Why digital marketing teams should work on defining AI foundations now

We’re in a period of transition and AI is being introduced and adopted across every user journey and area of the business. A new era of agentic AI is also emerging. It’s certainly a learning journey – an exhilarating but occasionally very bumpy journey that is still rapidly evolving including the products that you use.

There are lots of way you can use AI in digital marketing and website management – it is changing and will change how we manage our digital channels. There are good sides and bad sides to using AI on your website and different teams are at different stages of this journey.

Effective AI is resolutely not plug and play, even though it might feel like it is when you instantly ask ChatGPT a question or ask it to generate a blog post. If you want to get the best out of AI for digital marketing, take advantage of agentic capabilities emerging within CMS platforms, and reduce any associated risks, then it is critical to build the right foundations, even if you are not using AI now.  

These “foundations” span several different areas:

• Alignment: Ensuring digital marketing is aligned with your overall company’s approach to AI, which in turn is aligned to your company strategy and values.
• Ownership: Ensuring there is the right ownership of using AI for marketing to manage risks and make clear decisions.
• Awareness and confidence: Ensuring there is the right awareness of areas such as the product landscape to make informed decisions and also establish confidence to use the tools.
• Core reference material: Defining brand rules and reference material in place for the AI to generate content that is optimised and on brand and start to use agents effectively.
• Prompts, agents and skills:  Defining the detail behind the AI within AI platforms itself, even if you don’t have the detail locked down now.

Let’s look explore each of these five foundational areas.

1. Alignment

Does your company have an overall AI strategy or defined approach to AI? You may not or it might be already out of date given the current pace of evolution.
In our view it is important for leadership teams to set out a strategy, agenda or even some kind of manifesto that helps to establish both the spirit and limit of AI deployment and innovation and provides a guide across all company initiatives. This can then act as an overall guide that sets the tone around the level of ambition and ensures an ethical approach. 
Every digital marketing team has the opportunity to use AI now, so it is important to ensure there is alignment with your broader company strategy. This is critical as the AI-driven output on your external channels is visible to the outside world.

2. Ownership

AI happens within organisations across multiple processes and different products. Employees are also likely to use unofficial AI solutions for work purposes. This means in practice there is no overall owner of AI within any organisation.
While this distributed pattern of ownership encourages innovation, experimentation and learning, it also contributes to risks, increased costs and scattergun approaches.  However, ownership and responsibility for different aspects of AI, or the use of AI in particular disciplines can be defined.

When it comes to using AI for digital marketing and website management, it helps to define the people who will make decisions and be ultimately responsible for it. This will almost certainly be a mix of marketers and technologists. Your IT function, for example, will be responsible for any security aspects of using AI. Your marketing team may be the de facto owners of your content and digital marketing tools including your CMS.

AI is being used now. Take time to formalise the roles relating to AI decisions to ensure you progress in the best way possible. Using a RASCI matrix can be very helpful here and roles and responsibilities should also extend to your digital agency.

3. Awareness and confidence

It is challenging to keep up with what is happening with AI right now. There is a lot happening at great pace and once you’ve got your head around something, it may already be out of date. It is impossible to keep up with all the details, but having mechanisms for general awareness about AI, what it can do, and the direction of travel can help the right teams to make the right decisions at the right time. Being confident in using AI tools is also important – something which usually comes with a little exposure and experimentation.

Areas of AI to keep an eye on include: 

• product updates, especially across your digital marketing stack
• the effective use of AI prompts and skills to support general confidence
• any technical and security updates
• developments with agents and how these are being deployed
• integrations and the potential to integrate AI (for example using the MCP protocol)
• general trends and the overall direction of travel.

4. Core brand reference material

One of the challenges of using AI and AI agents for content generation and optimisation is that the output can be off-brand and generic. Recognising the need for governance and context, platforms like Sitecore and Optimizely have built in places to establish brand guidelines, brand reference material and related rules that can help define the output of any content that is generated. 
This helps provide some brand consistency to AI-generated content, reducing risk and also speeding up the creation of the content, as well as establishing trust in using the AI agents. Sitecore recently published a good blog post about the importance of building this context for AI which is worth reading. Other LLM services like Claude and ChatGPT also have the potential to establish reference material that can also help define brand rules.

At its core, brand reference material to guide AI output and agent actions should cover areas such as:

• Core brand message and positioning covering elements such as company mission and purpose.
• Key facts about the business that are frequently referenced.
• Tone of Voice (TOV) guidelines.
• Specific content rules, for example treatment of capitalisation and numbers.
• Any audience and persona specifications and related guidance.
• Any rules around language and localising content.
• Any taxonomy and metadata rules.
• Any prohibited phrases or content.
• Image, design and formatting rules.
• Any rules relating to data privacy and / or regulatory compliance.

While some digital marketing and brand teams have these covered, in practice they are not always complete or current and are not actively maintained. The AI era gives more impetus to define and update these kinds of documents.

5. Prompts, skills and agents 

Having the general rules brand and content is important, but then you can go even further by having specific approved prompts and / or skills within the AI, or setting up agents within the CMS, which are more specific and might help drive more granular rules around specific content formats, workflows, and more. While the reference content provides the context, this layer of governance is more defined within the AI or AI-powered CMS itself.

This is really about defining the detail around formatting. layout, targeting different content to particular personas and so on. If you’re not using your AI at the moment but plan to do so in the near future, then consider starting to think about the specifics.

Setting the foundations

AI is here to stay and is impacting digital marketing and website management. Getting the foundations in place now will help put you in a position of strength to get the best out of AI.

If you’d like to discuss using AI for your website or digital marketing channels, get in touch!