Tuesday, November 10, 2026

Enterprise Network Modernization, Part 2: Where to Find Evidence That Isn't Just Slides

In Part 1 I argued for drawing your own map before buying anything. In short: the traffic lives in four domains (data center, campus and branch, WAN and secure access, multi-cloud and hybrid). Three planes govern it (operations and intelligence, security and trust, consumption and sourcing). A physical foundation sits under both. That makes 24 topics, each with exactly one home, decided by who owns the buying decision.

This part is about the question that bothers me even more: once the map is on the wall, where do you get the evidence to decide anything?

The conversation usually jumps straight to "so what do the analysts say?", and that quickly shrinks to looking at the top right corner of a chart. I understand the instinct. It is also how a lot of expensive decisions get made.

Not all evidence is equal

Roughly, from weakest to strongest in terms of relevance to your environment:

  1. Vendor material.
  2. Analyst reports.
  3. Peer conversations and references.
  4. Independent or lab testing.
  5. Your own data and your own proof of concept.

Breadth runs the opposite way. Analysts see the whole market and your data sees only you. So you need both, and you need to know which question each one can answer. Analysts can tell you where the market is heading and who the players are. They cannot tell you whether it fits your network.

Reading analyst output without being fooled

A few habits that have saved me time.

Read the text, not just the picture. The quadrant or the wave is a summary of a long document, and the interesting parts are usually the strengths, cautions and inclusion criteria. Check the date and the scope. A vendor can be a "Leader" in a market defined in a way that has nothing to do with what you want to buy. Remember what the chart does not measure: your fit. And remember who often pays for the copy you are reading. A lot of full reports are paywalled, and what is public is frequently a vendor's reprint or press release. I treat a press release as what it is.

Different firms also do different jobs. Gartner, IDC and Forrester are strongest at defining categories and comparing vendors across them, and they give you breadth. Gigamon, KuppingerCole, Frost and Sullivan and similar firms often go narrower and deeper into features and approach. I find it useful to ask how deep any piece of analysis actually goes. Roughly: does it stop at logos, at checkboxes, at features, at approach, at architecture, or does it get to what the architecture means for you? Most public material stops somewhere around the third step. The value is in the last two, and that is where your own architects and your field experience have to fill in.

And label forecasts as forecasts. Take a line like Gartner's earlier prediction, as reported in the trade press, that by 2026 sixty percent of new SD-WAN purchases would be part of a single-vendor SASE offering. That is a forecast, made at a point in time, by someone with a view. Fine to read, risky to quote as adoption. Always ask who counted, and when.

What the coverage looks like on my own map

I tagged each of the 24 topics with whatever named public analyst evaluation I could find. The pattern was clear enough to be useful. Security and operations are well covered. There is a Gartner quadrant for hybrid mesh firewall (September 2026), Forrester Waves covering SASE and zero trust platforms (both Q3 2025) and network analysis and visibility (Q4 2025), and a Gigamon Radar for network observability (version 6, April 2026). Wireless LAN has both a Gartner wired and wireless LAN quadrant (May 2026) and an IDC MarketScape (2025). Carrier WAN services and managed services have Gartner quadrants too, for global WAN services (March 2026) and managed network services (April 2026).

At the edges it thins out quickly. For edge or micro-DC, IoT and OT networking, optical transport, cloud-native networking, cloud interconnect, service mesh overlays and passive infrastructure, I did not find a named public evaluation I could point to. To be fair, I may have missed some, and paywalled research may cover more than I can see.

That is not a verdict on those topics. Often it is the opposite, because these tend to be the early ones. It just means you should lean harder on field evidence and your own testing, and be more careful about confidence. A gap in analyst coverage is a flag, not a conclusion.

The routine I follow: gather, churn, synthesize

It is boring, and it works.

Gather from three lanes. Market definers for the big picture. Specialists for depth. Then the field: lessons learned, post-mortems, honest conversations with people who have done it, plus your own research. Do not skip the third lane, because it is the only one that is about reality instead of positioning.

Churn is the unglamorous middle. Put everything on the map, topic by topic. Date-stamp each finding. Mark each number as forecast or observed. Cross-check anything important against a second independent source before believing it. Write down where sources disagree and where there is nothing at all. This is where most people stop caring, and it is where most of the value is.

