6 Ways to Transform Your Business with AI with Google Cloud’s VP & GM of Applied AI

At this point, “We should do something with AI” has become the corporate version of “We should totally grab coffee sometime.”
It sounds ambitious, but unless it turns into a concrete plan, nothing actually changes.

In a recent SaaStr session, Google Cloud’s VP & GM of Applied AI (Duncan Lennox) shared how thousands of companies are already
making AI real, not theoretical. Google Cloud surveyed around 2,500 businesses and found that roughly 61% already have
generative AI in production; of those, 86% report at least a 6% annual revenue lift, and 74% see ROI within the first
year of adoption. That’s not “someday” impactthat’s P&L-level change on today’s dashboards.

So how do you move from AI slides in your board deck to AI systems that actually drive revenue, cut costs, and delight customers?
Let’s walk through six practical ways to transform your business with AI, inspired by Google Cloud’s applied AI approach,
plus real-world examples of what’s working right now.

Why “Applied AI” Matters More Than “Cool AI”

Google Cloud talks a lot about applied AIwhich is code for “AI that your CFO can see on a spreadsheet.”
Instead of focusing on standalone experimental chatbots, applied AI is about weaving models into your existing products,
workflows, and data so they quietly move key metrics: revenue per user, ticket resolution time, churn, margin, and more.

Google’s own guidance emphasizes starting from business value: choosing use cases by impact and feasibility rather than hype,
and aligning AI initiatives with measurable outcomes and owners.
At the same time, new infrastructure like Google Cloud’s AI Hypercomputer platform, with custom Axion CPUs and seventh-generation
TPUs, is making it easier to train and deploy massive models at scale, while keeping performance high and costs under control.

In short: the technology is ready. The question is whether your strategy is.

1. Start with Business-Backed Use Cases, Not Shiny Demos

Lead with ROI, not novelty

The fastest way to waste money on AI is to start with “What can this model do?” instead of “Which business problem hurts the most?”
Google Cloud recommends a business value–driven approach to selecting generative and traditional AI use cases:
map problems, estimate impact (revenue, cost, risk), and score feasibility (data readiness, integration complexity,
regulatory constraints).

Common high-ROI starting points include:

• Customer support: AI agents for tier-1 issues, auto-drafted responses, and knowledge search.
• Sales and marketing: automated personalization, email and ad copy generation, lead scoring.
• Operations: document processing (invoices, contracts), routing, anomaly detection.
• Product: AI-assisted features inside your app (recommendations, summaries, insights).

Google’s own examples show companies prioritizing use cases where AI can cut manual effort by 30–70% or unlock entirely
new experiences customers will pay for.

Design small wins with big visibility

Instead of a giant multi-year “AI transformation” project, start with one well-defined use case with:

• A single, accountable business owner.
• Clear success metrics (for example: “Reduce support handle time by 25% in six months”).
• A tight feedback loop between end users and the AI team.

That first success becomes your internal case study. It’s much easier to get budget for the next five AI initiatives once
you’ve shown the first one paid for itself in under a yearwhich is exactly what three-quarters of Google Cloud AI customers
are seeing.

2. Build a Modern Data Foundation (Before You Stack AI on Top)

Here’s the tough-love part: if your data is scattered across 14 systems, half of it is duplicated, and nobody agrees on what
“active customer” means, AI will politely expose that chaos in record time.

Leaders at Google Cloud repeatedly highlight that the real bottleneck isn’t the modelit’s data readiness:
how well you’ve integrated, governed, and secured the information you want AI to reason over.

Make data reliable and AI-ready

Practical steps for an AI-ready data foundation include:

• Centralize critical data in a cloud data warehouse or lakehouse, so AI tools can access consistent,
  governed information instead of fragile spreadsheets and exports.
• Clean and label your dataespecially for high-value domains like customers, revenue, and products.
• Set clear access controls to keep sensitive data protected while still making it usable for AI.

When companies do this well, the payoff is big: Google Cloud reports organizations seeing strong ROI from AI once they pair
powerful models with high-quality, well-governed data.

3. Put AI in the Apps Your Teams Already Use

Not every employee wants to live inside a developer console or a standalone chatbot tab. In fact, many don’t want “AI” at all
they want less manual work inside the tools they already use every day.

