Welcome to this week’s edition of the Smarter Than Busy. This week, we dive into the chasm between “trying AI” and “scaling AI,” plus look at two major updates from Google and OpenAI that are changing how small businesses work.

Let’s dive in!

📰 AI News That Impacts SMBs

Insights and stories at the intersection of AI and small business: what’s changing, why it matters, and how it actually hits the ground

McKinsey: State of AI 2025 — What’s Actually Going On

McKinsey’s 2025 AI report basically confirms what a lot of us already see on the ground: almost every company claims they’re “using AI,” but in reality, most are just testing things in small pockets of the business. It’s experimentation, not transformation.

The new big wave is agentic AI (AI that can actually take a sequence of actions instead of just answering a prompt). Still, barely anyone is using it in a meaningful, scaled way.

Meanwhile, there’s a tiny group of companies that McKinsey calls “AI high performers” who are capturing almost all of the upside. Why? They rebuilt processes, invested in data infrastructure, and aligned leadership. Everyone else is still poking around with tools.

On top of that, workforce changes are accelerating, and AI governance is now a major blocker.

🔑 Key Insights: McKinsey State of AI 2025

• Usage is Wide, Scaling is Narrow: While 88% of companies use AI somewhere, only one-third have managed to scale it across their organization.

• The Agent Gap: 62% are experimenting with agentic AI (AI that takes multi-step actions), but fewer than 10% have scaled these systems in any function.

• Impact is Uneven: Only 39% report meaningful company-wide financial gains, and the profit effect for most is still under 5%.

• Innovation is Strong: 64% say AI is directly enabling new products, services, or ways of working.

• Workforce Changes: Expect varied effects: 43% anticipate steady headcount; 32% expect reductions; and 13% expect increases.

• AI Risk is Pervasive: Over half of companies have experienced an AI-related failure, with inaccuracy being the top recurring problem.

💡 Why This Matters

Every business now touches AI in some shape or form. But very few are actually seeing revenue or operational gains from it. The gap between “we tried a tool” and “we redesigned a workflow around AI and its printing ROI” is getting wider by the month.

And suppose your competitors fall into that “high performer” category. In that case, they’re building advantages you’re not going to close overnight—speed, cost efficiency, better products, and more automation baked right into their operations.

👉 Read The Full Post

Goldman Sachs: Small Businesses Are Surprisingly Optimistic — Thanks to AI

Goldman Sachs surveyed 1,400 small business owners ahead of its 10,000 Small Businesses Summit. Even with rising costs, tighter access to capital, and general economic uncertainty, owners are surprisingly optimistic about the year ahead.

The common thread behind that optimism: AI is finally becoming a practical, everyday tool that saves time and reduces stress.

That time back is a big part of why confidence is going up, not down.

🔑 Key Points: Goldman Sachs Small Business Survey

• High Optimism: 78% of small business owners are optimistic about the next 12 months, despite rising costs and economic uncertainty.

• AI is Delivering: A significant 94% of AI users report that the technology is actively helping their business.

• Practical Use Cases: Owners are using AI for everyday tasks like menus, flyers, hiring materials, and inventory forecasting, rather than complex tech projects.

• Productivity, Not Replacement: AI is primarily viewed as a productivity enabler, allowing businesses to do more without replacing employees.

• New Growth Playbook: Large companies are demonstrating the shift by growing revenue without adding staff, highlighting how AI-driven efficiency is changing the growth model.

💡 Why This Matters

For small business owners, the message is clear: AI is becoming a genuine advantage, not a buzzword. It’s giving everyday operators more speed, more clarity, and more breathing room.

The old growth model, “hire more people to get more done,” is getting replaced by something much more flexible. AI gives you leverage without adding headcount, helping your team punch well above its weight.

This isn’t about cutting jobs. It’s about removing busywork so your people can focus on the real work.

Adaptability is the new competitive edge, and AI is making it accessible to everyone.

👉 Read The Full Post

🛠️ Tool Updates / Feature Highlights

Two fresh updates this week that make everyday work simpler without adding headaches

Google Launches Gemini 3, its ‘most intelligent’ AI model yet

Google and DeepMind have released Gemini 3, their most advanced multimodal AI model, marking one of the most significant capability jumps in the Gemini lineup.

It’s significant because Gemini 3 moves beyond traditional chatbot behavior, offering deeper analysis, richer media understanding, and the ability to execute multi-step operations across Google Workspace and enterprise systems.

🗓️ Released November 18th 2025

⚙ Access Level: Only available on paid Workspace Business, Enterprise, and Education plans (not supported on personal or free Google accounts).

