AI in Marketing Takeaways for Leaders: October 2026

30 Oct

AI adoption is accelerating. Is your business keeping up strategically?

One thing I'm increasingly noticing through conversations and projects is how differently AI is being adopted across businesses, and sometimes even within the same organisation.

A marketing team might be using ChatGPT or Gemini extensively for research, content and campaign planning. Someone in finance may have developed their own approach to reporting and analysis. Operations could be experimenting with automation, while other employees are barely using AI at all.

There's a lot to be encouraged by. People are finding practical ways to improve how they work, often without waiting for formal direction.

But it raises some important questions for leadership. Do you know how AI is being used across your organisation? Which activities are creating genuine value? Are teams sharing what works? Are there appropriate controls around company and customer data? And is any of this connected to a wider business strategy?

October's developments make these questions increasingly relevant.

Google, OpenAI, Microsoft and others are bringing AI further into the everyday software and processes businesses rely on. AI agents are becoming more capable of completing work across multiple systems. Advertising platforms are using increasingly sophisticated intelligence to determine which brands and products reach customers. Meanwhile, the number of available tools, models and applications continues to grow.

For marketing and business leaders, the opportunity is considerable. Better customer experiences, more efficient operations, stronger decision-making and new opportunities for growth are all within reach.

The challenge is identifying where to focus, how to bring people and processes together, and how to turn growing AI capability into measurable commercial progress.

With 2027 planning approaching, it feels like a particularly good time to consider that bigger picture.

Here are the AI developments leaders should be aware of this month.

1. AI discovery and advertising are evolving together

October brought further evidence that AI is changing how customers discover, research and evaluate businesses.

Brainlabs published its State of AI Visibility 2026 research, examining AI discovery and referral performance across a substantial sample of client accounts.

Across 54 accounts representing more than 250 million sessions, AI referral traffic increased by 163.5% between the periods examined, while traditional organic sessions declined by 10.5%.

Those figures need context. AI referral traffic was growing from a much smaller base, and the findings represent Brainlabs' client sample rather than the market as a whole. Nevertheless, they illustrate how customer discovery is becoming more distributed.

The research also highlighted the differences between AI platforms. ChatGPT accounted for more than 90% of AI referral sessions in the sample, while other platforms displayed different patterns of traffic growth and conversion performance.

At the same time, advertising opportunities are expanding.

On 5 October, OpenAI announced new visual advertising formats within ChatGPT, alongside expanded measurement partnerships and brand suitability capabilities. Later in the month, it expanded ChatGPT Ads to another 26 countries, bringing availability to 89 markets worldwide.

Google followed on 28 October, announcing the expansion of advertising in AI Mode to ten additional countries, with the UK among the markets planned for further rollout during Q4. Google is also testing advertisements within follow-up AI conversations in the US.

Together, these announcements reflect an important change. Customers can increasingly ask detailed questions, explore alternatives and receive relevant commercial suggestions within the same conversational experience.

The relationship between information, recommendations, advertising and purchasing decisions is becoming more closely connected.

Leadership takeaway:

Marketing leaders should be considering AI discovery as part of their wider visibility, brand and acquisition strategy.
That includes understanding where customers are researching, how AI platforms describe and recommend the business, which competitors appear, and whether these interactions generate valuable commercial outcomes.
It also means ensuring product information, service propositions, customer evidence and brand content are accurate and accessible across the environments AI systems use.
AI advertising deserves measured experimentation, but investment should follow evidence of commercial value. The same principles of incrementality, acquisition cost and customer quality still apply, even when the advertising environment looks very different.

2. Meta's advertising intelligence is changing what marketers need to control

Meta provided a series of updates during October describing significant developments in its advertising technology. These included its Generative Ads Model (GEM), Generative Recommenders (GenRecs) and the next generation of its Andromeda advertising retrieval system.

Although the underlying technology is complex, the commercial direction is relatively straightforward. Meta is improving its ability to understand advertising content, products, customer interests and likely outcomes, using increasingly sophisticated AI models to determine which advertisements should be shown to which people.