Synthesize into one page per topic. Market context. Adoption, dated and labelled. Considerations. Challenges. The vendor landscape, if you want to show one, on axes you define yourself and disclose, rather than something that looks like somebody else's chart. And a recommendation. Then add one more line: where does practice differ from what the analysts say? That is the field check, and it is often the most interesting sentence on the page.

Your own data is the best evidence you have

The good news is that you probably already own more of it than you think.


None of that comes from a slide. All of it is more relevant to your decision than any quadrant.

The field matters just as much, so write down your own lessons learned while they are fresh, and ask peers what went wrong, not only what went well. When you talk to vendors, ask for the reference that did not go well. I wrote some questions along those lines years ago for self-driving networks (Few Questions You Should Ask Your fav Self-Driving Networks Vendor). The technology names have changed, the questions mostly have not.

Rate the evidence, not just the technology

This is a small habit with a big effect. For every topic, rate two things separately. How attractive is the technology for us? And how confident are we in the evidence behind that? A topic can be attractive and poorly evidenced, which usually calls for a small, reversible experiment and not a big commitment. A topic can be well evidenced and unattractive, which usually means leave it alone. Keeping the two scores apart stops a persuasive demo from doing both jobs at once.

A quick worked example: quantum-safe

Take one topic and run it through the routine, quickly.

Market context: Gartner's 2026 trends for enterprise networking name quantum-safe networks, agentic NetOps and network sovereignty among the themes to watch, and its 2026 Hype Cycle points at regulatory, sovereignty and quantum pressures on network security. So the direction is not in doubt.

Adoption: here I would be careful. What I can find publicly is mostly forecasts and vendor announcements, not measured adoption. I would mark it forecast and leave it there.

Considerations: some data has a long shelf life, and some of your equipment has a longer life than your planning cycle. Whether your vendors support what comes next in firmware matters more than what they say on a roadmap slide.

Challenges: before you can protect anything, you need to know where cryptography is used, and in most organizations nobody has a complete answer.

Own data: the cryptography inventory from the table above. Field: ask peers who have started one how long it took and what surprised them.

Recommendation, and this is a judgement call: start with the inventory, and put crypto agility questions into your next network RFPs. Do not start a rip-and-replace Programme on forecasts. Confidence: reasonable on direction, low on timing.

Notice what happened there. Nobody bought anything, and we now know what to do next.

Where I would start on Monday

Draw your own version of the map, and argue about it until the boxes stop moving. Pick four to six topics, not twenty-four. For each one, find out what the market says, find out what your own data says, and write the two side by side. Mark everything as forecast or observed, with a date. Rate the evidence separately from the technology. And before any budget gets approved, ask the person presenting to show you where it failed for someone else.

None of this is clever. That is rather the point. Stay grounded and keep it real.

HTH...

A Tech Artist 🎨

Sources and further reading

Full reports are largely paywalled. What follows is what I could verify publicly, by title and date, including vendor reprints and press coverage where that was all that was visible.

Posted by at No comments:
Labels: , , , ,

The Agreement Trap: Fixing Group Paralysis in the AI-Era Operating Model




Most enterprises do not have an AI technology problem. They have a collective-action problem, and AI makes it more expensive.

Only 5% of companies capture AI value at scale, while 60% report minimal gains despite substantial spend (BCG, 2025). Just 39% see any enterprise-level EBIT impact from AI (McKinsey, 2025). The models are not the bottleneck. The way groups of people decide is.

Three classic pieces of organizational research explain this better than any technology roadmap I have seen. In this post I connect them, turn them into a simple framework, and tie it back to the operating model every CXO is now redesigning for AI.

Executive summary

  • Three theories explain the gap. Smith and Berg show why groups freeze. Harvey shows why they agree on what nobody wants. Pfeffer and Sutton show why they know what to do but do not do it.
  • Together they form one failure chain. A group can fail to decide (Stuck), decide wrongly through false consensus (Silent), or decide rightly and not act (Stalled). There is no fourth way to miss the right action.
  • AI raises the cost of each failure. Pilots approved by polite consensus, agent programmes nobody owns, and value nobody measures are the same three failures at machine speed.
  • The fix sits in the operating model. Five levers matter: decision rights, team design, workflow, metrics and incentives, and leadership behavior.
  • CXOs can start in 90 days. Diagnose the top 10 AI decisions, install dissent and decision-latency mechanisms, and move execution authority to outcome-owning human-agent teams.