🚀 How Gemini 3 Helps

Gemini 3 upgrades every SMB workflow by boosting accuracy, reducing manual review, and automating more analysis.

• Parses complex materials fast: Ingests text, images, audio, and video to produce transcripts, summaries, metadata, and structured insights—cutting hours of manual sorting.

• Supports financial and legal decisions: Assists with modeling, supply chain adjustments, and contract/terms evaluation, enabling teams to move faster with fewer errors.

• Automates multi-step tasks in Workspace: Runs analyses, generates insights, and completes procedural work directly inside Docs, Sheets, and Drive.

• Handles nuanced reasoning: Deep Think mode improves decisions that require layered logic across finance, legal, and strategic planning.

• Builds interactive tools: Generates working UI components and lightweight app experiences from plain-language instructions—letting small teams ship more without engineering resources.

🧩 Key Features

• Multimodal comprehension: High-accuracy processing across text, images, audio, video, and code.

• Deep Think mode: Stronger step-by-step reasoning for complex analytical tasks.

• Agentic workflow execution: Orchestrates and completes reliable, multi-step processes across connected tools and systems.

• Workspace-native operations: Work directly within Docs, Sheets, and Drive to analyze data and generate content.

• Enhanced tool + API usage: Executes long-running tasks and integrates with internal systems reliably.

• Interactive UI/app generation: Creates dynamic interfaces and lightweight apps from natural language prompts.

• High-capacity media output: Produces long-form transcripts and structured insights from extended multimodal inputs.

• Scaled performance: Handles massive context workloads and automation at enterprise-level volume.

💼 Gemini 3 Use Cases (Multimodal & Workspace Focus)

These use cases show how Gemini 3 can manage mixed media and streamline tasks in Google Workspace.

1. 🏥 Example Use Case #1: Payroll Specialist (Sarah)

Sarah is a Payroll Specialist at a small accounting firm that handles payroll for various small business clients. She ensures payroll is accurate, organized, and delivered on time.

Her tasks include collecting timesheets, checking hours and pay rates, preparing payroll drafts, updating employee details, tracking tax deadlines, sending pay stubs, and answering client questions.

But one task in particular takes a lot of her time: entering payroll data from mixed-format timesheets and contractor invoices.

Each pay period, Sarah receives unclear and slow-to-process formats. She has to open each file, find the correct values, and retype them into the payroll system. Handwritten photos and poorly formatted PDFs slow her down and increase the risk of mistakes. This repetitive process takes hours each week.

⚡ How Sarah Uses Gemini to Help

Using Gemini 3’s multimodal document-parsing and Workspace-based workflows, Sarah can:

• Gather all timesheets and invoices into a single folder or Gmail label: Gemini’s multimodal parsing reads PDFs, images, photos, spreadsheets, and email text directly.

• Extract the payroll-relevant data: Gemini pulls employee names, hours worked, rates (if included), totals, contractor invoice amounts, and dates.

• Normalize the mixed formats: Whether the source is a photo, PDF, spreadsheet, or email, Gemini converts the extracted data into one consistent structure.

• Generate a clean payroll table: All data is output as a ready-to-review table directly into Google Sheets.

• Flag potential issues: Highlighting missing hours, inconsistent rates, duplicated records, and unusual totals automatically.

• Prepare import-ready rows for the payroll system: Gemini formats the data so Sarah can paste or upload it directly into her payroll software.

• Gemini provides a short review summary: a concise text outlining any items requiring her attention before running payroll.

🎯 Result

Sarah no longer manually retypes payroll information. Gemini 3 extracts and organizes all payroll data into a clean structure within minutes. This eliminates repetitive data entry, reduces errors, and saves hours each payroll cycle.

With less time spent on copying, Sarah can focus on reviewing anomalies and ensuring accurate payroll processing.

2. 🏥 Example Use Case #2: Quality Control Supervisor (Luis)

Luis is the Quality Control Supervisor for a residential and commercial cleaning service. He ensures the cleaning team maintains high-quality work by visiting properties post-cleaning, evaluating cleanliness, and communicating issues back to the team.

His most time-consuming task is documenting site inspections. He takes photos, types notes, and sometimes records videos during property visits. Afterward, he must sort through the media, rewrite summaries, and send messages, resulting in a slow and repetitive process.

⚡ How Luis Uses Gemini to Help

Using Gemini 3’s multimodal capabilities and Workspace-based workflows:

• Luis uploads photos and videos from the inspection into a single Drive folder or Gmail thread.

• Gemini reads all the media directly: images, videos, screenshots, and text.

• Gemini analyzes photos/videos and, multimodally, extracts all relevant quality issues. It identifies: missed areas, damage, streaking, clutter left behind, or items that need revisiting.