Its GenRecs research describes systems capable of understanding similarities between products and customer preferences, including situations where a new product has limited historical engagement data. Andromeda is designed to identify relevant advertisements from an enormous pool of potential candidates at much greater speed and scale.

As these systems become more capable, advertising platforms can take greater responsibility for decisions previously influenced through manual targeting, segmentation and campaign configuration. This has implications for the skills and priorities within marketing teams.

Audience understanding, proposition, creative quality, product information, measurement and commercial objectives become increasingly important inputs. A platform can become very good at predicting who might respond to an advertisement. It still depends on the business providing compelling reasons to choose its products or services, and understanding whether those responses lead to profitable customers.

Leadership takeaway:

Review how your marketing team spends its time as AI takes on more execution and optimisation.
There may be opportunities to reduce manual campaign management and redirect effort towards customer insight, creative testing, proposition development and performance analysis.
Clear commercial objectives and trustworthy measurement are particularly important when the platform is making a growing proportion of delivery decisions.
The opportunity is to make marketing teams more effective, with skilled people concentrating on the decisions that have the greatest influence on performance.

3. AI is becoming a connected layer across business systems

One of October's key themes was the continuing shift towards AI working across existing business applications.

Google Cloud introduced its Gemini agent on 8 October, describing a system designed to work with organisational knowledge, use connected tools and complete tasks across documents, email and business systems.

Microsoft announced further integration between Copilot, Dynamics 365 and Teams, including 30 new CRM skills designed to bring customer information and actions into everyday workflows.

OpenAI also expanded its partnership with ServiceNow on 28 October, connecting AI capabilities with operational workflows covering areas such as IT service requests, software development and security.

These are different platforms and approaches, but they point towards a similar development. AI is becoming increasingly capable of operating across the applications people already use, bringing together information and completing work that previously involved moving between several systems or departments.

Consider a relatively straightforward customer enquiry. An employee might need to find the customer record, check previous correspondence, review an order or booking, consult an internal knowledge base, prepare a response and update the relevant system.

Connected AI could potentially support several of those activities within one workflow. The same principle applies to internal reporting, project management, purchasing, finance and administration. However, the effectiveness of these systems will depend heavily on the quality of the underlying information, the reliability of integrations and the business processes they are designed to support.

Disconnected data, inconsistent procedures and unclear ownership can become more problematic when AI starts working across departmental boundaries.

Leadership takeaway:

Start identifying workflows that involve repetitive administration, multiple systems or unnecessary handoffs between people and teams.
Consider where AI could improve efficiency, accuracy, responsiveness or the quality of decisions.
Before introducing further tools, review what is already available within your existing technology environment, particularly Google Workspace, Microsoft 365, CRM systems and other core platforms.
The best opportunities may come from connecting existing capabilities and improving the underlying process, rather than adding another standalone AI application.

4. AI agents are moving into sales, customer service and commercial operations

October also provided some practical examples of AI moving beyond internal productivity into customer-facing and revenue-generating activities.

Microsoft introduced further capabilities for its Sales Development Agent and Service Agent, designed to support activities including prospect research, customer communication, sales follow-up and service case management.

xAI shared how its own sales operation is using shared Grok Bots for account research, quote preparation and deal review.

The interesting aspect of that example is its approach to shared knowledge and processes. Rather than individual employees building separate AI assistants, the organisation has developed specialised agents that can be used across teams, with common instructions, information and controls.

Salesforce also announced an interesting UK example through Tottenham Hotspur. The club has connected more than 4.6 million fan accounts across a wider digital ecosystem, supporting a new AI assistant called Ask Spurs. According to Salesforce, the assistant has achieved an 86% engagement rate, with 20% of interactions resolved by Agentforce alone and other enquiries able to move to human support.

The reported results are encouraging, although engagement and resolution are different measures, and the case study is still an early example rather than proof of wider industry performance.

What makes the example particularly useful is the connection between customer information, different business systems and the actual service experience. For organisations operating across multiple customer touchpoints, such as hospitality, leisure, retail, property or professional services, similar opportunities are worth exploring.