Three diagnoses, one failure chain

1. Stuck: why groups freeze (Smith and Berg)

Kenwyn Smith and David Berg's Paradoxes of Group Life (1987) argues that group life is made of opposing pulls that cannot be resolved. They group 12 paradoxes into three clusters:

  • Belonging: identity, involvement, individuality and boundaries.
  • Engaging: disclosure, trust, intimacy and regression.
  • Speaking: authority, dependency, creativity and courage.

Their companion paper in Human Relations makes the key point: these conflicts are "released" rather than resolved once a group sees them as essential to its life. A group that treats a paradox as a problem to settle, picking one side to win, oscillates or freezes.

Where it fails: before the decision. No decision is made.

2. Silent: why groups agree on what nobody wants (Harvey)

Jerry Harvey's The Abilene Paradox tells the story of a family that makes a hot, 53-mile drive to Abilene that nobody wanted. Each member assumed the others wanted to go. Harvey's insight is that the hard organizational problem is managing agreement, not conflict.

He describes a recognizable pattern:

  • Members privately agree on the problem and on the fix.
  • They fail to say so, and send misleading signals instead.
  • The group acts against what its members actually want.
  • Frustration, blame and subgroups follow, and the cycle repeats.

The deepest driver is fear of separation from the group. Mark Murphy in Forbes recently framed the same issue as a team-design problem: silence gets read as agreement.

Where it fails: at the decision. The wrong decision is made.

3. Stalled: why knowing is not doing (Pfeffer and Sutton)

Jeffrey Pfeffer and Robert Sutton's The Knowing-Doing Gap (2000) shows that firms usually know what to do. Five things stop them doing it (chapter summary):

  • Talk substitutes for action.
  • Memory substitutes for thinking.
  • Fear prevents acting on knowledge.
  • Measurement obstructs good judgment.
  • Internal competition turns colleagues into rivals.

Stanford GSB's excerpt stresses the last one: zero-sum internal contests destroy the cooperation that interdependent work needs.

Where it fails: after the decision. The right decision is not executed.

Why AI raises the stakes

AI adoption is broad, but value is narrow. Every major 2025 study finds a small leading group pulling away from a large stalled majority.



The three failure modes map directly onto these numbers:

  • Stuck shows up as decision latency. AI shifts authority, identity and dependency all at once. Those are exactly Smith and Berg's paradoxes of speaking and belonging. When leaders are unsure whether a human or an agent owns an outcome, they default to committees. Gartner expects at least 15% of day-to-day work decisions to be made autonomously by 2028 (W.Media). Decision rights written for a human-only firm will not hold.
  • Silent shows up as pilot theatre. Many AI programmes get approved because nobody wants to look like the sceptic in the room. The approvals look like consensus. The cancellations that follow are the drive home from Abilene.
  • Stalled shows up as the value gap. Most firms know AI needs workflow redesign, yet McKinsey's high performers are almost three times more likely to actually do it. MIT NANDA blames a "learning gap" and brittle integration, not model quality. That is the knowing-doing gap in its purest form.

The human signal is moving the wrong way. Manager engagement fell from 30% to 27% in 2024 (Gallup). Positive views of GenAI rise from 15% to 55% where leadership support is strong, yet only about a quarter of frontline staff say they get that support (BCG). Managers are the layer through which all three failures flow.

My framework: the Movement Stack

A group misses the right action in only three ways, and each is fixed through the same five operating-model levers. The failure modes are mutually exclusive by symptom: no decision, a wrong decision, or a right decision left undone. Together they cover the whole path from tension to action.

 


The five levers are the parts of an operating model that shape how groups decide and act:

  1. Decision rights: who decides, and by when.
  2. Team design: how humans and AI agents share the work.
  3. Workflow: how work moves end to end.
  4. Metrics and incentives: what gets counted and paid.
  5. Leadership behavior: what leaders model every day.

Diagnose first, then apply the lever. Adding more dissent, a Silent fix, to a Stuck team only adds paralysis.

If you are Stuck: create decisiveness

  • Decision rights: one named owner per decision, with a decide-by date.
  • Team design: define what humans own above the loop and what agents do inside it.
  • Workflow: sort decisions into reversible and irreversible, and give the reversible ones a fast lane.
  • Metrics and incentives: track decision latency as a board KPI.
  • Leadership behavior: name the paradox out loud and hold both sides instead of forcing a choice.

