Skip to content
ArchitectureTechnology RadarHorizon ScanningInnovationStrategy

Emerging Technology Scanning

Level:All levels — SA, EA, TS
Duration:Half-day workshop
Deliverable:Technology radar template + horizon scanning process + hype vs signal evaluation criteria

Quick Navigation


Before we start — the one thing to hold onto

Technology moves faster than most organisations can absorb. Not everything new is relevant — the skill is filtering hype from genuine opportunity. I've seen organisations discover technology shifts only when a competitor has already adopted them or when a vendor announces EOL. By then, the response is rushed, expensive, and poorly designed.

Scanning = filtering signal from noise — structured process with sources, cadence, evaluation criteria, and decision outcomes. If scanning doesn't change decisions, stop doing it.

Scanning isn't reading tech blogs. It's a discipline — monitor what's coming, evaluate rigorously, place technologies on a radar, and connect every placement to an investment or governance decision.

Keep it in mind.


1. Why scanning matters — The cost of being surprised by technology shifts

Most organisations discover technology shifts only when a competitor has already adopted them or when a vendor announces EOL. By then, the response is rushed, expensive, and poorly designed. The cost of being surprised is always higher than the cost of scanning.

Without scanning, every technology shift becomes a crisis:

flowchart TD
    SHIFT["⚡ Technology Shift"] --> NO_SCAN["❌ Without Scanning\nDiscover late\nPanic response\nRushed adoption"]
    SHIFT --> WITH_SCAN["✅ With Scanning\nSee it early\nDeliberate response\nPlanned transition"]

    NO_SCAN --> COST_H["High Cost\nPoor design\nCompetitive disadvantage\nForced migration"]
    WITH_SCAN --> COST_L["Low Cost\nWell-designed\nCompetitive advantage\nSmooth transition"]

The cost of surprise:

Scenario Without scanning With scanning
Competitor adopts new technology Panic adoption — rushed, poorly designed Already evaluated — deliberate response
Vendor announces EOL Crisis migration — unplanned cost Already planned — smooth transition
New paradigm emerges (e.g., AI agents) Confusion — "what do we do?" Already trialled — know what works
Regulation changes Compliance gap — reactive Already designed in — proactive

The scanning value chain: Monitor — track sources systematically. Raw signal list. Filter — apply relevance criteria. Shortlist of technologies. Evaluate — assess against strategy and risk. Technology assessments. Decide — adopt, trial, assess, or hold. Radar placements. Act — fund pilots, update standards, plan migrations. Investment decisions.

Lead time economics: Without scanning — technology shift discovered at competitor launch. 0 months lead time → panic adoption → 18 months to catch up. Cost: 3x normal adoption cost (rushed, rework, poor design). With scanning — technology shift discovered 12 months early. 12 months lead time → deliberate evaluation → 6 months to adopt. Cost: 1x normal adoption cost (planned, well-designed).

The ROI of scanning isn't just cost avoidance — it's competitive positioning. The organisation that sees shifts first adapts first.

Scanning creates lead time — don't wait to be surprised.


Try it yourself — The surprise audit

When was the last time your organisation was surprised by a technology shift?

Scenario What happened Cost Could scanning have prevented it?

If you can name two or more, you need a scanning practice.


2. What to scan — Sources, signals, and how to structure the horizon

Without a structure for what to scan, teams either scan randomly (whatever catches their eye) or not at all. Random scanning produces random results — a collection of interesting articles with no decision value.

Structure scanning across three horizons and five source types:

flowchart TD
    SCAN["📡 What to Scan"] --> HORIZON["🗓️ Three Horizons"]
    SCAN --> SOURCES["📚 Five Source Types"]

    HORIZON --> H1["H1: Now\n0-6 months\nReady for adoption"]
    HORIZON --> H2["H2: Next\n6-18 months\nMaturing rapidly"]
    HORIZON --> H3["H3: Later\n18-36 months\nEarly signals"]

    SOURCES --> S1["Industry analysts"]
    SOURCES --> S2["Academic research"]
    SOURCES --> S3["Vendor roadmaps"]
    SOURCES --> S4["Community / OSS"]
    SOURCES --> S5["Internal experiments"]