An AI assistant could potentially answer questions about availability, support a booking journey, help customers navigate complex services or provide more consistent follow-up after an enquiry.

Leadership takeaway:

Look beyond the chatbot itself and assess the entire customer or commercial process.
Where are enquiries delayed? Which routine requests consume disproportionate employee time? Where are customers repeating information? Are sales opportunities being followed up consistently?
AI can help address these challenges, but reliable customer information, clear escalation routes and human oversight remain essential.
Success should be measured through outcomes such as customer satisfaction, resolution quality, conversion, retention and time saved, rather than simply the number of interactions handled by AI.

5. Business-wide AI adoption is accelerating, but capability needs to develop alongside it

Another area we wanted to explore based on updates in October was how AI is actually being adopted within organisations.

OpenAI published research on AI use among European businesses, reporting that weekly active users of agentic tools among its EU small and medium-sized business customers had increased 8.1 times since May. This reflects activity within OpenAI's customer base, so it should not be interpreted as a measure of adoption across all European businesses. However, it demonstrates how quickly usage can grow once organisations start moving beyond occasional experimentation.

Uber also announced the company-wide rollout of ChatGPT Work to more than 30,000 employees, following an initial pilot involving 1,000 people. The rollout is supported by approved integrations, an internal library of AI skills and employee training across functions including operations, finance, marketing and product development.

Providing access to AI tools is relatively straightforward. Helping an organisation use them consistently and effectively is a much broader challenge. Within one business, employees may already have very different levels of experience. Some will be developing sophisticated workflows, building their own assistants and finding ways to reduce repetitive work. Others may lack the confidence or knowledge to use basic AI capabilities.

There can also be significant differences between departments. Marketing teams often have immediate opportunities for content, research, analysis and creative work. Finance, operations, HR and customer service may have equally valuable applications, but different data requirements, risks and practical barriers.

Without coordination, knowledge can remain isolated within individual teams. Successful experiments may never be shared, and different departments can end up purchasing similar tools or developing separate solutions to the same problem.

Leadership takeaway:

Treat AI capability as an organisational development priority.
Understand how confidently different teams are using AI, identify employees who are already developing valuable applications and create opportunities for that knowledge to be shared.
Training should be relevant to people's actual responsibilities, supported by practical examples and clear expectations.
It is also worth involving employees in identifying opportunities. The people closest to a process often understand its inefficiencies better than anyone.
A coordinated approach can help businesses build capability more evenly while retaining the initiative and experimentation that make AI adoption valuable in the first place.

6. More capable, affordable AI is making platform choice and spending harder to manage

Updates this month also reinforced how quickly the economics of AI are changing. Anthropic introduced Claude Haiku 5.5 on 7 October, designed for faster, high-volume and cost-sensitive tasks. The company reported average running costs around 75% lower than the previous Haiku generation. This is encouraging for businesses considering larger-scale AI applications, particularly where a relatively straightforward task doesn't require the most expensive model.

But lower model costs do not necessarily mean lower overall AI expenditure. An organisation may now be paying for AI within Microsoft 365, Google Workspace, CRM systems, creative platforms, analytics tools and specialist applications, alongside individual subscriptions and usage-based services. As more workflows become automated, the volume of AI activity can increase substantially.

HubSpot's introduction of new AI spending controls in October is a useful illustration of how software providers are responding to this challenge.

Other developments are also making platform selection more interesting. Anthropic contributed its Agent Skills standard to an industry foundation, supporting a more interoperable approach to giving AI systems specialised knowledge and workflows.

Perplexity introduced a fully local version of its Computer agent for supported Macs, providing another option for organisations considering how sensitive information should be handled.

These announcements illustrate that businesses have more choices around capability, cost, integrations, portability and data handling. There is unlikely to be one ideal platform for every task or organisation.

