Last Updated on May 14, 2026 by Craig Allen Keefner
NAMA Los Angeles (Post-Show) + the 2024–2025 Census: growth is real, but the category is changing
The NAMA show in Los Angeles is now in the books—and the 2024–2025 State of Convenience Services Census is out. Together, they confirm what operators already feel day-to-day:
This is no longer “just vending.” It’s becoming a distributed network of self-service retail endpoints, increasingly defined by software, connectivity, and format flexibility.
The “vending industry” label will persist, but what’s forming underneath is clearer than ever:
A network of intelligent, self-service retail endpoints—managed like platforms, not like machines.
Review — NAMA 2026: AI-drives opportunities for unattended retail; 9 takeaways to consider
Interesting Vending Industry News
-
FTC Requires 365 Retail Markets to Divest Micromarket Kiosk Business to Clear Acquisition of Cantaloupe
- The FTC is forcing a structural remedy (divestiture): 365 Retail can complete its $848M acquisition of Cantaloupe only if it divests Cantaloupe’s Three Square Market (TSM) micromarket kiosk business to Seaga. That’s the FTC saying the deal, as-is, would likely reduce competition.
- The FTC is also imposing conduct/behavioral constraints: the proposed order includes non-discriminatory integration/interoperability obligations, a monitor to oversee compliance, and a 10-year prior-notice requirement for future micromarket-kiosk acquisitions. Those are meaningful ongoing constraints, not just a one-time asset sale. Federal Trade Commission
- What it implies about the market: the FTC characterizes this as combining the two largest micromarket kiosk providers, with concerns it could raise kiosk/software costs that may be passed through into higher micromarket food prices and/or reduced quality.
What the Census actually says
For this Census, vending includes traditional vending machines and smart coolers. Definitions and methadology matters. Trust but verify is good advice.
Industry size and “real” growth
- Total industry revenue: $31.1B in 2025 (preliminary), up from $26.6B in 2023.
- Average annual growth since 2023: 8.1%, moderating to ~6.8% in 2025.
- After pricing, real growth is ~2–3% over 2024–2025.
Segment mix (2025 preliminary)
| Business line | 2025 revenue | Share of total |
|---|---|---|
| Vending (includes smart coolers) | $20.6B | 66% |
| Micro markets | $6.8B | 22% |
| Office Coffee Service (OCS) | $3.2B | 10% |
| Pantry | $471M | 2% |
Micro markets are explicitly called out as the primary engine of growth, with vending growing more modestly.
The “blur” is now official
- 41% of vending operators run smart coolers.
- Among those using them, smart coolers are ~8% of vending revenue and average ~180 coolers per company.
- ~70% of businesses operate multiple service models, and the Census says the distinctions between vending, smart coolers, and micro markets are “less rigid.”
What the show floor confirmed (NAMA 2026 LA): the stack is the product
From our NAMA 2026 show report, the strongest signal wasn’t a single “killer machine”—it was the steady march toward a common architecture:
- Hybrid edge + cloud (edge for resilience/latency, cloud for fleet analytics).
- Connectivity moving from “nice-to-have” to a core dependency as more functions become cloud-attached.
- Smart-store concepts are advancing, but wide-scale rollout remains constrained by investment and consumer education—while more pragmatic formats keep winning deployments.
Updated take: what NAMA will call it vs what it means
The structural transition (measurable, not hypothetical)
1) Vending isn’t disappearing—its definition is expanding
The Census explicitly includes smart coolers inside vending. That matters: it signals a category shift from “spirals and cans” to “controlled-access retail nodes.”
2) Micro markets are scaling—but the “next market” is smaller and more controlled
Micro markets are estimated at 90,000–100,000 operations in 2025. Growth is moderating as large vending-bank conversions are largely done; operators increasingly target mid-sized locations (100–250 employees) and tighten shrink control with controlled-access formats where needed.
3) Location economics are tilting industrial
In vending, manufacturing/factory is the largest share of placements (27%). Micro markets also skew heavily toward manufacturing in deployment mix. This reinforces the “hybrid workplace stabilization” reality: white-collar is important, but industrial is dependable.
The 2030 lens (what the Census projects)
- Total industry nominal growth through 2030: ~6.5% CAGR.
- Projected CAGRs by line (2026–2030): vending ~4.9%, micro markets ~10.3%, OCS ~7.9%, pantry ~6.5%.
Translation: the Census itself is telling you the mix shift continues—micro markets keep pulling the curve.
Bottom line
The Census validates the story, and the show confirms the mechanism:
- Growth is real (with pricing doing a lot of work).
- Format convergence is accelerating (vending now formally includes smart coolers; multi-model portfolios are the norm).
- Technology is central (connectivity + hybrid edge/cloud + product recognition are reshaping operations).
The “vending industry” label will persist, but what’s forming underneath is clearer than ever:
A network of intelligent, self-service retail endpoints—managed like platforms, not like machines.
NAMA Gap Analysis by TIG (2026)
Positioning: Solid on vending/unattended retail core. Lagging in broader “self-service infrastructure intelligence.”