If you are Silent: create real voice

  • Decision rights: collect independent written positions before any group discussion.
  • Team design: rotate a formal challenger role, and add an AI red-team agent that is told to disagree.
  • Workflow: run a pre-mortem and set kill criteria before any AI investment is approved.
  • Metrics and incentives: measure voice, such as recorded dissent and "my opinion counts" scores.
  • Leadership behavior: the most senior person speaks last and rewards bad news.

If you are Stalled: create motion

  • Decision rights: execution authority sits with the team that owns the outcome.
  • Team design: small, cross-functional human-agent teams replace functional committees.
  • Workflow: redesign end to end; every pilot scales or stops within 90 days.
  • Metrics and incentives: a few P&L-linked value measures, and no rankings that pit teams against each other.
  • Leadership behavior: treat reasonable failure as learning, and build the system rather than making every decision.

This is consistent with where operating models are heading anyway. McKinsey's work on the agentic organization describes small, outcome-focused teams of humans who supervise AI workflows, with governance that is real time rather than periodic. Those teams will only work if the three failure modes are designed out from the start.

What CXOs should do in the next 90 days

Treat collective action as an operating-model workstream with an owner, a baseline and a deadline, not as a culture campaign.

Days 0 to 30: Diagnose

  • Take the 10 largest open AI decisions and investments, and classify each as Stuck, Silent, Stalled or healthy.
  • Baseline three numbers: decision latency, pilot-to-production rate, and share of AI spend with a tracked P&L outcome.
  • Run Harvey's test anonymously with each leadership team: would members privately reverse a decision they publicly backed?

Days 31 to 60: Install the mechanisms

  • Give every AI decision a single owner and a decide-by date.
  • Make pre-mortems and kill criteria a gate for AI investments above an agreed threshold.
  • Collect written positions before every steering committee, and have the most senior person speak last.
  • Redraw decision rights for three end-to-end workflows, stating where humans sit in the loop, above it, or out of it.

Days 61 to 90: Move authority and money

  • Hand two or three workflows to small human-agent teams that own the outcome end to end.
  • Shift part of leadership incentives from functional KPIs to shared value measures.
  • Scale or stop every pilot older than 90 days.
  • Report the three baseline numbers to the board quarterly, next to the financials.

The test of success is simple: fewer open decisions, more pilots scaled or stopped, and dissent recorded before decisions rather than blame after them.

Limits of this argument

I stress-tested this framework the way I would review a colleague's work. A few honest caveats:

  • Scope. The framework does not cover honest error, where genuine debate still picks the wrong option. Nor does it cover drift, where a good decision goes stale because nobody revisits it. A fourth mode, failure to learn after acting, may be worth adding.
  • Evidence quality. The BCG and McKinsey figures come from large but self-reported surveys, and the link between workflow redesign and performance is correlation, not proof of cause. The MIT 95% figure is widely quoted but rests on a small sample, so I treat it as directional. The decision-making and psychological-safety data predate GenAI.
  • Regional fit. All three theories come from US research. In the hierarchical cultures common across APAC, I suspect Silent is the dominant failure mode, and "the leader speaks last" on its own will not unlock voice.
  • AI can make Silent worse. AI assistants tend to agree with whoever is asking. A red-team agent only helps if it is explicitly instructed to disagree and its objections are logged.

Smith and Berg would also warn against using this framework as an either/or tool. The aim is not to maximize dissent or speed. It is to hold speed and voice, autonomy and control, in useful tension.

HTH...

A Tech Artist 🎨

Sources

 

Posted by at No comments:
Labels: , ,

Tuesday, October 13, 2026

The Fallacy of Autonomous AI: Why Enterprise Value Lives in the Middle


In executive forums and vendor keynotes, the dominant narrative remains the same: fully autonomous, self-governing AI agents that will run end-to-end business functions with minimal human involvement. The vision is compelling. The economics and operating realities of most enterprises, however, make the near-term pursuit of that vision one of the higher-risk capital allocation choices currently available.

Early production evidence shows that organizations fail with agentic AI less because models lack raw capability and more because their process foundations, integration surfaces, data quality, identity and authorization models, and control systems cannot safely support the autonomy they attempt to deploy. Value is not binary between a conversational interface and a fully autonomous workforce. It sits in the middle of the spectrum—in bounded, goal-oriented agents applied to well-defined, high-volume Standard Operating Procedures (SOPs).


