GLF Pulse 2026: AI Gets Real, Partnerships Rule, and Growth Hinges on Execution

GLF Pulse Survey 2026: Execution Beats Vision for Telecom Product Teams

The latest GLF Pulse Survey 2026 shows a clear shift: generative AI has moved from pilot to production, partnerships are standard practice, and CAPEX is leaning into growth. Yet price erosion, margin squeeze, heavy regulation, and slow speed-to-market keep dragging performance down. Future winners won't be the most ambitious. They'll be the ones who can turn strategy into scalable, repeatable delivery.

Founded in 2016, the GLF brings together 170+ leaders across global connectivity and digital infrastructure. This year's report reflects input from 50+ organisations across Africa, APAC, Europe, LATAM, the Middle East, and the US-giving product leaders a grounded view of what's working and what's not.

What product leaders should take from the survey

• GenAI is in active use, especially in AI for Operations, product development, and go-to-market.
• Partnerships are now the default path for capability building across AI, services, and connectivity.
• CAPEX is shifting to new business lines and product innovation, not just upgrades.
• Enterprise ICT services and high-performance data connectivity lead value creation.
• Execution gaps persist: 84% lack dedicated innovation teams and launches still take 6+ months.

GenAI: From pilots to production-focus on integration, not hype

More than half of surveyed organisations are piloting or deploying GenAI. The biggest gains are in AI-driven optimisation, predictive maintenance, automated assurance, faster product cycles, and smarter sales motions. That's tangible efficiency and time-to-value.

Partnerships remain the fastest route to capability, but 2026 shows a shift back to building in-house skill. The risk: a growing talent gap. The question is no longer "adopt or not," but how to wire AI into products, processes, and economics without breaking them.

• Adopt a clear buy/build/partner strategy per use case. Don't mix by default-decide by ROI and dependency risk.
• Stand up an AI product council for guardrails on data, model choice, and responsible use.
• Put MLOps and promptOps next to DevOps. Treat models and prompts like versioned product assets.
• Instrument every AI feature with cost, latency, and quality metrics from day one.

Growth optimism with pressure on margins

Optimism is softening, especially among established internationals. Price declines and margin pressure remain the main headwind, now joined by tougher regulation across borders. Tuning legacy services won't cover the gap.

Your roadmap needs new engines of value. That means productising network capabilities, packaging enterprise services, and using automation to remove internal drag instead of adding headcount.

CAPEX is tilting to new services-make it count

Roughly 70% plan CAPEX for new business lines, and 80% will increase investment in product innovation this year. New service development tops the 3-year list, with international network expansion close behind. Analytics spend is easing slightly as attention shifts to automation and orchestration.

• Prioritise inter-carrier automation (trading, deal management, billing) to cut cycle times and leakage.
• Adopt industry APIs to reduce integration friction and speed partner onboarding TM Forum Open APIs.
• Modernise internal systems only where they unblock product delivery or reduce unit costs-avoid vanity upgrades.

Partnerships that ship: Make collaboration a capability

Partnerships now outrank cost-cutting, M&A, or reorgs as the top strategic lever. Nearly half prefer partners to build GenAI capability; over a third rely on them for product innovation. But many still struggle with governance, incentives, and integration.

• Set shared outcomes (ARR, attach rate, churn reduction), not just activity metrics.
• Standardise interfaces and schemas; align on SLAs, credits, and test plans before signing.
• Agree on data ownership, model usage rights, and audit paths up front.
• Create a partner "playbook" with pricing rules, legal templates, and integration kits to scale repeatably.

Where value is forming now

High-performance data connectivity remains the core engine. Content delivery and enterprise ICT are close behind, with enterprise services rated the single largest growth opportunity. Messaging and CPaaS still matter, but focus is shifting to fewer, higher-value plays.

Practical takeaway: package connectivity with enterprise services (security, observability, SASE, managed cloud) to improve margin and stickiness.

Product innovation is priority #1-delivery is the gap

Most organisations build new products internally, yet 84% lack dedicated innovation teams. Launch cycles of six months or more remain common. Budgets, conservative culture, legacy systems, and thin partner execution slow everything down.

• Stand up a small "Product Acceleration" pod: product, Design, engineering, and finance. Give it clear guardrails and decision rights.
• Move to a rolling quarterly roadmap with monthly capacity planning. Fewer bets, higher conviction.
• Use discovery sprints with kill criteria. If a bet can't prove demand or unit economics in 30-45 days, stop.
• Pre-build go-to-market assets (pricing packs, demos, ROI calculators) while engineering finishes MVP.

Metrics that matter

• Time-to-learn: days from idea to decision (ship/iterate/stop).
• Time-to-live: lead time from commit to production for one change.
• Attach rate: % of connectivity deals with at least one enterprise add-on.
• Gross margin by SKU and segment, tracked weekly.
• Automation coverage: % of quotes, orders, and settlements processed straight-through.
• Partner contribution: revenue, margin, NPS, and support tickets per partner.

90-day action plan for telco product teams

• Week 1-2: Pick three growth bets (e.g., enterprise ICT bundle, AI-assisted NOC feature, inter-carrier automation). Define success metrics and kill criteria.
• Week 3-6: Run discovery sprints. Validate demand, integration complexity, and unit economics. Secure two design partners per bet.
• Week 7-10: Build MVPs with shared components. Set up continuous delivery and automated tests. Pre-wire pricing, packaging, and pilot contracts.
• Week 11-13: Launch controlled pilots. Track adoption and margin in real time. Kill or scale based on evidence, not opinions.

Risk is rising-plan for impact, not perfection

Geopolitical instability leads 2026 risks, followed by hyperscaler moves. Regulatory shifts and intense competition stack on top. Build resilience into product and partner choices now.

• Run scenario tests on cross-border services and supply chains.
• Bake compliance into design. For AI features, align with practical frameworks like the NIST AI RMF.
• Keep a fast exit path from critical vendors and models to limit lock-in.

Bottom line

The industry agrees on the right themes: invest in innovation, use AI where it moves the needle, and build with partners. The gap is repeatable execution. Teams that shorten decision loops, productise partnerships, and ship faster-without burning margin-will take the next wave.

If you're closing the AI skills gap on your team, these curated pathways can help Complete AI Training: Courses by Job.