inquiry, service modification — was measured, mapped, and redesigned against a target cost-per-transaction and cycle time. The redesign was not blue-sky; it worked within the operator’s existing OSS/BSS architecture, which in most cases meant legacy systems with limited integration. The gains came not from system replacement but from eliminating handoffs, reducing rework loops, and automating decision points that had previously required manual intervention. In the Southern European engagement, the provisioning process went from seventeen handoffs to seven, and the first-time-right rate moved from sixty-one per cent to eighty-four per cent within nine months.
The second was channel migration. Every operator in the programme had invested in online self-service portals, but adoption was low — typically between eight and fifteen per cent of the customer base for service transactions. The barriers were not technological but operational: the self-service channel could not handle the same transaction set as the assisted channel, error rates were high enough that customers learned not to trust it, and there was no systematic programme to migrate customers from phone to web. The transformation practice treated channel migration as an industrial programme, not a marketing campaign. It involved three sequential steps: aligning the self-service transaction set to cover eighty per cent of assisted-channel volume, fixing the failure modes that drove customers back to the phone, and then actively managing migration through a combination of IVR deflection, first-call-resolution scripts that walked customers through the self-service process, and — in two engagements — deliberate extension of assisted-channel wait times for transactions that were fully available online.
The sequencing mattered. In the Central European engagement, the partner initially pushed migration before the self-service failure modes had been resolved; the result was a spike in repeat contacts as frustrated customers abandoned the web portal mid-transaction and called the service desk. The lesson — fix the destination before you redirect the traffic — became a standard discipline across subsequent engagements. Self-service penetration across the programme moved from an average of twelve per cent to between thirty-five and forty-eight per cent within eighteen months.
The third was delivery industrialisation. Field operations — installations, fault repairs, equipment swaps — were restructured along manufacturing principles: standardised job packs, optimised routing, pre-staged equipment, and completion-rate targets that replaced the traditional metric of dispatches-per-day. The shift in metric was significant. Dispatches-per-day incentivised volume; completion-rate incentivised getting it right first time, which reduced costly repeat visits. In the Central European engagement, repeat-visit rates fell from twenty-two per cent to nine per cent, and the average field workforce handled fourteen per cent more completions per day without additional headcount.
What Travelled and What Did Not
By the end of 2010, the practice had run six engagements across five countries. The method — baselining, channel migration, delivery industrialisation — transferred with remarkably little adaptation. The forty-seven-process map from the first engagement became a template that was refined with each deployment but never fundamentally redesigned. The channel-migration playbook was reusable almost in its entirety. The field-operations model required local adjustment for regulatory differences in workforce management — different countries had different rules about technician working hours, subcontracting, and union consultation — but the core metrics and methods were portable.
What did not travel was the politics.
Every operator had its own internal power structure, and every transformation programme threatened someone’s authority. In the Southern European incumbent, the head of network operations had built a department of eleven hundred people over fifteen years; the baselining exercise implicitly valued that department’s output at less than the executive had claimed, and the resistance was personal, sustained, and conducted through the operator’s board rather than through the programme governance. It took the chief executive’s direct intervention — and the commercial model’s escalation mechanism, which gave the transformation partner standing to raise operational blockers at board level — to break the impasse.
In the Central European operator, the politics were different but equally obstructive. The operator had recently been through a merger, and the transformation programme became entangled with the integration programme in ways that neither side had anticipated. Process owners from the legacy organisations defended their workflows not because the workflows were efficient but because they represented territorial claims in the merged entity. The partner’s programme director spent, by her own estimate, forty per cent of her time on stakeholder management — not because the method required it, but because the organisational context demanded it.
In the Eastern Mediterranean engagement, the political challenge was regulatory. The operator was subject to workforce-reduction restrictions that limited the pace at which headcount savings from process automation could be realised. The commercial model had assumed a faster savings trajectory; the regulatory constraint meant that the thirty-per-cent OPEX target was reached, but six months later than contracted, and the partner absorbed the cost of the extended delivery period.
The pattern was consistent: the mechanism travelled; the politics never did. Each engagement required a bespoke political strategy that could not be templated, could not be predicted from the outside, and consumed a disproportionate share of senior leadership attention. The transformation partner learned, by the third engagement, to budget twice as much senior stakeholder time as the method alone would suggest — a lesson that no process map or playbook could encode.