Where Delivery Organizations Misalign

Three persistent misalignments currently dilute returns:

Business leadership is drawn to the right-hand side of the spectrum - multi-agent coordination, autonomous objective setting, and dynamic self-governance. These capabilities sit beyond current enterprise readiness for the majority of processes. Without deterministic foundations, such initiatives routinely remain high-cost proofs of concept that cannot be scaled under acceptable risk and audit conditions.

IT and technology leadership frequently anchors on the left - broad deployment of conversational assistants and copilots. These improve individual productivity but leave transactional systems, exception handling, and cross-system orchestration largely untouched. The result is diminishing returns once basic knowledge-work assistance is in place.

Engineering and architecture teams often design complex multi-agent frameworks, long-term memory structures, and non-deterministic loops before the underlying business processes have been standardized and instrumented. The outcome is unpredictable tool-call behavior, elevated latency, rapid token cost growth, and limited operational yield.


A Structured Maturity Progression

Organizations extract the highest near-term economic yield by progressing deliberately across a maturity spectrum rather than attempting large jumps. The progression is not linear, and readiness must be demonstrated at each stage through explicit metric thresholds rather than qualitative judgment alone.

Stage 1 – Minimal Reactive Search (Guided). Rule-based logic, deterministic decision trees, and structured knowledge retrieval. Prerequisite: clean operational taxonomies, API endpoint governance, and data hygiene.

Stage 2 – Emerging Conversational Assistant (Guided). Context-aware language models providing synthesis, multi-turn assistance, and draft outputs. These should be treated as advisory. Outputs require human review before any system of record is updated.

Stage 3 – Basic Goal-Oriented Agent (Enterprise Value Zone). Agents that perform task decomposition, structured tool use (APIs, queries, enterprise search), and multi-step execution against bounded SOPs. This is where the majority of near-term enterprise investment should concentrate. Focus areas are high-volume, relatively standardized processes with clear success criteria and limited exception paths.


Readiness criteria for Stage 3 (tied to metrics):

  1. Task Completion Rate (TCR) sustained above the human baseline for the target SOP.
  2. Human Intervention Index (HII) below a defined threshold that still delivers net capacity gain.
  3. Tool-Call Precision above 90 % with drift held inside acceptable bounds.
  4. Cost-Per-Resolved-Workflow (CPRW) demonstrating clear advantage versus manual or traditional automation.
  5. Hallucination & Policy Breach Rate at or below the organization’s residual risk tolerance, with automatic escalation active.
Stage 4 – Intermediate Continuous Improver. Introduction of closed-loop feedback, limited self-correction within pre-approved boundaries, and longer-term contextual memory. Human-in-the-loop evaluation remains mandatory for edge cases and policy boundaries. Progression requires the Stage 3 metrics to hold under increasing volume and complexity.

Stage 5 – Advanced Strategic Automatic Agent (Cautionary Horizon). Multi-agent coordination, autonomous objective setting, and dynamic self-governance. This remains an R&D and controlled-pilot domain. Entry requires the full Stage 3–4 metric set to demonstrate sustained performance inside human operational error rates, plus robust real-time circuit breakers, full decision provenance, and continuous policy auditability.

Aligning the Maturity Spectrum with a Multi-Year Delivery Path

The agentic spectrum does not exist in isolation. It maps directly onto the critical path most organizations must follow to convert generative and agentic AI investment into durable impact. Early production programs that skip foundation work—process standardization, stable integration contracts, identity models for non-human actors, and evaluation harnesses—consistently stall at the pilot stage regardless of model capability.





A practical three-year sequence looks like this:

Year 1 (Foundation). Establish process visibility for priority SOPs, clean integration surfaces, baseline operational metrics (TCR, HII, CPRW), and the control framework. Deploy Stage 1–2 capabilities where they reduce knowledge-work friction. Avoid large multi-agent builds.


Year 2 (Targeted Deployment). Concentrate investment in Stage 3 Basic Goal-Oriented Agents on a small number of high-volume, well-bounded workflows. Gate expansion on sustained metric performance (TCR above human baseline, HII declining, CPRW advantage, breach rate inside tolerance). This is the primary value-creation window for most enterprises.