That’s why Google is weaving AI into WorkspaceGmail, Docs, Sheets, Meet, Chat, and moreso knowledge workers can brainstorm,
summarize, analyze, and draft without leaving their existing workflows.

Copilots, not robots

The most successful deployments treat AI as a copilot, not a replacement:

• Sales reps get email drafts they can tweak, not send blindly.
• Analysts get starting points for models and dashboards they still verify.
• Managers get meeting summaries and action lists they can refine.

New tools like Workspace Studio let non-technical teams create AI agents and automations through natural language, then share
them across the company, dramatically reducing the need for custom script-writing.

The end result: more people benefit from AI, fewer tickets land on IT’s plate, and AI stops being a “special project” and
becomes part of everyday work.

4. Reinvent Customer Experience with AI Agents

Your customers don’t care what model you’re using. They care whether they can get the right answer in 30 seconds at 11:47 p.m.
from their couch while holding a slice of pizza.

Consumer brands are already using Google Cloud AI to personalize experiences and automate support in ways that simply weren’t
possible a few years ago. For example, Papa John’s is working with Google Cloud to use AI for personalized offers, smarter
messaging, chat-based ordering, and even optimization of delivery routes and store operations.

How AI agents transform CX

Applied AI can help you:

• Deflect routine requests with conversational agents that are grounded in your knowledge base and policies.
• Personalize journeysthink tailored recommendations and offers based on behavior, history, and context.
• Support human agents with suggested replies, summaries of prior interactions, and instant knowledge search.

Google highlights customer service as one of the clearest generative AI use cases, with companies reducing support costs while
improving customer satisfaction scores at the same time.

5. Differentiate Your Products with Generative AI Features

It’s no longer enough to be “AI-powered” in your marketing copy. Customers are starting to expect concrete, useful AI features
inside the products they buyespecially in SaaS.

Google Cloud showcases hundreds of real-world generative AI use cases where companies have built entirely new product
capabilities: automated content generation, advanced search and summarization, AI-driven analytics, and more.

Examples of product-level AI differentiation

• Analytics platforms that let users “chat with their data” instead of writing complex queries.
• Learning apps that generate personalized practice sessions based on what users struggle with.
• Vertical SaaS tools that draft contracts, proposals, or compliance documents tailored to industry rules.

Behind the scenes, Google’s AI infrastructureincluding custom chips like Axion CPUs and Ironwood TPUsis built to handle
training and inference for these high-value product features efficiently.

The lesson from the SaaStr ecosystem: the winners aren’t just using AI in their slidesthey’re shipping AI features their
customers can’t imagine living without.

6. Invest in Skills, Governance, and Responsible AI

You can’t transform your business with AI if only three people in the org understand how it worksand one of them is an intern
on a six-month contract.

Google and others are heavily investing in AI education for business leaders and technical teams, from foundational training
on generative AI and prompt engineering to advanced leadership courses on how to weave AI into strategy and operations.

Build skills and guardrails at the same time

A mature applied AI program includes:

• Training for both builders (developers, data scientists) and consumers (business users, managers).
• Governance around what data can be used, where AI is allowed in workflows, and how outputs are reviewed.
• Responsible AI policies to manage bias, security, and compliance, especially in regulated industries.

Companies that get this right turn AI from a risky experiment into a reliable capabilityand they’re usually the ones
capturing outsized value while competitors are still stuck in pilot purgatory.

How Google Cloud’s Ecosystem Accelerates AI Transformation

AI transformation isn’t just about models; it’s about the ecosystem around them. Google Cloud is expanding that ecosystem
fastfrom infrastructure and partnerships to industry-specific solutions.

Recent developments include:

• Partnerships with other cloud providers so enterprises can access Google’s Gemini models even when their
  core workloads sit elsewhere, such as Oracle’s cloud platforms.
• Specialized quantitative AI models from partners like SandboxAQ, aimed at heavy numerical workloads such as
  financial modeling and scientific computing, now available via Google Cloud.
• Custom silicon and AI Hypercomputer infrastructure designed to deliver massive performance for training and
  inference in a single integrated platform.

