When ChatGPT launched at the end of 2022, we all scrambled to master prompt engineering. Our LinkedIn feeds were filled with ultimate prompt guides and secret prompt formulas for coaxing better outputs from our new AI partners.

Two short years later, AI platforms are evolving beyond the simple prompt-and-response format. The big shift isn't about AI getting smarter, although it has. It's about AI getting more focused. Instead of engineering better prompts, the focus is on designing better contexts.

The same AI chats can now take place in a customised environment where you build in the knowledge needed to consistently deliver more meaningful results. Tools like Google's NotebookLM, Anthropic's Claude Projects, OpenAI's Custom GPTs, and Perplexity's Spaces learn about your specific world.

It's yet another game-changer in how businesses work with AI that promises to transform how organisations manage and leverage their knowledge.

Why Context Design Matters

While prompt engineering focuses on getting the right output from each interaction, context design creates an AI environment that truly understands your organisation's needs, constraints, and objectives.

Think of context design like setting up a new employee for success. Just as you wouldn't expect a new team member to execute a task well without proper onboarding, documentation, and guidance, you can’t expect AI to deliver consistent, high-quality outputs without proper context.

This is crucial for three reasons:

The complexity of business knowledge: Modern organisations don't just need to analyse single documents or create individual pieces of content. They need to synthesise insights across vast amounts of information while maintaining consistency with their brand, strategy, and compliance requirements.

Improved efficiency and consistency: Rather than having team members repeatedly explain basic context each time they prompt, and hoping they get those details right each and every time, custom AI environments offer opportunities for standardisation.

This allows teams to focus on higher-value activities instead of constant AI supervision and correction. They can upload documents from a central repository covering standards like your style guide, buyer personas, or market insights, and focus on the unique aspects of their projects instead.

Enables more sophisticated applications of AI: When AI tools are given a richer context—from brand voice to market positioning to customer insights—they become capable of more nuanced and valuable outputs. For example, rather than creating generic blog suggestions, a well-contextualised AI can suggest topics based on your content gaps, that references your latest research, or addresses a target account’s specific pain points—all while staying true to your brand voice and strategic objectives.

What Context Design Actually Means

Context design shapes how your AI tools understand and support your business. You can’t dump every source of information you have into an AI system and hope for insights to spill out. You need to thoughtfully select and organise the information that matters most.

It starts with asking the right questions:

What knowledge is truly essential for understanding our brand, market, and our point of view on the subject?
What guidelines and parameters will ensure consistent, quality outputs?
How will we maintain and update this context as our needs evolve?

Let’s look at some categories of information you could provide to get more precise outputs tailored to your needs.

Brand Intelligence: Your Story and Voice

At the heart of context design lies your brand intelligence. This goes far beyond basic style guides—it's about capturing the soul of your brand. AI needs to understand your tone of voice, but it’s also helpful to provide examples that perfectly embody it. Think of the marketing pieces that made you proud, the ones that captured your brand's essence exactly right.

Your brand story elements matter just as much. Those pivotal moments that shaped who you are as a company, your origin story, and the key narratives that define you—these all help AI understand your brand at a deeper level. Consider how you translate abstract concepts into visual elements, and don't forget to include those success stories where your brand voice created genuine connections with your audience.

Subject Matter Expertise: Bridging the Knowledge Gap

AI knows a lot, but it is a generalist. Context design means you can equip AI with the most current and accurate information, making it more capable of emulating the subject matter expertise that makes your brand authoritative in your space.

