SME AI Research Round-up — May 2026

01 · Industry Survey · March 2026

Goldman Sachs — 10,000 Small Businesses Voices

Babson College & David Binder Research · n=1,256 SMBs · all 50 US states + DC + PR

What it means — in plain English

Three out of four small-business owners are already using AI, and almost all of them say it's helping. But only 1 in 7 have actually woven it into how they run the business day-to-day. The gap between "I use ChatGPT sometimes" and "AI runs my workflows" is the real prize — and most owners say they'd take help getting there.

76%

Currently use AI

93%

Report positive impact

14%

Fully integrated to core ops

73%

Want training & support

Primary benefit — 84% cite increased efficiency and productivity
Workforce impact — 87% say AI augments rather than replaces employees; 67% expect AI to increase revenue
Top barriers — 50% data-privacy concerns · 49% lack technical expertise · 48% can't choose the right tools
Survey window — Jan 27 to Feb 4, 2026; release dated 16 Mar 2026

Source: goldmansachs.com — Small Businesses Embrace AI But Need Training and Support to Fully Harness It

02 · Op-Ed · Fast Company

Small businesses should be a much bigger part of the AI conversation

Fast Company · Op-ed + reporting · How SMEs actually capture value from AI today

↓ Read the source

What it means — in plain English

Nearly every small business has tried AI for something — usually writing or simple reports. The ones winning are using it for repetitive tasks, 24/7 customer chat, marketing optimisation, and lead-response speed. The mental shift: stop "experimenting with ChatGPT occasionally" and start baking it into the daily workflow.

90%

Use AI for ≥1 process

93%

Find it cost-effective

$Bs

Saved annually (SMB cohort)

Concrete value areas

Where AI creates real value — automation of repetitive tasks · 24/7 customer service · marketing optimisation · lead-response speed · data-driven decision-making
Headcount unlock — AI lets a small team scale operations without proportional hiring, levelling the field vs. larger competitors
Where to start — pick ONE time-consuming problem (lead follow-up, CS wait time, inventory forecast), run a low-cost pilot, then make it daily
Invest in training — adoption becomes strategic, not superficial; the article's core argument is that SMEs deserve more attention because the gap from "tool user" to "AI-native operator" is closeable

Source: fastcompany.com — Small businesses should be a much bigger part of the AI transformation conversation

03 · Live Case Study · Claude + Manus

Rebel Cheese recovered $400K from shipping overcharges using AI

Texas vegan cheese maker · Mark-Cuban-backed · built by non-technical founder Kirsten Maitland

↓ Read the source

What it means — in plain English

A non-technical founder built an AI agent that audits every shipping invoice line-by-line against the carrier contract, even looking at box photos for tiny bulges that trigger surcharges. It has clawed back $400,000 in overcharges. The catch: it took months of nights, multiple agent attempts, and a human still reviews flagged items. The ROI is enormous, but it's not yet plug-and-play.

$400K

Recovered to date

<$40K

Recent monthly recoveries

1/8″

Box-bulge threshold

Engineers on the team

The stack — Anthropic Claude (analysis & reasoning) + Manus (agentic orchestration) — compares weekly invoices line-by-line against the shipping contract
What it catches — dimensional surcharges, zone-pricing errors, mis-applied fees; even uses photos of boxes to flag dimensional disputes
The build cost — months of iteration, long nights refining prompts, multiple agents tried before Claude+Manus stuck; not a weekend project
Human-in-the-loop — exec assistant still uploads invoices; Maitland reviews flagged discrepancies before disputing
The takeaway for operators — AI agents for SMEs deliver real ROI, but deployment still requires founder time, technical curiosity and tolerance for iteration. This is exactly the "implementation gap" the Goldman Sachs study identifies

Source: fastcompany.com — Texas Rebel Cheese AI shipping · Fortune — Rebel Cheese × Mark Cuban × AI shipping · anthropic.com — Claude Cowork

Goldman Sachs — 10,000 Small Businesses Voices

Small businesses should be a much bigger part of the AI conversation

Rebel Cheese recovered $400K from shipping overcharges using AI

What this means for the Gulf SME operator