The Marketing Leaders Guide To AI
This week I wrote for Marketing Unfiltered (a dedicated Marketing Leadership newsletter I edit) and I went almost 2k words into AI, the top tools, the prompts to use and free courses to help Marketing & Growth leaders get ahead.
We are told this constantly.
“AI is a lifetime technology shift”
We have heard almost every major tech leader compare the shift to running water and electricity.
We have lived through two major shifts:
(i) the internet as a lifetime shift and
(ii) mobile (turning us into the anytime-anywhere generation) and like most big shifts, we have seen thousands of tools and companies spring up.
Today AI can seem too big, “it seems like a new language” to understand and many are wary of diving into something that seems too good to be true.
We are flooded with millions of tools all offering to do amazing work and letting us down with its output.
So Harry and I wanted to tackle AI tools, prompts, and free courses that are highly rated.
Today I have dived deep into AI, liaised with trusted AI consultants from senior Marketing backgrounds, and asked the writers of Marketing Unfiltered, & the senior marketers from The CMO Circle WhatsApp Group to share their tools and prompts and any courses they have been on.
The following is broken down into sections (if you just want to jump to the tools you can click here)
The AI Use Cases
The Tools
The Prompts
The Courses
The Legal Side
The Areas To Consider With AI
💡 Tip💡 Here’s a quick dummies guide to AI and how to shape your thinking (for the near future) AI is an editor, a creative assistant, a personalised agent and an editor, it all depends on how good your questions (prompts) are and the tool you use for output.
Click on the image below to skip the breakdown and see the full list of tools, prompts and courses
👩💻 AI - The Proven Use Cases 📈
Improve your writing or rewrite/reword to improve how you communicate
Organise data and analyse data
Create original images using your prompts
Research - speed up research
Summarize longer-form information (PDFs, long decks, complicated documents)
Surface insights from long data or docs
Spot trends, synthesize information, and identify business opportunities
Build connections with colleagues/service providers with insights from their writing or docs or helping to rewrite and optimise their work
» With these use cases in mind, it makes it easier to go through the tools we have been recommended or tried and tested ourselves and then you select the areas you need help with.
⚒️ The Tools ⚒️
We have collated 25+ tools that have been used and approved by many marketing leaders. We have broken down into a handful of useful categories:
Free analytics tools
Editing tools for video and audio
Image creation tools
Search and sales tools
Brand tracking tools to help you forecast more effectively.
Check out the tools section on the Google Sheet here
💡 Tip💡 Here is a simple 5-step approach to win at AI tools without it feeling daunting
Collect what are the reoccurring issues you are facing at work or your team is facing
Find and test a set of tools to answer my problem (ChatGPT vs Claude Vs Gemini for instance) and test out prompts that help break down the problem(s)
Share learnings with your team(s) and centralised lessons on your shared department space (think Google docs, Notion etc)
Run sessions with your team(s) where you share prompts and approve certain tools
Replace your current tool (let’s say Google) and default to a tool you trust from testing (let’s say Perplexity or Claude) and add their app onto your phone and desktop. Then use a set of prompts to co-create steps forward.
⌨️ The Prompts ⌨️
For those who aren’t familiar, prompts are the best way to query AI tools with chat interfaces to answer your questions or solve problems most effectively.
We asked two trusted friends Adam from Podup (who offers AI consultancy via Best Practice AI) for their prompts, I asked Katie McPhee who has a number of years working with and for startups in the AI/automation space and Katie is now running generative AI workshops for marketing teams.
And, I asked Claude (why not ask AI to help with AI 😉) the best prompts to use on Claude to get the most out of it.
Here are my favourite simple prompts
ChatGPT Help me problem solve: I am facing a challenge related to <add problem> here is an example <performance marketing drop>. Please generate a list of problem-solving strategies that I can use to overcome this challenge. Include practical steps for implementing each strategy and potential outcomes.
ChatGPT help me research topic: I am preparing to research on <topic>. Please help me create a comprehensive outline that includes key sections, subtopics, and points to include. Please ensure you Include suggestions for supporting evidence and data I should include in each section.
ChatGPT help me explain simply: I need to explain a technical process related to <topic> to someone with little to no <topic> background. Please break down the process into simple, easy-to-follow steps, using clear language and analogies where appropriate to help with their understanding.
ChatGPT write social media caption: Write me a compelling social media caption for this <image>, using a <type of> tone and focusing on <topic>.
Claude prompt for personal development plans: Can you help me create a personal development plan for my <title of role>, with a focus on <topic>
Claude / ChatGPT prompt for customer insights: Tell me 10 frustrations, 10 desires and 10 fears of <buyer> looking for <topic>
Claude/ChatGPT prompt for improved ad copy: Use the AIDA framework to write ad copy on the following ad for a <topic> <launch date>
Claude/ChatGPT prompt for list of analogies: can you help me build a list of analogies for a <topic> with <feature>
Any AI tool - prompt to break down something very simply explain <topic> to me like i'm a child
💡 Tip💡Lenny’s Newsletter has a great overview of how to think about getting the most out of prompts
👩💻 The Courses 👨💻
We added a number of free courses to help you understand AI and the tools, with free offerings from Google, Microsoft and Harvard.