Current Industry Knowledge

AI models have knowledge cutoff dates that can leave them months or even years behind current developments. Bridge this gap by providing:

Recent industry developments and breakthroughs
Updated technical specifications and standards
Current market statistics and trends
Latest regulatory changes and compliance requirements
Emerging technologies and their impact on your sector

Technical Depth

It can help if AI understands the nuances that set your solutions apart. This means providing:

Detailed documentation on your products or services
Common challenges and your approaches to solving them
Integration considerations and technical dependencies
Performance benchmarks and technical comparisons
Technical FAQs and troubleshooting guides

Point of View Development

What truly sets your people apart is their unique perspective on industry challenges and solutions. Help your AI understand:

Your brand's stance on key industry issues
Why you've chosen certain approaches over others
Your vision for where the industry is heading
How you differ from conventional wisdom in your space
The reasoning behind your strategic decisions

Market Context: Understanding Your Landscape

Your AI needs to see the bigger picture of where your brand fits in the market. This means providing rich competitor analysis that goes beyond surface-level comparisons. Help your AI understand not just who your competitors are, but how they position themselves and, crucially, where you differentiate.

Industry trends play a vital role here too. Both the established patterns and emerging shifts in your market can help AI understand the context of your messaging. Include market research that illuminates your sector's dynamics and evolution. And don't forget the regulatory framework—those compliance requirements and industry-specific regulations that keep your content within bounds.

Audience Intelligence: Knowing Your People

Your AI needs to understand your audience as well as you do. This means creating buyer personas that go beyond demographics, exploring the psychographics that drive decision-making. Map out the paths your customers take to purchase, showing how different segments might approach your brand differently.

The voice of your customer is invaluable here. Direct quotes and feedback from actual customers can help AI understand the real language and concerns of your audience. Complement this with detailed pain point analysis—break down the challenges your customers face and how your solutions address them.

Performance Insights: Learning from Experience

Past performance offers crucial lessons for future success. Share your top-performing content with your AI and ask it to analyse what worked and why it worked. Campaign analytics tell important stories about your audience's preferences and behaviours. How do different segments respond to various approaches? What messaging resonates most strongly?

This empirical evidence helps AI understand the practical application of your brand guidelines and audience preferences.

Putting It All Together: Implementation Guidelines

With these foundational elements in place, you need clear guidelines for how to use them. You might create specific parameters for tone calibration—when should content be formal versus conversational? What industry-specific language requirements exist? How should emotional resonance vary across different content types?

Quality controls are also essential. Establish clear fact-checking processes, citation standards, and review triggers for sensitive content. Define your output specifications clearly—what format requirements exist for different content types? What are your standards for calls-to-action and brand voice consistency?

Maintaining the Foundations

Your knowledge foundation isn't static—it needs regular maintenance to stay relevant and effective. Schedule regular reviews of core documentation and updates to market intelligence. As new performance data emerges, refresh your insights accordingly. Most importantly, continue adding new case studies and success stories that demonstrate your evolving brand narrative.

Quality assurance shouldn't be an afterthought. Monitor for any drift from core brand values. Regular audits of AI outputs against brand standards help ensure consistency and effectiveness. Build feedback loops for continuous improvement and document edge cases that require special handling.

Remember, this isn't about constraining AI or replacing your team—it's about empowering them with the knowledge needed to truly represent your brand. Think of it as creating a living, breathing knowledge ecosystem that grows and evolves alongside your business.

By thoughtfully designing this context, you're not just feeding information into a system—you're creating an AI partnership that understands your brand at a fundamental level and can help your people to tell your story effectively across all channels.

A Note on Privacy and Security

It’s important to note that privacy and security considerations deserve your serious attention.

Before implementing any context-aware AI solution, visit the platform’s website and carefully evaluate the privacy and security implications of uploading proprietary information. Every major AI platform, and even subscription tiers within platforms, has distinct policies regarding:

Data Retention: How long your uploaded information remains in their systems
Model Training: Whether your data and conversations can be used to train or improve AI models
Access Controls: Who can view or interact with your uploaded content
Data Sovereignty: Where your information is stored and processed

Do your research before uploading anything that you wouldn’t want out in the public realm, and take time to understand what happens to that data. Organisations should:

Review platform-specific privacy policies and security documentation
Understand data retention and deletion mechanisms
Verify compliance with industry regulations
Establish clear policies for acceptable content uploads
Implement approval processes for the use of sensitive content
Implement access controls and monitoring systems
Document data handling procedures for audit purposes

The New Generation of Context-Aware AI Tools

AI platforms are evolving rapidly to meet different knowledge management needs. While some excel at research and analysis, others create persistent workspaces that maintain context across conversations. Some focus on collaborative environments where teams can build and query shared knowledge bases.