Year 3 (Selective Scale and Controlled Horizon Work). Expand proven Stage 3 agents across additional process families. Begin limited Stage 4 continuous-improvement loops only where the Stage 3 metrics remain stable under volume. Keep Stage 5 activity inside explicit R&D or tightly governed pilots with full circuit breakers and provenance



Enterprise Prerequisites Before Agent Deployment

Process slicing alone is insufficient in most large organizations. Tribal knowledge, inconsistent execution, legacy system interfaces, and data quality issues routinely constrain agent performance more than model capability. Before significant agent investment, organizations should establish process visibility, stable schema-constrained integration contracts, identity and least-privilege models that extend to non-human actors, and baseline operational metrics against which agent performance can be compared. Agents should augment or orchestrate existing automation and workflow platforms rather than be positioned as a wholesale replacement.

Four-Phase Delivery Lifecycle

A disciplined delivery sequence reduces the risk of unbounded pilots. Each phase includes explicit metric gates before progression.

Phase 1 – Process Slicing and Pattern Identification. Deconstruct target workflows into discrete, repeatable SOPs. Isolate sub-steps that are deterministic or tightly bounded. Establish baseline TCR, HII, and cycle-time measures.

Phase 2 – Bounded Tool Provisioning. Provide agents with structured, schema-validated tool interfaces. Default to read-only access. Restrict write or action permissions to the narrowest scope required. Measure initial Tool-Call Precision and establish drift monitoring.

Phase 3 – Telemetry and Human-in-the-Loop Intercepts. Instrument prompts, reasoning traces, tool calls, latency, and token consumption. Implement automatic escalation when an agent exceeds iteration limits, confidence thresholds, or policy boundaries. Track HII and Hallucination & Policy Breach Rate in real time.

Phase 4 – Progressive Autonomy and Gatekeeping. Expand the autonomous execution window only after TCR, HII, CPRW, and breach-rate metrics demonstrate sustained advantage over the human or traditional automation baseline. Maintain explicit gates and rollback criteria tied to these measures.

Governance, Controls, and Operating Metrics

Vanity engagement metrics provide limited insight. Governance should rest on operational, risk, and economic measures, overlaid with authorization, auditability, and evaluation controls. The five core metrics—Task Completion Rate, Human Intervention Index, Tool-Call Precision & Drift, Cost-Per-Resolved-Workflow, and Hallucination & Policy Breach Rate—serve both as performance indicators and as the quantitative basis for readiness gates between maturity stages. Authorization is enforced at the tool and action level. Full decision provenance and immutable logging support auditability. Evaluation gates covering quality, safety, and compliance precede any expansion of autonomy.

Strategic Path Forward

Enterprise advantage with agentic AI will not come from being first to declare autonomous systems. It will come from disciplined architectural and operating-model execution that concentrates investment where foundations can support reliable performance, measured by explicit thresholds on the operational metrics above.

Organizations should stop treating AI as a binary choice between a conversational interface and an autonomous workforce. The highest near-term return lies in Basic Goal-Oriented Agents applied to repeatable, bounded workflows. This requires clean integration contracts, operational telemetry, progressive autonomy gates tied to TCR, HII, precision, CPRW, and breach-rate performance, and a control framework that treats non-human actors with the same rigor applied to human ones. Autonomy should be earned through demonstrated metric performance rather than assumed through model capability.

The middle of the spectrum is not a compromise. It is the zone in which most enterprises can currently convert AI investment into measurable operational and economic outcomes. The organizations that treat the maturity spectrum and the multi-year critical path as two views of the same operating model—rather than as separate technology and strategy exercises—are the ones currently converting agentic investment into durable results.

HTH...

A Tech Artist 🎨

Posted by at No comments:
Labels: , , , ,

Monday, October 12, 2026

Enterprise Network Modernization, Part 1: Draw the Map Before You Buy the Boxes




Ask five people in the same company what “network modernization” means and you will get five answers, each one shaped by the last vendor presentation that person sat through. One will say SD-WAN. Another will say SASE, or zero trust. The data centre folks will talk about AI fabrics and 800G optics. The cloud team will tell you the network is somebody else's problem now, it lives inside a VPC. And someone from finance will ask why all this costs so much and whether we can just rent it.

None of them is wrong, which is exactly the problem. Everybody is describing one piece and nobody is holding the whole thing. So the roadmap turns into a shopping list, the shopping list turns into a set of projects, and a few years later nobody can quite say who asked for what, why, or whether anything actually got better.

I wrote about the same itch for security a while back (The Enterprise Security Meta Model). Same spaghetti, different kitchen. This post is the networking version, and then a second part on something that bothers me even more: once you have the map, where do you get the evidence to decide anything?