1. Strategic Scope Gap
Where NAMA is strong
- Vending, micro markets, unattended retail
- Payments, cashless, telemetry
- Operator-focused business models
Gap
- Under-positioned in full-spectrum self-service ecosystem
- Kiosks (QSR, healthcare, government)
- Digital signage + wayfinding
- Accessibility (ADA / EAA / EN 301 549)
- Edge AI + compute lifecycle
Implication
They are seen as category association (vending), not platform authority (self-service)
2. AI & Edge Compute Gap
Current state
- Some coverage of smart vending, computer vision, frictionless retail
Gap
- No structured framework for:
- Edge AI vs cloud decisioning
- Retrofit vs replace economics
- AI hardware stack (Intel, Hailo, ARM, etc.)
- Lifecycle planning (5–7 year deployments)
- “No guidance on adding a $150–$200 edge AI module to extend lifecycle”
- “No framework for when inference belongs on-device vs cloud”
Examples — Meldx easily adds AI reporting and analysis.
In some deployments Aramark explicitly uses AI (e.g., computer vision + machine learning) in its “smart vending” and self-checkout offerings, which implies AI in the operational stack behind those endpoints (item recognition, tracking, analytics/insights). https://www.aramarkconnected.com/smart-vendingaramarkconnected.com
If we look at classic snack/soda vending backend optimization (forecasting, route optimization, dynamic planograms) across their general fleet, Aramark doesn’t publicly document that at a “yes/no, companywide” level in a way we can verify—only these AI-forward product lines are clearly stated.
Examples of AI called out in our NAMA 2026 show review:
- Hybrid edge AI + cloud AI architectures (e.g., Grabot, Aeritek, Moneta Market): edge AI on-device for resilience/latency (e.g., age verification, ordering) and cloud AI for analytics/training, dashboards, remote monitoring.
- Computer vision “smart coolers”:
- Aeritek A‑Pop “powered by Vision AI” (multi-camera CV; can run edge or cloud AI; aims at item recognition + shrink control).
- Cantaloupe Coke & Go Cooler using “dynamic vision and AI” to charge for items removed (“buy & go”).
- AI-enabled cashierless / smart-store systems:
- Cantaloupe Smart Aisle using AI + 3D cameras + weighted shelves to build a virtual cart.
- Smart stores broadly described as using AI computing, smart shelves, and robotics to improve economics and usability (but scaling still uncertain).
- AI as an operator “copilot” (analytics/insights):
- Cantaloupe Seed Copilot: AI-powered natural-language Q&A over operator data/insights (Spotlight).
- “AI-powered” unattended retail nodes beyond vending:
- CeresGo described as an “AI powered vending machine” with an intelligent operating platform for real-time monitoring, inventory management, and analytics.
- Biometric + AI for verification:
- Innovative Technology age estimation tool; AI Tended Bar facial recognition for age/ID checks and linking payment to tabs (your piece notes privacy sensitivity).
Opportunity — “AI for unattended ≠ AI for kiosks ≠ AI for healthcare”
3. Accessibility & Compliance Gap
In regulated sectors, lack of accessibility isn’t a gap—it’s a deal killer
Critical miss
- Limited leadership on:
- ADA (2010 refresh implications)
- European Accessibility Act (EAA)
- Section 504 enforcement (2026 deadline)
Why it matters
- Accessibility is becoming a deployment blocker, not a feature
- Especially in:
- Healthcare (HIMSS alignment)
- Government kiosks
- Public retail environments
Gap severity: HIGH
NAMA risks being non-authoritative in regulated environments
4. Enterprise Deployment Gap
NAMA focus
- Operators (route-based, vending fleets)
Gap
- Limited guidance for:
- Enterprise IT buyers
- CIO / CTO / Digital Transformation leaders
- “Digital Front Door” strategy
Missing frameworks:
- TCO modeling at scale
- Device lifecycle + serviceability
- OS fragmentation risk
- Cloud dependency risk
5. Standards & Governance Gap
Current
- Some payment standards, industry best practices
Gap
- No central role in:
- Cross-standard mapping (ADA vs EN vs EAA)
- ANSI alignment
- “standards risk scoring”
This is a major opening for:
- American National Standards Institute alignment narratives
- Global compliance frameworks
6. Content / Media Gap
NAMA
- Event-driven content (NAMA Show, Coffee Tea & Water)
- Member-focused updates
Gap
- No:
- Deep-dive intelligence briefs
- ROI tools (retrofit vs replace)
- C-suite explainers
- Independent vendor-neutral analysis
Translation
NAMA informs operators.
It does not guide strategic buyers or investors.
7. Global Strategy Gap
NAMA
- Primarily North America
Gap
- Limited perspective on:
- China (AI retail, QR ecosystem)
- Europe (regulation-driven deployments)
- Asia-Pacific scale deployments
This is a big miss given:
- China = largest self-service deployment lab globally
- EU = regulatory driver
- US = fragmented innovation
Bottom Line
NAMA is operationally relevant but strategically narrow. We read the brief interview on vending times – Short take: solid association messaging… but very little operational or strategic depth.
They own:
- Vending
- Operator economics
- Payments evolution
They do NOT own:
- Self-service infrastructure strategy
- Accessibility leadership
- AI deployment frameworks
- Enterprise-scale guidance
Where TIG Wins
Where we fill the gaps:
- Pillars → hardware / software / services
- Standards hub → ADA, EAA, EN 301 549
- AI hub → edge, retrofit, Intel/Hailo
- ROI tools → Gumroad assets
- Global lens → kioskasia.org