    H1 --> DECIDE["Decisions\nNot just information"]
    H2 --> DECIDE
    H3 --> DECIDE

Three horizons:

Horizon Timeframe What to look for Action
H1: Now 0-6 months Technologies ready for adoption Evaluate for standards
H2: Next 6-18 months Technologies maturing rapidly Trial and assess
H3: Later 18-36 months Emerging signals, early-stage Monitor and learn

Five source types:

Source type Examples
Industry Gartner, ThoughtWorks Technology Radar, InfoQ
Academic arXiv, conference proceedings, research labs
Vendor Vendor roadmaps, release notes, beta programmes
Community Open source projects, developer forums, meetups
Internal Team experiments, proof of concepts, hackathons

Not all signals are equal. Evaluate signal quality before investing in evaluation: Strong signal — multiple independent sources, clear trajectory, major vendor backing. Evaluate now. Medium signal — some sources, emerging community, unclear trajectory. Monitor closely. Weak signal — single source, early-stage, speculative. Monitor passively. Noise — marketing hype, no substance, no independent validation. Ignore.

Use multiple sources — no single source is reliable alone.

Three horizons (Now, Next, Later) + five source types = structured scanning — not random blog reading.


Try it yourself — The source check

Which of these sources are you currently scanning?

Source type Currently scanning? Gap
Industry analysts
Academic research
Vendor roadmaps
Community / OSS
Internal experiments

If you're only using one or two, your scanning is incomplete.


3. The technology radar — Building, maintaining, and communicating one

Technology assessments that live in documents nobody reads have no impact. Teams need a shared vocabulary and a visual tool to communicate technology position — what to adopt, what to trial, what to watch, and what to avoid.

The technology radar (popularised by ThoughtWorks) is a visual tool for communicating technology position:

flowchart TD
    RADAR["🎯 Technology Radar"] --> RINGS["💍 Four Rings\nAdopt · Trial · Assess · Hold"]
    RADAR --> QUADS["📐 Four Quadrants\nTechniques · Platforms\nTools · Languages"]
    RADAR --> VELOCITY["⚡ Adoption Velocity\nHow fast is it moving\nthrough the rings?"]

    RINGS --> ADOPT["Adopt\nProven, recommended\nUse as standard"]
    RINGS --> TRIAL["Trial\nWorth pursuing\nPilot in 1-2 teams"]
    RINGS --> ASSESS["Assess\nWorth exploring\nResearch and experiment"]
    RINGS --> HOLD["Hold\nProceed with caution\nDo not start new"]

    VELOCITY --> FAST["Fast Velocity\nRapidly moving to Adopt\nHigh strategic urgency"]
    VELOCITY --> SLOW["Slow Velocity\nStuck in Assess\nMay not be worth it"]

Rings (adoption readiness):

Ring Meaning Action
Adopt Proven, recommended for production use Use as standard
Trial Worth pursuing, limited production use Pilot in 1-2 teams
Assess Worth exploring, understand potential Research and experiment
Hold Proceed with caution, do not start new Migrate away if possible

Quadrants (technology categories):

Quadrant What it covers
Techniques Architecture patterns, practices, methodologies
Platforms Infrastructure, cloud, runtime platforms
Tools Development tools, CI/CD, observability
Languages & Frameworks Programming languages, frontend/backend frameworks

Adoption Velocity — measuring momentum: Adoption Velocity measures how quickly technologies are moving from assess to trial to adopt. It's calculated by tracking ring changes over time.

Velocity What it means Implication
Rapid Moved 2+ rings in 2 quarters High urgency — evaluate immediately
Steady Moved 1 ring in 2 quarters Normal cadence — follow the process
Stalled No movement in 2+ quarters Question relevance — why is it stuck?
Declining Moved backwards (e.g., Adopt → Trial) Risk signal — investigate why

Building a radar: Gather — collect technologies from scanning sources. Evaluate — assess each against strategy, risk, and readiness. Place — assign to ring and quadrant. Measure velocity — track ring changes over time. Communicate — share with teams — quarterly update. Act — move technologies between rings as they mature or decline.