Microsoft - https://microsoft.github.io/AI-For-Beginners/
OpenAI - https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/
Coursera - https://www.coursera.org/learn/big-data-ai-ethics
👩⚖️ The Legal Side 👨⚖️
We spoke to a friend of Marketing Unfiltered Mark Stafford - Mark is a Partner @ leeandthompson.com, he suggested the following:
In the UK, the legal position as to AI-generated copyright works is that the person or persons who made the arrangements for the creation of the work is the first owner of copyright in that work. Ordinarily, that means the owner of the AI system rather than the individual providing the input instructions.
However, many generative AI systems assign all rights in the output to the user. For example, OpenAI (owner of ChatGPT and the AI image generator Dall-E) has a section in its Terms of Business which assigns all rights in and to the AI's output (i.e the words / images generated) to the user (i.e the person providing the input instructions).
Ownership of Content. As between you and OpenAI, and to the extent permitted by applicable law, you (a) retain your ownership rights in Input and (b) own the Output. We hereby assign to you all our right, title, and interest, if any, in and to Output.
There are a couple of important points to note though.
First, there will always be a question mark over the originality of the works the AI has created. OpenAI notes in its Terms of Business that its Output may not be unique, and there is also a risk that its Output is not original, and is substantially similar to one or more third-party works on which it has been trained (so potentially infringes copyright). OpenAIs Terms of Business state:
Similarity of Content. Due to the nature of our Services and artificial intelligence generally, output may not be unique and other users may receive similar output from our Services. Our assignment above does not extend to other users’ output or any Third Party Output.
OpenAI gives no warranties as to its Output being non-infringing, and its maximum liability in the event a claim is made in respect of any commercial use of the Output is very low, limited to the greater of $100 or one year's subscription fee. If the AI system has been trained using a data set, where all permissions for the training have been granted from the relevant rightsholders, then infringement will not be an issue, however where the AI's training has used data from unknown and unlimited sources, there will always be some risk.
Second, if you are using AI works that have been created from a third party inputting the prompts, you should make sure that the rights in the AI works are validly assigned to you by the person who provided the input instructions, so you can use the AI content freely.
Under UK law, any copyright created by an employee in the course of their employment is automatically assigned to the employer, however for AI content that does not apply, as the works are not created by the employee (they are created by the AI and then assigned to the user). If you are an employer, then you will need to make sure that there is an express assignment of rights from any employees inputting the prompts. It may already exist in your standard employment contract, however you should check to make sure (and get advice from a lawyer if you are in doubt). The same goes for freelancers, there needs to be a clause in the freelance agreement expressly assigning any IP that is created in the product of the freelancer's services to you, otherwise the rights in the AI content will still be owned by the freelancer, which could cause problems later on.
The 🔟 Areas To Consider With AI As A Department Leader
Trust: There will need to be a phase where you create a trusted environment for AI tools and prompts and review and review the output of your teams and the tools they use
Improve Internal & External: Use AI tools to improve training, knowledge sharing and communication between teams and cross-functional departments - many have overlooked the power of internal data and improving comms and jumped to varied quality of output
Data! General AI data access is old data, not up-to-date data, so unless you are uploading large real-time data sets it should be a constant thought of which tool you are using and what data it is referencing with answers and recommendations
Quality > Quantity: Quality will always beat out quantity - ensure your team still has a high-quality barrier over demand to chase the quantity of content created
Hallucinations: AI tools still hallucinate and will mislead your team members, be mindful of blind trust and testing AI without testing
Power Up: Many team members will use this as their opportunity to boost productivity and increase their output, the question will be is this helping and how can they teach their new skills to your team members
Branding Your AI: If you use the same tools as others the output won’t be unique or built on the same training data, so prompts and style guides will be essential to win
AI IQ: There will be an IQ problem for LLMs - many LLMs will have an issue with their data and understand your request and then where these queries take place:
Small (on your own device, like your iPhone) and the abilities it has to perform on your device and on your own data
Large size models (AI in the cloud) - for the best phase of AI you will need to understand the capability of the model you are using and if you are a better use case to query via an app.
Doing → Reviewing: Going from doing the work to reviewing the AI work and outputs means you will be under pressure to reshape your team(s) and will be judged on the performance of output and an expected ramp of performance. be mindful when the shift from doing to reviewing becomes another financially driven decision
Build Your Own AI Tools: There is a huge opportunity to build internal tools that enable you and your team(s) to produce insights, research and higher quality work - work with your CFO and CPO to invest early into your internal tool kit and evolve as the external LLMs do
This was a huge piece of work so if you have appreciated this post please do share it on your favourite WhatsApp group, slack channel or on LinkedIn.
Copy and paste this link https://www.marketingunfiltered.co/p/ai
If you enjoyed this article, I break down the impact of AI and how to consider your thinking for the next 12-18 months here.
Here are the rest of my AI posts