Let's explore the leading platforms in this space and their unique capabilities for different aspects of context design.

NotebookLM - The AI Research Assistant

NotebookLM transforms source documents into interactive knowledge bases, powered by Google's Gemini model.

You can upload up to 50 sources, including up to 500,000 words per source document. The system handles PDFs, Google Docs, website links, and YouTube videos. Unlike other AI tools, NotebookLM maintains precise citations to your source materials, making it easy to verify information.

One standout feature you might already be aware of is the ability to generate natural-sounding audio discussions about your documents, effectively turning written content into podcast-style conversations. Not long after launch, a conversation between two ‘hosts’ went viral for their reactions to finding out they weren’t real.

You can customise these audio overviews to focus on specific topics or areas of expertise, and they'll continue playing while you query sources or ask questions. For example, you could upload several industry reports and have NotebookLM create an audio overview highlighting key trends, focusing on specific aspects that matter to your team.

It’s very easy to do. Prior to publishing this article we gave NotebookLM the entire text along with 1827 Marketing’s own customer persona for context. Here’s the podcast that was generated in just a few minutes:

The tool particularly shines when you need to:

Extract specific information from multiple lengthy documents
Create study guides or learning materials from complex content
Transform written reports into conversational audio summaries
Identify connections across multiple sources

NotebookLM Business is in the pipeline and will offer enhanced capabilities including higher usage limits, expanded customisation options, and the ability to share notebooks with team members.

Claude Projects - The Context-Aware Workspace

Powered by Anthropic’s powerful models, Claude Projects creates dedicated AI workspaces that maintain context across multiple conversations and documents. Available to Pro and Team users, its key advantage is the ability to remember and reference information from previous chats within the same project.

Each project features a 200K token context window (equivalent to a 500-page book) and supports multiple file formats including PDF, DOCX, CSV, TXT, HTML, ODT, RTF, and EPUB. You can also set persistent custom instructions that shape how Claude interacts, ensuring consistent responses aligned with your needs.

A distinctive feature is Artifacts, which provides a dedicated window alongside your conversation for generating and editing content like code snippets, documents, graphics, or diagrams. For team users, there's also a shared project activity feed.

It excels at tasks requiring:

Long-term context maintenance
Consistent style and tone across multiple sessions
In-depth document analysis with follow-up questions
Iterative content development

For instance, you could create a project for your quarterly newsletter, upload your style guide, past editions, performance data, and recent blog posts, and have Claude help develop content while maintaining editorial consistency. You might start with outlining and drafting, refine the tone across pieces, and develop executive summaries - all while keeping your source materials and brand voice consistent.

Perplexity Spaces - The Collaborative Research Hub

Perplexity Spaces combines document management with AI-powered search, allowing teams to build shared knowledge bases that can be queried using both uploaded files and web sources. Its standout feature is the flexibility in how you source answers - you can search your private documents, the web, both simultaneously, or simply engage directly with the AI model.

Pro users can upload up to 50 files per Space (25MB per file), and the system supports common business formats including Excel, PowerPoint, Word, PDF, and CSV. The tool provides auto-generated summaries of uploaded files while maintaining access to the original documents.

You can customise each Space with specific AI model preferences and custom instructions about how the AI should respond. Pro users can invite up to 10 collaborators, while Enterprise Pro users get unlimited collaboration capabilities.