The radar = Adopt / Trial / Assess / Hold + Adoption Velocity — a shared vocabulary for technology position and momentum.


Try it yourself — The radar draft

Take three technologies your organisation uses or is evaluating. Where would you place them?

Technology Ring Quadrant Velocity

If you can't place them, you don't have a radar. You have opinions.


4. Separating hype from opportunity — How to evaluate emerging technologies rigorously

Every new technology arrives with a wave of hype — conference talks, vendor marketing, social media excitement. Following hype leads to wasted investment. Ignoring everything leads to missed opportunities. The skill is evaluating rigorously at the right moment.

The Gartner Hype Cycle describes the typical trajectory of emerging technologies:

flowchart LR
    TRIGGER["🚀 Innovation\nTrigger"] --> PEAK["📈 Peak of\nInflated\nExpectations"]
    PEAK --> TROUGH["📉 Trough of\nDisillusionment"]
    TROUGH --> SLOPE["📈 Slope of\nEnlightenment"]
    SLOPE --> PLATEAU["✅ Plateau of\nProductivity"]

    PEAK -.->|"Do NOT adopt\nhere — max hype"| DANGER["⚠️ Danger zone\nUnrealistic promises"]
    TROUGH -.->|"Best time to\nEVALUATE"| OPPORTUNITY["💡 Opportunity\nHype gone, reality clear"]
    PLATEAU -.->|"Safe to\nADOPT"| SAFE["✅ Safe zone\nProven, mainstream"]

Hype cycle stages:

Stage What happens Architect's response
Innovation Trigger New technology appears; excitement spikes Monitor — don't adopt
Peak of Inflated Expectations Maximum hype; unrealistic promises Assess — evaluate critically
Trough of Disillusionment Hype fades; real problems emerge Best time to evaluate — hype is gone, reality is clear
Slope of Enlightenment Real use cases emerge; practical value Trial — pilot with real use cases
Plateau of Productivity Mature, proven, mainstream Adopt — if it fits strategy

The evaluation criteria:

Criterion Question
Strategic fit Does this support our business strategy?
Maturity Is this production-ready or still experimental?
Community Is there an active community and ecosystem?
Talent Can we hire or develop skills for this?
Risk What's the blast radius if it fails?
Alternative Is there a proven alternative that works?

Hype detection checklist — red flags: Vendor-only narrative — "Only our product does this." No failure stories — "Everyone who tries it succeeds." Buzzword density — every sentence has 3 buzzwords. No production references — "We have great demos" (no production deployments). Unrealistic timelines — "Adopt in 2 weeks, transform in 2 months."

Green flags: Multiple independent sources — different people, different contexts, same conclusion. Failure stories available — "Here's what went wrong and how we fixed it." Production references — "Running in production for 12 months at scale." Clear limitations — "Here's what it does NOT do well." Realistic timelines — "Expect 6 months to production maturity."

Evaluate at the Trough of Disillusionment — hype is gone, reality is clear.


Try it yourself — The hype check

Take a technology you're currently evaluating. Run it through the hype check:

Red flag Present? Green flag Present?
Vendor-only narrative Multiple independent sources
No failure stories Failure stories available
Buzzword density Production references
No production references Clear limitations
Unrealistic timelines Realistic timelines

If red flags outnumber green flags, you're evaluating hype, not opportunity.


5. Connecting scanning to strategy — How horizon intelligence becomes investment decisions

Many organisations run scanning exercises that produce reports nobody acts on. Scanning that doesn't change decisions is just reading. The gap between "interesting technology" and "funded initiative" is where scanning dies.