• It converts all inspection findings into a structured, easy-to-read report. Gemini outputs a standardized QC summary listing each issue, room by room, with attached reference images.

• Gemini generates clear, actionable action items for the cleaner: It turns each issue into a fix list with simple corrective instructions.

• It highlights patterns or recurring problems: Gemini identifies if certain types of issues keep appearing (e.g., baseboards missed, dusting incomplete).

• It prepares a message Luis can send to the cleaner or owner: Gemini drafts a concise communication that includes a full report, a list of what needs addressing, a deadline or priority, and recommended improvements.

🎯 Result

Luis no longer spends hours sorting photos or writing inspection notes. Gemini 3 automatically analyzes images and videos from walkthroughs, generating standardized quality control reports in minutes.

Quality checks happen faster, making it easy to spot recurring issues, and everyone on the team communicates better. Luis can manage multiple locations without feeling rushed, ensuring clients receive cleaner, more consistent standards. It’s a win-win for everyone!

👇 See ‘Gemini 3’ Demo

OpenAI Releases GPT-5.1 for Paid ChatGPT Users

OpenAI recently launched GPT-5.1, the newest model in the GPT-5 family, (for paid users) it began rolling out in mid November. This update brings faster performance, stronger instruction-following, more natural conversation, and new controls that make outputs more predictable and on-brand.

GPT-5.1 is not a total reinvention—it’s a refinement—but the improvements directly address the pain points small businesses experience every day: inconsistent tone, unpredictable drafts, slow responses, and limited automation.

🗓️ Released November 12th 2025

⚙ Access Level: Only available on ChatGPT Plus, Team, Enterprise, and EDU plans (not included on the free tier).

🚀 How GPT-5.1 Helps

GPT-5.1 reduces the need for revisions and double-checking. It provides precise, straightforward answers, making it ideal for replying to customers, writing updates, and decision-making.

• Instant Mode: Quickly manages emails, replies, and summaries without hassle.

• Cleaner Writing: Produces solid first drafts, minimizing extensive rewrites.

• Consistent Voice: New Steerability Controls ensure a unified tone across all communications.

• Better Decision-Making: Thinking Mode outlines the pros and cons of pricing, scheduling, and purchasing.

• No Downtime: Automatically integrates with existing systems.

• Team Alignment: Shared tone settings ensure consistent communication.

• Complex Problem-Solving: Codex Max efficiently handles extensive codebases and workflows.

🧩 Key Features

• Two Modes: Instant for quick tasks, Thinking for in-depth reasoning.

• Task Complexity Adaptation: Routes queries to the appropriate model.

• Tone Control: Maintains your business style consistently.

• Reliable Instructions: Correct formatting and structure on the first try.

• Codex Max: Designed for tackling large projects without memory issues.

💼 GPT-5.1 Use Cases (Tone control / structured, high-quality text generation)

1. 🏥 Example Use Case #1: Real Estate Agent (Ava)

Ava is the Lead Agent for a small residential real estate business. Her job is to guide buyers and sellers through every stage of a property transaction.

She prepares listings, conducts showings, explains contract terms, gathers updates from lenders and inspectors, and keeps deals moving forward across multiple active clients at once.

One task in particular consumes most of her time: Responding to the constant stream of client and partner messages that all require clear, customized, professional explanations and updates.

⚡ How Ava Uses GPT-5.1 to Help

Using ChatGPT 5.1’s abilities: reading text, summarizing, drafting, rewriting, organizing information, and adjusting tone.

• Ava pastes incoming texts or emails into ChatGPT, which identifies the client’s questions and those that need addressing.

• ChatGPT produces a polished response that matches her preferred tone control setting (Friendly, Professional, or Calming if needed).

• Ava updates on deals by noting: “inspection cleared, lender waiting on documents, appraisal scheduled for Friday."

• For clients who are stressed or anxious, ChatGPT adjusts its tone to be reassuring, empathetic, or firm, guided by Ava’s guidance.

• On busy communication days, Ava groups multiple messages together. ChatGPT arranges them into clear, ready-to-send drafts for quick replies.

• Ava prepares property highlights, and ChatGPT transforms them into a polished listing description for seller approval.

• To coordinate with vendors, she summarizes key points, and ChatGPT transforms them into concise, actionable instructions for inspectors, photographers, or lenders.

🎯 Result

Ava no longer spends her evenings rewriting messages or repeating the same explanations. ChatGPT 5.1 transforms raw notes, lengthy email threads, and client inquiries into polished, professional responses in just seconds.

Her communication is now faster, more precise, and more consistent. Clients feel supported and informed, leading to smoother transactions.