It's best suited for:

Collaborative research projects
Building searchable knowledge bases
Fact-checking against multiple sources
Maintaining project documentation with AI assistance

Consider using it for organising research materials where you need to regularly cross-reference internal documents with current web information.

Custom GPTs - The Specialised AI Creator

Open AI’s Custom GPTs, introduced in late 2023, allow you to create purpose-built versions of ChatGPT without coding knowledge. The key differentiator is the ability to give the AI specific instructions, knowledge, and capabilities that persist across conversations with any user.

Custom GPTs can access the web, access Code Interpreter and Data Analysis, generate images via DALL-E, and connect to external tools through APIs.

They work well for:

Creating focused tools for specific tasks
Building knowledge-based assistants
Automating repetitive workflows
Providing consistent responses to common queries

For example, you could create a GPT that knows your product specifications and can consistently answer technical questions, or one that helps analyse specific types of data.

ChatGPT Canvas - The Interactive Editor

Canvas reimagines how users interact with AI for writing and coding tasks. It provides a dedicated editing workspace alongside the chat interface.

Unlike document-focused tools such as NotebookLM, Canvas focuses on the creation and refinement process. Think of using it when you need to maintain more control over the editing process while leveraging AI assistance.

Canvas allows for direct text editing alongside AI suggestions. You can highlight specific portions of text to ask questions or request edits, and basic markdown formatting is supported. It also includes features for adjusting the reading level, modifying length, and one-click formatting options. The interface maintains version history through an undo button, and a "Show changes" feature helps track modifications.

While you can't currently create a permanent knowledge library, Canvas allows you to upload files to the current conversation for analysis. For example, you might upload a market research report and use Canvas to extract key findings and co-create an executive summary, with the ability to refine and edit the content directly in the workspace.

After its beta period, Canvas will be available to free users and will automatically open when ChatGPT detects a scenario where it could be helpful.

Adobe Acrobat AI Assistant - The Document Specialist

Adobe's AI Assistant is an add-on built specifically for document interaction, bringing conversational AI capabilities to PDF workflows. Unlike general-purpose AI tools, it's designed to understand document structure and maintain the context of complex documents.

Key features include:

Intelligent Citations: A custom attribution engine helps verify the source of AI Assistant's answers
Easy Navigation: Clickable links for quick location of information in long documents
Streamlined Content Creation: One-click "copy" button for sharing insights in emails, presentations, and reports

It's particularly effective for:

Extracting specific information from long PDFs
Generating document summaries with accurate citations
Navigating complex technical or legal documents
Converting document information into other formats
Creating document insights as the basis for content creation

For example, you might use it to quickly pull key statistics from annual reports, summarise technical documents, or repurposing complex white papers and case studies into shorter, more snackable formats.

More affordable PDF viewers such as UPDF offer the same capabilities.

Beyond the Tools

The real power of context design isn't in any single platform or feature. It's in how it changes the questions we ask of AI.

Instead of "How do I write the perfect prompt?", we can ask "What does this AI need to understand about our business?" Instead of optimising individual interactions, we can focus on building AI environments that grow more valuable over time.

For B2B marketers, this could be particularly powerful. When AI understands your full context - be that sales cycles or subject matter expertise, compliance requirements or competitor positioning - it transforms from a content generator into a strategic asset. You can stop worrying about whether it captures your brand voice or technical accuracy, and start exploring how it can identify patterns across customer behaviours, spot gaps in your content strategy, or adapt complex ideas for different audience segments.

This shift in thinking - from crafting instructions to designing environments - will separate organisations that merely use AI from those that truly leverage it. The tools are just the beginning. The real question is: how will you reshape your marketing function around AI that actually understands your business?

Ready to transform how your approach to AI? As B2B marketing specialists, we help firms like yours build sophisticated digital strategies that go beyond basic AI automation. Talk to us about your vision.

Book a Demo

Beyond the Prompt: How Context-Aware AI is Advancing B2B Marketing