Leadership takeaway:

Develop a clearer picture of your existing AI technology and spending.
Understand which platforms employees are using, where capabilities overlap, what the full costs are and which tools are delivering measurable benefits.
When selecting solutions, consider:
- Suitability for the business task
- Integration with existing systems
- Security, permissions and data handling
- Ease of use and employee adoption
- Running costs and commercial value
- Flexibility as technology changes

For many businesses, consolidating existing subscriptions and making better use of capabilities they already have could deliver immediate improvements.

7. Creative AI is advancing, making quality and content verification increasingly important

AI-generated creative content continues to improve in quality and accessibility. During October, Canva announced Craft, its new image-generation model, followed by Brush, designed to understand existing designs and make requested edits across creative assets.

Adobe also expanded its collaboration with NVIDIA, focused on improving AI-powered creation, editing, production and personalisation across products including Photoshop, Premiere and Frame.io.

For marketing teams, these developments offer practical opportunities to reduce production time, explore creative ideas and adapt materials for different channels and audiences. But as content becomes easier to produce, reviewing and managing it becomes increasingly important.

A business can now generate considerably more copy, imagery, presentations and campaign variations with relatively limited resource. Without clear creative direction and appropriate quality controls, that can also lead to inconsistent messaging, inaccurate claims or a growing volume of fairly undifferentiated content.

Content authenticity is another developing consideration. Google expanded access to its SynthID Detector globally in English during October, making it easier to identify supported AI-generated imagery, video and audio from Google and participating partners.

It is an important development, although the tool cannot identify every piece of AI-generated content. The wider direction is clear: businesses increasingly need to understand how their content was created, where AI has been used and what standards should apply before material is published.

Leadership takeaway:

Review AI as part of your complete creative and content process.
That includes strategy, briefing, production, brand consistency, factual accuracy, intellectual property considerations, consent and final approval.
There is considerable value in making production more efficient, particularly where teams are under pressure to create more content across more channels.
The strongest results will still depend on the quality of the thinking, customer insight and creative judgement behind the work.

8. AI governance is becoming part of everyday operational management

As AI becomes more connected and capable of taking action, the importance of governance continues to grow.

Amazon Web Services published guidance during October on evaluating AI agents in production, highlighting the need to monitor reliability, accuracy and performance beyond initial testing. This is an important consideration because AI systems can produce convincing results while making mistakes, particularly when completing more complex tasks across several applications.

It is also worth distinguishing between the different levels of responsibility businesses give AI. An employee asking AI to summarise an internal document creates a different risk profile from an agent capable of changing customer records, sending communications, adjusting advertising budgets or updating financial information. The latter requires much greater consideration of access, approval and accountability.

There are also wider questions around the information employees are sharing with AI platforms, particularly where personal, commercially sensitive or confidential customer data is involved.

For UK businesses, existing data protection, confidentiality and contractual obligations remain relevant regardless of whether AI is involved. A business may already have sensible IT security policies, but those arrangements will not necessarily cover every new AI application or workflow employees have introduced.

Leadership takeaway:

Establish a proportionate approach to AI governance that reflects the activities being undertaken.
At a minimum, businesses should understand:
- Which AI tools and agents are being used
- What information they can access
- Which actions they are permitted to take
- Where human approval is required
- How outputs and performance are checked
- Who is responsible for the process and its outcomes
Low-risk experimentation should remain straightforward. More sensitive or commercially significant activities need stronger oversight.
As usage develops, these controls should become part of normal business management rather than something addressed only when an issue arises.

Leadership Perspective: Does your business have an AI strategy for 2027?

Looking across October's developments, I think there is a broader leadership challenge emerging. AI capabilities are becoming more sophisticated, accessible and commercially relevant. But the pace of development makes it difficult for individual organisations to understand which opportunities deserve attention and how everything should fit together.

I've started working on a couple of projects that look at exactly this challenge, both within marketing specifically and across wider business operations. One observation is just how varied the starting point can be.

Businesses may have employees who are already highly capable AI users, alongside colleagues who would benefit from relatively basic training and support. Different departments may be using separate tools, developing different approaches or solving similar problems independently.