All of this is in service of applied AI: making it practical to embed intelligent capabilities into business processes without
every company having to invent its own AI research lab.

Experiences from the Field: What Actually Works

Turning the six principles above into reality looks different in every organization, but the patterns are surprisingly
consistent. Let’s look at how three types of companies typically experience AI transformation with Google Cloud–style applied AI.

1. The digital-native SaaS company

A mid-stage SaaS startup selling a workflow platform might start with a clear, measurable goal: reduce churn by 10% and lift
expansion revenue by 5% within a year. Instead of building a standalone chatbot, they embed AI directly into the product:

• New customers get AI-generated onboarding checklists tailored to their industry and use case.
• Existing users see in-app suggestions for underused features based on their behavior patterns.
• Account managers get auto-generated health summaries before renewal calls.

Because the startup already has a relatively clean, cloud-based data stack, connecting product usage data, CRM signals, and
support history to Google Cloud AI tools is straightforward. Within months, they start to see lower support volume per customer,
higher feature adoption, and stronger renewal ratesall traceable back to specific AI-powered interventions.

The internal mood shifts from “AI is a risk” to “AI is how we keep winning deals.” The board starts asking, “What’s the next
AI feature we can ship?” rather than “Why are we spending so much on experimentation?”

2. The established retailer or restaurant chain

A large consumer brandthink retail, quick-service restaurants, or hospitalityoften begins with a fragmented tech stack and
deep operational complexity. Their AI journey usually starts in two places:

• Personalized marketing and loyalty: better segmentation, smarter offers, improved targeting.
• Operational efficiency: inventory forecasting, labor planning, and demand prediction.

By centralizing data in the cloud and using AI to analyze customer behavior, they can move from mass promotions to tailored
experiences: different notifications, offers, and content based on each customer’s history and preferences. At the same time,
AI models help them predict demand more accurately by store, time, and product, cutting down on waste and stockouts.

A pattern emerges: once leadership sees that AI-driven personalization can lift conversion and average order value, they gain
the confidence to experiment with bolder ideas like voice-based ordering, AI-powered menu recommendations, or chat-based
customer service. The story is similar to what we’re seeing with brands partnering with Google Cloud to modernize ordering
and loyalty experiences.

3. The mid-market enterprise with legacy systems

For a mid-market manufacturer, insurer, or logistics firm, the biggest barrier is often the legacy systems that nobody wants
to touchbut everyone relies on. These companies usually can’t “rip and replace,” so they lean on applied AI that wraps around
existing systems.

Common starting points include:

• Document-heavy workflowsclaims processing, contracts, invoiceswhere AI can extract key fields and route
  cases automatically.
• Knowledge search across decades of PDFs, manuals, and policies, turning static documents into a searchable,
  conversational knowledge base.
• Predictive maintenance where AI models use sensor and operational data to signal when equipment needs
  attention before it fails.

These organizations often see early wins not because they deploy the flashiest AI, but because they target mundane, labor-heavy
tasks where even a 20–30% improvement translates into millions in savings and happier employees. Over time, as trust in AI grows
and the data foundation matures, they expand into more strategic uses like pricing optimization or scenario planning.

Across all three examples, the pattern is the same: start small and valuable, build on a strong data and governance foundation,
put AI where people already work, and keep the focus on measurable business outcomes.

Conclusion: Turning AI Hype into Applied ROI

AI doesn’t transform your business just because you’ve signed a cloud contract or added “AI” to your product page.
Transformation happens when:

• You choose high-impact use cases tied to revenue, cost, or risk.
• Your data is trustworthy enough for AI to reason over.
• AI flows into the tools your teams and customers already use.
• You invest in people, processes, and governancenot just models.

The data from Google Cloud and the insights shared at SaaStr are clear: companies that embrace applied AI are not just
shaving minutes off workflowsthey’re seeing meaningful revenue growth and fast payback periods, often in less than a year.

If you treat AI as a side project, you’ll get side-project results. If you treat it as a core capabilitygrounded in business
value, powered by a modern data stack, and woven into everyday toolsyou’ll be much closer to the kind of transformation
Google Cloud’s VP & GM of Applied AI describes: AI not as a buzzword, but as a quiet engine behind your next stage of growth.