If you are an executive, the map and the evidence section are the parts to read, and you can skim the technical asides. If you run or design networks, the asides are for you, and I would like to hear where you disagree.

Why I start with a map

Without a shared map, three things tend to happen.

First, overlap. Two teams quietly buy the same capability under different names. The firewall team and the WAN team both end up with a cloud security service, each with its own policy engine, each convinced theirs is the strategic one.

Second, gaps. Something important has no owner, and it is usually the thing that sits between two teams. Who owns the path from a branch office to a SaaS application? Ask three people, get three answers, and a fourth who says “it depends”.

Third, sequencing mistakes. Agentic operations get funded before anyone trusts the telemetry. Segmentation projects start before anyone knows what talks to what. You can guess how those go.

There is a second reason, and it is less obvious. Categories keep moving. Gartner published a Magic Quadrant for Single-Vendor SASE in mid-2024. A year later the same space was being evaluated as SASE platforms, and the 2026 edition is dated July. The technology did not change overnight. The label did. If your strategy is organized around somebody else's category names, it ages with their marketing calendar. I would rather own the map and then lay the analyst categories on top of it.

The map


Two questions organize the whole thing. Where does the traffic live? And what governs it? The first gives four domains. The second gives three planes. Under both sits a physical foundation that nobody wants to talk about and everybody depends on. 


That is 24 topics plus the foundation. It is not a magic number, it is just where I ended up after moving things around many times.

The one rule that makes it work: every topic has exactly one home, and the home is decided by who owns the buying decision, not by what the product brochure says. Consultants call this mutually exclusive and collectively exhaustive. I just call it not double-counting.

So SASE platforms sit under WAN, because that is where the traffic path gets bought, even though half of what they do is security. Quantum-safe appears once, under security. Segmentation appears once, under security, even though it gets enforced in campus switches, in DC fabrics and in cloud constructs. The moment you let a topic appear in two boxes, the map turns into a pile of overlapping boxes, and a pile is not a map.

Where the traffic lives

Data center. This is where the money and the heat are. Enterprise DC fabric is the familiar leaf-spine, EVPN-VXLAN world that carries the applications you already run. AI fabric is a different animal: the back-end network that ties GPUs together, with a few enormous, synchronized, loss-sensitive flows instead of lots of small ones, and the Ethernet or InfiniBand debate that comes with it. My view, and it is only a view, is that most enterprises are not building GPU clusters. They are renting capacity or calling APIs, and for them the AI question is less about 800G switches and more about connectivity, data movement and security around that rented capacity. Whether you need the AI fabric topic as a priority depends on which of those two worlds you actually live in. Edge or micro-DC is compute pushed closer to where data gets created, such as stores, plants and branches. The network question there is mostly about how many of them you will end up with and who is going to run them.

Campus and branch. The least glamorous domain and the one most of your users actually touch. Wired and wireless LAN are where Wi-Fi 7 and its predecessors live. Private 5G is for the cases where Wi-Fi is the wrong tool, such as large sites, mobility, or industrial areas. IoT and OT networking covers the devices that cannot run an agent, cannot be patched on your schedule, and were sometimes bought by a department that never told IT.

WAN and secure access. SD-WAN is the overlay that steers traffic across whatever transport you have. SASE platforms bundle that with cloud-delivered security, and the interesting argument inside most organizations is whether the network team or the security team owns the decision. Carrier WAN services are the underlay: the ISPs and telcos that carry it all, and the contracts that make switching hard. Optical or photonic transport sits below that, and it matters when you need a lot of bandwidth or predictable latency between a handful of sites. The conversation moves down the stack, and the people you need in the room change.

Multi-cloud and hybrid. Native cloud networking is every cloud's own constructs (VPCs, VNets, transit gateways), each slightly different. Multi-cloud platforms try to give you one policy and one view across them. Cloud connect, or SDCI, is the software-defined private on-ramp to those clouds. Service mesh and PaaS overlays are application-level networking inside platforms like Kubernetes, and they are usually owned by platform teams, not network teams, which is where a lot of quiet friction comes from.

What governs it

Operations and intelligence. Three items, and the order matters. Observability and digital experience is how you see. Automation and orchestration is how you act, deterministically, from a specification. Agentic NetOps is how you let software decide, and that is a different thing from automation, not a bigger version of it. Automation is governed by the spec: it either does what was written or it does not. A learned system is governed by measuring its behavior over time, with confidence levels and drift. Treat them as the same thing and you will govern one of them badly. You cannot automate what you cannot see, and you should not let an agent act on what you cannot verify.