Scanning outcomes must connect to a decision chain that ends in investment or action:

flowchart LR
    SCAN["📡 Scan\nSources + Signals"] --> EVAL["🧐 Evaluate\nCriteria + Scoring"]
    EVAL --> PLACE["🎯 Place\nRing + Quadrant"]
    PLACE --> DECIDE["⚖️ Decide\nAdopt / Trial\nAssess / Hold"]
    DECIDE --> FUND["💰 Fund\nInvestment\nor Pilot budget"]
    FUND --> MEASURE["📊 Measure\nOutcomes\n+ Velocity"]

    BREAK1["❌ Chain breaks here\nNo evaluation"] -.-> SCAN
    BREAK2["❌ Chain breaks here\nNo decision"] -.-> PLACE
    BREAK3["❌ Chain breaks here\nNo funding"] -.-> DECIDE

Scanning-to-strategy mapping:

Scanning output Strategy action
"Technology X is ready for adoption" Add to standards register
"Technology Y is worth trialling" Fund a pilot
"Technology Z is emerging" Monitor; assign to team for learning
"Technology W is declining" Plan migration; deprecate

The decision framework:

Radar placement Velocity Decision Funding
Adopt Steady or Rapid Standardise Include in platform budget
Trial Steady Pilot Pilot budget (3-6 months)
Trial Rapid Accelerate pilot Increased pilot budget
Assess Any Learning budget 10-20% time or hackathon
Hold Declining Migration budget Deprecation planning

The investment decision tree: Technology placed on radar → Adopt ring → Velocity: Steady → Standardise in next budget cycle. Velocity: Rapid → Accelerate — immediate budget reallocation. Trial ring → Strategic fit: High → Fund pilot (dedicated team, 3-6 months). Strategic fit: Low → Monitor only — no funding. Velocity: Rapid → Increase pilot scope. Assess ring → Multiple signals → Fund learning (hackathon, 20% time). Single signal → Monitor passively. Hold ring → Currently in use → Fund migration. Not in use → Document rationale, revisit in 6 months.

Track scanning investments as a portfolio, not individual bets: Adopt — standardisation — 60% of tech budget. Trial — pilots — 25% of tech budget. Assess — learning — 10% of tech budget. Hold — migration — 5% of tech budget. If >70% is in Adopt, the organisation isn't scanning enough. If >40% is in Assess, the organisation isn't deciding enough.

Scanning → Radar → Decision → Investment. If the chain breaks, scanning is wasted.


Try it yourself — The chain check

Where does your scanning chain break?

Link in chain Working? If not, why?
Scanning → Evaluation
Evaluation → Radar placement
Radar placement → Decision
Decision → Investment

If any link is broken, your scanning isn't changing decisions.


6. Governance of emerging technology adoption — How to experiment without creating chaos

If every new technology requires full governance review, teams stop experimenting. If no governance applies, teams adopt anything and create a fragmented landscape. The answer is tiered governance — match the rigour to the maturity.

Emerging technologies need different governance than established standards:

flowchart TD
    GOV["🏛️ Tiered Governance"] --> EXP["🧪 Experiment\nH3 — Emerging\nNo governance\nLearn freely"]
    GOV --> TRIAL["🔬 Trial\nH2 — Maturing\nLightweight\nPilot scope, review"]
    GOV --> ADOPT_FULL["✅ Adopt\nH1 — Ready\nFull governance\nStandards, exceptions"]

    EXP --> RULES_EXP["Rules:\nDefined scope + timeline\nNo production impact\nDocument learning"]
    TRIAL --> RULES_TRIAL["Rules:\nPilot scope defined\nReview outcomes\nSuccess criteria set"]
    ADOPT_FULL --> RULES_ADOPT["Rules:\nStandards register\nException process\nCompliance required"]

Governance tiers:

Governance level Technology maturity What governance applies
Experiment H3 — emerging No governance — learn freely
Trial H2 — maturing Lightweight — pilot scope, review outcomes
Adopt H1 — ready Full governance — standards, exceptions

Rules for experimentation: Defined scope and timeline — experiments end — they don't become permanent. No production impact — experiments are safe — no customer risk. Documented learning outcome — learning is preserved whether it succeeds or fails. Success → progress to trial — clear path from experiment to adoption. Failure → documented — failed experiments are valuable — don't lose the learning.

Rules for trials: Pilot scope defined — 1-2 teams, specific use case, not everything. Success criteria set — know what "success" means before starting. Review outcomes — formal review at end of pilot. Go/no-go decision — trial leads to Adopt or Hold — not indefinite.