By reducing her writing workload, Ava frees up several hours each week that she can reinvest in property showings, negotiations, and building client relationships, all without the need to hire additional administrative support.

2. 🏥 Example Use Case #2: Tutoring Assistant (Mia)

Mia works as the Lesson Planning Assistant at a small tutoring business that helps students in different grades and subjects.

Because she receives information in various formats—like PDF report cards, parent emails, and teacher feedback Mia has to read and interpret each one separately. This takes a lot of time, as she has to summarize and convert the scattered information into organized lesson plans.

The biggest bottleneck of her job involves reviewing the scattered information for each student and creating a clear, organized lesson plan for the tutor.

⚡ How Mia Uses GPT-5.1 to Help

Using ChatGPT 5.1’s text-based capabilities and structured workflows:

• Mia collects essential student information (parent emails, teacher feedback, progress notes, session summaries) and inputs it into ChatGPT.

• ChatGPT analyzes the data to assess student strengths, weaknesses, and concerns, providing a concise overview of academic status and growth areas.

• A summary for tutors is generated, outlining key teaching priorities for the upcoming lesson.

• A structured lesson plan is created, including sections like Warm-up, Core Skill Practice, Applied Tasks, and Next Steps.

🎯 Result

Mia no longer spends hours creating lesson plans. ChatGPT 5.1 consolidates parent updates, teacher feedback, past notes, and student goals into structured plans within minutes.

The business achieves personalized instruction without needing extra staff, enhancing overall instructional quality and efficiency.

👇 See ‘CHATGPT 5.1’ (Prompting Guide)

🕵️‍♂️ Feature Spotlight Deep Dive

Each week we Spotlight a feature from one the major LLM / and Generative AI platforms that can help you gain time and work Smarter Than Busy.

In this weeks edition we are talking about ChatGPT Projects.

Let’s dive in.

The Advantage over “Just Using Normal Chats”

Regular chats can hold context, but they are not built for long-running work. Things drift. Instructions get buried. The model’s “understanding” shifts as the conversation grows.

Projects solve that. Yes, they organize your files and notes, but that’s not the most powerful feature.

The real power is that a Project gives ChatGPT a fixed foundation: the rules, background, goals, examples, and constraints that define how it should think for this specific job.

Instead of a general assistant, you get a purpose-built brain that stays locked onto one mission and doesn’t wander.

✨ The Power of ChatGPT Projects

• You can group multiple conversations under that Project to provide aggregate context, rather than scattering them across your chat history.

• You can upload reference files such as briefs, datasets, or docs to keep the AI aligned.

• You can set custom instructions specifically for that Project so ChatGPT behaves according to the rules and context of that workstream.

• The specific project keeps that context active, meaning the files, chats, and instructions all influence outputs without you having to re-explain things every time.

• On business/team plans, you can share Projects, so teammates operate on the same contextual foundation.

🧩 Key Features

• Workspace Segmentation: You split initiatives by project rather than cluttering everything into one chat.

• Persistent Custom Instructions: You define how the AI should think for this specific workflow, and that context is permanent.

• File Integration: Upload everything the AI needs and keep it centralised.

• History Aggregation: All conversations in the project share the same context bucket.

• Collaboration: Teams see the duplicate files, chats, and instructions when they enter the project.

• Context Memory: The Project becomes the single, non-volatile "memory zone" for the entire workflow.

• Focus & Relevance: Each Project stays clean, without unrelated chatter bleeding in.

👇 Below are tutorials and resources to help you master Projects this week:

• Projects in ChatGPT (Source OpenAI)

📚 Smarter Stack

You’re Not Behind (Yet): How to Learn AI in 17 Minutes Mastering AI in 30 Days – A Beginner-Friendly Roadmap. I came across this video and it breaks down a straightforward seven-step path to level up your AI skills.

33 Google AI Features You Won't Believe are Free 30 Free Google AI Tools – How to Actually Use Them. I like this one because it shows how far you can get with Google’s free AI stack:

AI photography prompts: how to create photorealistic images AI Photography Prompts – How to Create Photorealistic Images. I came across this guide and it does a great job showing how clearer, more intentional prompts lead to far more realistic AI photos.

STOP Wasting Money on More AI Tools. Do THIS instead AI Workflows – Turning ChatGPT, Gemini, Claude & Co. into a Real System. I like this video because it stops the “which AI is best?” nonsense and treats each model as a specialist.

And that’s a wrap! 🎬

Thanks for spending a few minutes with this week’s Smarter Than Busy. I hope you found something useful, inspiring, or just plain interesting. I’ll be back next week with more insights, tools, and resources to make your work a little smarter (and your week a little lighter).

If this edition helped you, share it with someone who could use it too.

Until Next Week.