Greetings earthlings!

Welcome to Spark Intelligence – your AI navigator in the creative and marketing world, written by Emma, co-founder of Spark. Whether you're testing the waters or fully embracing AI, this week’s Spark Intelligence gives you quick wins and strategic insights to keep you ahead of the curve.

Here’s what’s inside this week:

1. What questions should board members be asking about AI?

2. Copyright and IP deep dive

3. Small AI win of the week - regular competitor analysis

4. AI Accelerators underway

5. Deep Research coming out of our ears!

Let’s get into it.

1. What questions should board members be asking about AI?

I sit on a couple of boards: Spark’s of course, and also as a NED for an NHS Mental Health Trust (which is a real privilege). So I’m spending a lot of time thinking about how we are taking AI forward in our organisations, how its disrupting traditional business models and how we are managing both the opportunities and risks. For board members and advisors of both agencies and brands, the central question is ensuring AI not only aligns with but enables your strategy and your values.

So, what should be on the agenda when discussing AI in the boardroom? Here are the key questions we think board members should be asking:

Governance & Accountability

💡 Who owns AI strategy in our business, and do we have the right expertise at the leadership level to make informed decisions?

🔍 Do we have clear accountability for driving AI adoption, ensuring oversight on data privacy, client trust, and creative quality?

Strategic Oversight

🔍 Do we know what impact AI is likely to have on our business model and industry positioning?

🤝 How can we use AI to make us even better at the things we are known for? How can it help us create higher quality work for our clients?

⚙️ Where can AI free up time without disrupting creativity or reducing the value of our work?

📈 How will AI change the way we price, package, and deliver our services? Are we using it for innovation or productivity?

🚀 Are we leveraging AI to create new revenue streams?

Risk Management and Ethics

⚖️ What happens if AI generates misleading insights, biased content, or incorrect recommendations? How do we check our work and who’s accountable if things go wrong?

Positioning

📢 What’s our AI stance - do we want to proactively lead the conversation with clients, or are we simply reacting to their questions and concerns?

🤝 How do we talk about AI with clients, without undermining their perception of the value we bring?

🛑 Are we being transparent with clients about how AI is used in their work?

Performance and Monitoring

📊 How do we track AI’s impact beyond efficiency—are we measuring its effect on creative quality, campaign performance, and client satisfaction?

Regulatory Compliance

🔐 How are we staying ahead of AI regulations and ensuring compliance with emerging laws?

📜 Do we fully understand copyright, IP, and licensing when using AI-generated content? How do we protect our clients and ourselves?

If you’re interested in exploring these topics for your business we run AI Leadership programmes specifically for agencies and in house creative and marketing teams. If you’d like to learn more, book a 15 minute chat below.

2. Copyright and IP deep dive

One of the most common questions we hear from leaders is about copyright and intellectual property when using AI for client work. Join me on March 4th for an expert-level webinar, where AI policy specialist Alexandra Ralph will take us on a detailed run though of these critical topics.

While we’re on the topic of copyright, the UK government is reviewing AI copyright issues right now and their consultation closes this week on 25th February. Make sure you have your say. 

3. Small AI win of the week - regular competitor analysis

Struggling to keep up with competitors’ changes? Here’s a quick way to keep track of your (or your client’s) competitors with ChatGPT Scheduled Tasks.

First, a quick recap of where to find scheduled tasks in ChatGPT:

You need to be on a paid plan (which you should be anyway to keep your data safe!). Go to the drop down menu top left of the ChatGPT window where you choose the model you are using, and select ‘ChatGPT 4o with scheduled tasks’.

Next, write a prompt describing what you would like it to go off and do, and how often you want it to do it. Here is my simple scheduled task prompt for competitor analysis:

I found this daily task gives me really helpful nudges. But if I was doing a deep competitor review or starting some positioning work either for myself or a client, I would want to start with much more nuanced information than this Task gives me. I did spend a lot of time pushing the Task further to break down more detail, but I hit the limit of what scheduled tasks can do! No complaints, it is still in Beta after all.

So if you are hitting the limits of what you can do with Tasks, or want a more structured starting place for a deep competitor review, here’s the prompt I used in a conversation with good old ChatGPT 4o. It gave me more nuance, looking for market differentiation and compares directly to another brand:

Make this prompt your own - build on it, iterate and improve it over time - and let me know how you how you get on. Share it back with your builds in the comments. I’m looking forward to hearing what the Spark community hive mind comes up with.

4. AI Acceleration is underway 🚆

This month we’ve been travelling up and down the country from Oxfordshire to Leeds, massively enjoying working with our current AI Accelerator clients: mark-making* and Analogue. Two agencies serving very different sectors, both looking to integrate AI deeper and grow their team's capabilities.

We work with two agencies / in-house teams a month across your leadership team, strategy team and design team, bringing in experts at the top of their field to each area. If you’d like to be one of the next teams we work with, then get in touch for a chat.

5. We have Deep Research tools coming out of our ears!

They’re all at it, first Google launched Google Deep Research, then Open AI launched their Deep Research, and now Perplexity has a feature called Deep Research. They are all called the same thing, agh!

We haven’t had a chance to do a head-to-head comparison yet (watch this space) - but here’s a summary of the consensus amongst AI thought leaders:

Strategists: When to use Deep Research tools vs your regular LLM (aka ChatGPT, Gemini, Claude, Co-Pilot)

Not all AI research tools are created equal. Deep Research tools and standard LLMs serve different purposes, and knowing when to use each can make all the difference in your workflows. The key differences:

• Deep research tools dig into complex topics, delivering structured, well-referenced insights.

• LLMs are great for quick answers, creative ideation, and general language tasks.

How to use them

Deep Research tools:

• Best for specific, in-depth queries requiring extensive analysis.

• Take longer to process (5-8 minutes for OpenAI, a few minutes for Google).

• Deliver detailed, high-quality reports—think graduate-level research.

• Use reports as a launchpad for deeper insights, and expect more reliable information than LLMs.

LLMs:

• Great for fast, broad-stroke responses and brainstorming.

• Provide quick answers but may lack depth and accuracy.

• Require fact-checking to ensure reliability.

• Use outputs as inspiration, but always verify critical details.

Bottom line

Choose the right tool. Need detailed analysis? Go deep research. Want ideas fast? Tap straight into an LLM.

That’s it for today. Have a great week and let me know what you thought by hitting an option below 👇

Cheers,
Emma 
Co-Founder, of Spark AI