The governance escalation model: Experiment (no governance) → Success → Trial (lightweight governance) → Success → Adopt (full governance) → Declining → Hold (migration governance). At each transition: increase documentation requirements, increase review frequency, increase compliance scope, decrease freedom to deviate.

Anti-patterns: Governance-free zone — "Experiments never end." Fix: define timeline; review at end. Governance creep — "Every experiment needs a committee." Fix: keep it lightweight — one sponsor, one objective. Orphan experiments — "Nobody documented what we learned." Fix: require learning documentation as exit criteria. Shadow adoption — "We experimented in production." Fix: experiments must not affect production.

Govern less for experiments, more for adoption. Let teams explore; govern what goes to production.


Try it yourself — The governance check

What governance applies to your current technology experiments?

Experiment Governance level Rules in place?

If everything is at the same level, you're not tiering governance.


7. Building a scanning practice — Who does it, how often, and what it produces

Scanning done as a one-time exercise produces a snapshot that's outdated within weeks. Scanning must be a practice — assigned to people, on a cadence, producing regular outputs that drive decisions.

A scanning practice has five elements — who, cadence, inputs, outputs, and communication:

flowchart TD
    PRACTICE["🏗️ Scanning Practice"] --> WHO["👤 Who\n1-2 architects\n20% time allocation"]
    PRACTICE --> CADENCE["📅 Cadence\nMonthly source review\nQuarterly radar update"]
    PRACTICE --> INPUTS["📥 Inputs\nIndustry + Vendor\nCommunity + Internal"]
    PRACTICE --> OUTPUTS["📤 Outputs\nRadar + Recommendations\nAction items"]
    PRACTICE --> COMMS["📢 Communication\nShare with all teams\nQuarterly town hall"]

    WHO --> HEALTH["✅ Healthy Practice\nDecisions change\nTeams consult radar\nBoard informed"]
    CADENCE --> HEALTH
    INPUTS --> HEALTH
    OUTPUTS --> HEALTH
    COMMS --> HEALTH

Practice elements:

Element What to establish
Who 1-2 architects with scanning as part of their role
Cadence Quarterly radar update; monthly source review
Inputs Industry sources, vendor updates, team experiments
Outputs Technology radar, recommendations, action items
Communication Share radar with all teams — quarterly town hall

The scanning health check:

Dimension Healthy Unhealthy
Cadence Quarterly update Last update >6 months ago
Coverage All four quadrants covered Only one quadrant scanned
Action Radar changes lead to decisions Radar is published but ignored
Participation Teams contribute observations Only architects scan
Velocity tracking Ring changes measured over time No tracking of technology momentum

Assign scanning to 1-2 people, quarterly cadence, radar as output, decisions as outcome — make it a practice, not a one-time exercise.


Try it yourself — The practice check

Do you have a scanning practice?

Element In place? Gap
Who
Cadence
Inputs
Outputs
Communication

If fewer than three are in place, you don't have a practice. You have an ad-hoc activity.


Putting it all together

Here's the complete picture in one diagram. This is the mental model worth internalising.

flowchart TD
    PROBLEM["⚡ Technology Surprise\nOrganisations discover shifts\ntoo late — panic response"]
    SOLUTION["🔍 Scanning Practice\nMonitor · Filter · Evaluate\nDecide · Act"]
    TOOLS["🎯 Tools\nTechnology Radar\nAdoption Velocity\nHype Evaluation"]
    GOVERN["🏛️ Governance\nTiered by maturity\nExperiment → Trial → Adopt"]

    PROBLEM --> SOLUTION
    SOLUTION --> TOOLS
    SOLUTION --> GOVERN

    TOOLS --> OUTCOME["✅ Outcome\nLead time gained\nDecisions informed\nCompetitive advantage"]
    GOVERN --> OUTCOME

The foundation:

Scanning creates lead time — don't wait to be surprised by technology shifts. If you discover a shift when your competitor has already adopted it, you're too late. Evaluate at the Trough of Disillusionment — hype is gone, reality is clear. Don't adopt at the peak of hype. Wait for reality. Scanning is only valuable if it changes decisions. If scanning doesn't lead to Adopt/Trial/Assess/Hold decisions, stop doing it. Scanning = filter signal from noise — only valuable if it changes decisions.