Security and trust. Zero trust and segmentation is deciding who may talk to whom, and it is hard because it is mostly an organizational problem wearing a technical costume. Hybrid mesh firewall is the idea of consistent firewalling across on-prem, cloud and branch under one management plane. Quantum-safe is about cryptography that survives a future quantum computer, which matters now because some of what you send today will still be sensitive then. Network security basics is the unglamorous baseline controls that still need to be in place before anything fancier is worth discussing.

Consumption and sourcing. NaaS is a commercial model, not a technology. What you are really buying is a contract, so read the exit terms before you read the brochure. Managed and co-managed services answer a different question: who runs it, and who carries the pager.

Foundation. Fibre, cabling and optics. Twenty-year assets, slow to change, and the mistakes here are the most expensive to reverse.

What the map does not cover, and where it is weak

Satellite and LEO connectivity, DNS and IP address management, voice and collaboration, and whatever 6G turns out to be are not on it. Add them when they matter. The structure should make it obvious where they would go.

It is also not perfect. SASE straddles two worlds and any placement is a compromise. Edge or micro-DC could just as easily live under campus and branch. And a map is for deciding what to talk about, not for designing anything. A design still needs requirements, constraints and trade-offs, and no box on a slide replaces that.

What to do with it

Twenty-four topics cannot all be priorities, and if they are, nothing is. Pick a few. For an exercise like this, four to six is plenty.

For each candidate, five questions, in plain language:

  1. What business outcome is pulling on this? If you cannot name one, stop here.
  2. Where are we today, honestly?
  3. What breaks, or gets expensive, if we do nothing until 2030?
  4. What does it depend on? SASE depends on identity being in reasonable shape. Agentic operations depend on observability and automation. Segmentation depends on knowing what talks to what.
  5. What evidence do we have, and how good is it?

The fifth one is where most plans quietly fall apart, so the rest of this post is about that.

What comes next

The map tells you what to talk about. It does not tell you what to believe. In Part 2 I cover where I look for evidence before any of these topics gets a budget, how I read analyst reports without being fooled by the top right corner, and why I rate the evidence separately from the technology.  

HTH...

A Tech Artist 🎨
Posted by at No comments:
Labels: , , , ,

Wednesday, October 7, 2026

From Pass-Through to Proof: Overcoming the Gartner & IDC Blind Spot in Enterprise Advisory

 



What Traditional Analysts Miss

Traditional industry analysts prioritize market presence, revenue trajectory, and execution scale over technical depth. By evaluating products primarily through commercial readiness, logo adoption, and high-level feature checkboxes, legacy reports overlook underlying architectural mechanics. This creates a critical blind spot: market "Leaders" are often recommended based on brand stability rather than integration feasibility, frequently leading enterprises to adopt bloated, overprovisioned platforms that introduce technical debt and policy fragmentation.

What Architectural Analysts Bring

Emerging technical analysts address this deficit by evaluating solutions from the code and primitive level up. Rather than scoring marketing messaging or corporate viability, they test how an architecture actually translates into real-world capabilities—such as securing AI agents or fitting cleanly into multi-cloud environments. These briefings provide advisory teams with an objective, low-risk environment to stress-test technical trade-offs, evaluate niche versus platform strategies, and verify compatibility before making enterprise commitments.

Where Architectural Analysts Fall Short

Conversely, purely architectural evaluations lack the macro-level risk management that enterprise boards require. They do not measure financial solvency, global support infrastructure, long-term vendor viability, or broad peer-deployment history. A solution may possess an elegant architecture, but if the vendor lacks the scale or financial runway to survive an enterprise procurement cycle, recommending it introduces operational and commercial risk.

The Symbiotic Dual-Lens Framework

Neither evaluation model is sufficient on its own; true advisory leadership requires both. Traditional analysts validate market viability and governance compliance (answering “Will this vendor exist in five years?”), while architectural analysts validate technical fit and stack efficiency (answering “Will this actually work in our environment?”). Combining both frameworks bridges the gap between commercial safety and technical reality, enabling technology advisors to deliver strategies that satisfy both risk-conscious procurement teams and speed-driven technical architects.


HTH...

A Tech Artist 🎨

Posted by at No comments:
Labels: ,
Subscribe to: Posts (Atom)