The Engagement That Did Not Land
The fifth engagement, with a Northern European operator, failed. Not catastrophically — the programme was delivered, the baseline was completed, and some process improvements were implemented — but the contractual targets were not met, the commercial model produced a loss for the transformation partner, and the engagement was not renewed.
The reasons were structural, not methodological. The Northern European operator had a fundamentally different starting position from the other five. Its existing operational efficiency was already in the upper quartile for the European industry; its self-service penetration was already above thirty per cent; its field-operations processes had been through a lean-improvement programme two years earlier. The thirty-per-cent OPEX reduction target, which had been achievable against the inefficient baselines of the Southern and Central European operators, was arithmetically unrealistic against an already-optimised cost base. The transformation partner had underwritten a commitment that the baseline could not support.
The lesson was precise: the commercial model worked when the gap between current performance and achievable performance was wide enough to fund both the transformation cost and the shared savings. When the gap was narrow — when the operator was already reasonably well run — the model’s economics broke down. The partner was, in effect, contracting to deliver improvement that had already been captured.
This was not, in retrospect, a failure of execution. It was a failure of qualification. The practice had developed a strong method for delivering against inefficiency but had not developed an equally rigorous method for assessing, before contract signature, whether the inefficiency was large enough to make the commercial model viable. By the sixth engagement, a pre-contract diagnostic had been added to the sales process: a compressed, two-week assessment covering the same forty-seven processes, measured at sampling depth rather than census depth, designed to produce a rough efficiency profile that could predict — within useful bounds — the size of the improvement opportunity. The diagnostic would not have prevented the Northern European loss; but it would have made the risk visible before the contract was signed, and the practice would have priced or structured the engagement differently.
The Numbers, Taken Together
Across the five successful engagements, the results were consistent enough to constitute a pattern rather than a set of anecdotes:
| Metric | Before (average) | After (average) | Change | ||
|---|---|---|---|---|---|
| OPEX (indexed) | 100 | 69 | -31% | ||
| Service-delivery cycle time | 100 | 58 | -42% | ||
| Self-service penetration | 12% | 41% | +29 pp | ||
| First-time-right (provisioning) | 63% | 86% | +23 pp | ||
| Field repeat-visit rate | 21% | 10% | -11 pp |
The numbers were real, measured against validated baselines, and contractually significant — the partner’s compensation depended on them. They were not, however, uniformly distributed. The Southern European incumbent, which had the lowest starting efficiency, produced the largest gains. The Eastern Mediterranean operator, constrained by regulatory limitations on workforce restructuring, delivered the slowest trajectory. The variation was a function of starting position and local constraint, not of method.
What This Practice Actually Demonstrated
The conventional wisdom in the telecoms industry in 2008 was that operational transformation required either a full outsourcing arrangement — handing the operation to a managed-services provider — or a system replacement programme that would take three to five years and cost hundreds of millions of euros. This practice demonstrated a third path: a focused, method-driven, commercially aligned intervention that delivered measurable results within eighteen months without replacing the operator’s systems, outsourcing its workforce, or requiring capital expenditure beyond the transformation partner’s own delivery cost.
The mechanism was not complex. Data-driven baselining replaced assumption with measurement. Channel migration was treated as an operational programme, not a technology project. Field operations were industrialised using manufacturing principles that had been understood for decades. None of these ideas was new. What made them work at scale, across borders, was the commercial model that aligned the transformation partner’s incentive with the operator’s outcome.
The commercial model was not a contractual innovation for its own sake. It was the structural decision that made every other discipline land — because it converted advisory detachment into operational accountability, and gave the transformation partner standing to fight the political battles that the method, on its own, could not win.
But the model had a boundary condition, and the Northern European failure exposed it precisely. Where the gap between current and achievable performance was wide, the shared-savings model funded both the transformation and the reward. Where the gap was narrow, the model starved. The practice’s most important learning was not a process improvement or a delivery technique — it was a qualification discipline: the ability to measure, before commitment, whether the opportunity was large enough to sustain the commercial structure that made everything else possible.
Three years and six operators later, the transferable lesson is not about telecoms, or about OPEX, or about channel migration. It is about the relationship between commercial structure and operational behaviour. Incentive design is not a negotiating tactic; it is an operating-model decision. Get it right, and the method follows. Get it wrong, and no amount of process excellence will compensate.