Cheat Sheet — All the key terms

Concept One-Line Memory Key Action
Scanning Filter signal from noise Structured process, not blog reading
Horizons Now (0-6m), Next (6-18m), Later (18-36m) Scan all three
Technology Radar Adopt / Trial / Assess / Hold Shared vocabulary for technology position
Adoption Velocity Measures ring-change speed Track momentum, not just position
Hype cycle Evaluate at the Trough Don't follow hype
Strategy connection Scanning → Decision → Investment Chain must not break
Governance Light for experiments, full for production Match governance to maturity
Practice Quarterly cadence, assigned people Make it ongoing

How to know if this landed

You'll know this has landed when someone stops discovering technology shifts when competitors have already adopted them. Can explain the three horizons (H1, H2, H3) and what action each requires. Technology radar exists and is updated quarterly — not a one-time exercise. Adoption Velocity is tracked — know which technologies are accelerating. Evaluation criteria are defined — weighted scorecard, not gut feel. Scanning outputs connect to decisions — radar changes lead to investment or deprecation. Governance is tiered — experiments are lightweight, adoptions are full governance. And scanning practice is assigned — 1-2 architects with 20% time, quarterly cadence.


What changes when the mental model clicks

I've run this session with an insurance company where a competitor launched AI-powered claims processing — 10x faster. Company had no scanning practice — discovered the shift 18 months late. Rushed AI adoption — poor quality, no governance, expensive mistakes. The gap at the start is usually not about understanding scanning — it's about not having a structured practice.

What changes after this session:

Teams stop being surprised by technology shifts and start seeing them early. The technology radar exercise — "place your technologies in rings and quadrants" — is always the moment things click. People stop treating scanning as blog reading and start treating it as a decision-changing practice. Their results get better. They stop adopting at the peak of hype when the real problem was that they hadn't learned to evaluate at the Trough.

The scanning-to-strategy mapping exercise tends to immediately change how teams think about their scanning outputs. They start connecting radar placements to investment decisions. Their clarity goes up. They stop producing reports nobody acts on when the real problem was that the chain between scanning and investment was broken.


Book a Workshop

Ready to build a scanning practice that keeps you ahead of technology shifts?

→ Book a Training Session

or

→ Contact me directly

Half-day workshop includes scanning practice design — who, what, when, how, technology radar creation — build your first radar with real technologies, Adoption Velocity measurement — track technology momentum through the radar, hype vs signal evaluation — apply criteria to current emerging technologies, strategy connection — link scanning outputs to investment decisions, governance tiers — design experiment, trial, and adoption governance, and technology radar template (including Adoption Velocity metric) + horizon scanning process + hype vs signal evaluation criteria.

Related Trainings

Build / Buy / Partner / Participate — Ecosystem Strategy
Intermediate to Advanced (SA, EA, TS)5-6 hoursBuild/buy/partner/participate decision framework + vendor and ecosystem evaluation scorecard + exit and switching cost analysis template

How to decide what to build, what to buy, what to access through partners, and what to participate in through ecosystems — covering differentiation vs commodity analysis, vendor evaluation, total cost of ownership, interoperability velocity, exit strategy, and lock-in risk.

1 DayIntensive
StrategicSA · EA · TS
Operating Model Choices
Advanced (EA, TS)5-6 hoursOperating model design canvas + team topology guide + Conway's Law application worksheet + Communication Structure Heatmap

How the structure of technology teams shapes the structure of the systems they build — covering Conway's Law, centralised vs federated models, platform teams, product operating models, team topologies, and designing the organisation that produces the architecture you want.

1 DayIntensive
StrategicEA · TS

Next Step

Run this with your team

Every programme is adapted to your context before delivery — your systems, your constraints, your decisions. A short call is enough to work out the right shape and scope.