And while plenty of useful activity is happening, it isn't always easy for leadership to understand the combined value, cost or implications. That's understandable given how quickly the technology is evolving.

However, with AI now influencing marketing, customer experience, commercial processes and operational efficiency, there is a strong case for taking a more coordinated approach.

For many businesses, I think four areas deserve attention as part of their 2027 planning.

1. Understand where you are today

Start by reviewing how AI is currently being used across different teams.

Which tools are available? How frequently are they used? What are employees actually doing with them? Which applications are proving useful, and where are there gaps in capability, confidence or governance?

This creates a practical baseline rather than relying on assumptions about organisational adoption.

2. Identify where AI could create meaningful value

Look at the business priorities for the coming year and where AI could help achieve them. Within marketing, opportunities might include research, analysis, content production, campaign planning, reporting and customer insight.

Across operations, there may be opportunities in administration, customer service, document processing, internal knowledge, workflow automation, forecasting or communication.

Some opportunities will be relatively simple. Others may require significant changes to systems, processes or responsibilities. The important consideration is the business outcome being improved.

3. Prioritise opportunities and address the foundations

Not every application will justify investment. Assess potential opportunities against expected commercial impact, implementation complexity, cost, employee readiness, data availability and risk.

A modest improvement to a process used hundreds of times each week may deliver considerably more value than a technically impressive application with limited practical use.

There may also be foundational improvements required, including training, data quality, permissions, platform configuration or clearer ownership of existing processes.

4. Build a practical roadmap for 2027

Agree a manageable set of priorities.

Identify immediate improvements, select worthwhile pilot projects and establish who will be responsible for delivery.

Define how success will be measured, what support employees need and how the organisation will review progress.

Given the pace of change, the plan should be flexible enough to evolve as new capabilities become available.

A sensible AI strategy should provide clear direction without becoming an unnecessarily complicated exercise.

For founders, marketing directors and leadership teams, there's a valuable opportunity here to make AI adoption more purposeful, improve collaboration across departments and focus investment where it can make a genuine difference.

Leader's Checklist

A few questions worth discussing with your leadership team as you plan for 2027:

Visibility: Do we know which AI tools are being used across the organisation, and for what purposes?
Value: Can we identify which activities are improving efficiency, performance, customer experience or commercial outcomes?
Capability: Are there significant differences in AI confidence and competence between teams?
Alignment: Are departments sharing knowledge and working towards common objectives?
Security: Do we have appropriate policies, permissions and controls around AI use and company information?
Investment: Are we managing AI tools and spending as a coordinated part of our technology and business plans?
Direction: Do we have a clear set of AI priorities, responsibilities and measures of success for 2027?

You don't necessarily need an answer to every question immediately. But they provide a useful starting point for understanding how effectively AI is being adopted.

Final Thought

October has demonstrated just how quickly AI is becoming embedded in the systems, processes and decisions that underpin modern businesses. The technology is creating opportunities that would have been difficult or expensive to pursue even a relatively short time ago.

Marketing teams can reach customers in new ways. Employees can access information and specialist capabilities more easily. Operational processes can become more efficient. Customers can receive quicker, more relevant support. There is good reason for businesses to be ambitious about what AI could help them achieve. But ambition benefits from direction.

As we approach 2027, I'd encourage business leaders to look beyond individual tools and experiments and consider how AI fits within their wider commercial objectives, people, processes and technology.

Understanding where you stand today and agreeing where to focus next could be one of the most valuable planning exercises you undertake for the year ahead.

Considering your AI strategy for 2027?

At Saurce, we're supporting organisations with AI Strategy & Opportunity Reviews, helping leadership teams understand current AI adoption, identify opportunities across marketing and wider operations, and develop practical plans for making better use of the technology.

If you're considering how to approach AI more strategically, we're offering a free 30-minute AI strategy consultation to discuss your current position, challenges and opportunities.

Book a free 30-minute AI strategy consultation with Saurce

Useful Links & Further Reading

AI discovery, advertising and marketing performance