Practical guides
Small Ideas. Smarter Choices.
Practical tips at the intersection of marketing, artificial intelligence, creativity, and sustainability. From making better use of AI tools to communicating responsibly and reducing digital waste, each tip offers a simple idea worth considering, testing, or putting into practice. No lengthy guides. Just useful perspectives for the way we work, create, and communicate. For more practical tools and guides, explore our Resources page.
QUICK TAKE 01 I AI IN MARKETING
Your AI Doesn’t Know Your Customer
AI can produce a convincing marketing message in seconds. That doesn’t mean it understands the people you’re trying to reach. Before using AI-generated copy, check it against real customer questions, objections, and feedback. Look for assumptions about what customers want, language they wouldn’t use, and benefits your product cannot substantiate.
The test: Would your customer recognise themselves in this message, or does it sound like something written for everyone?
QUICK TAKE 02 I SUSTAINABILITY IN MARKETING
A Green Claim Needs More Than a Green Label
Words such as eco-friendly, green, and sustainable can sound reassuring, but they tell customers very little without evidence.
Instead of making broad environmental claims, communicate something specific and verifiable. Explain what has changed, which part of the product or process it affects, and what limitations remain. A measurable improvement is more credible than an ambitious promise.
The test: Can a customer independently verify your environmental claim, or are you asking them to take your word for it?
QUICK TAKE 03 | AI ART & CREATIVITY
A Beautiful Image Can Still Be a Bad Brief
AI image generators can produce striking visuals, but visual appeal alone doesn’t make an image effective.
Before choosing an AI-generated image, check whether it communicates the intended message, suits the audience, and works in its final setting. Look beyond colour and composition. Examine misleading details, visual inconsistencies, and anything that might undermine the message.
The test: If you removed the visual effects, would the idea behind the image still be worth communicating?
QUICK TAKE 04 | MARKETING STRATEGY
More Content Doesn’t Mean More Attention
Publishing every day can keep a brand visible, but frequency alone doesn’t make its communication valuable.
Before creating another post, examine what your audience actually responds to. Look at saves, meaningful comments, enquiries and actions taken, rather than relying only on impressions or likes. A useful piece of content that answers a real question may outperform a week of routine updates.
The test: If you published half as often, would your audience lose anything genuinely useful?
QUICK TAKE 05 | AI & SUSTAINABILITY
The Most Sustainable AI Output Might Be the One You Don’t Generate
Not every creative task needs a fresh AI-generated result. Repeatedly generating near-identical images, rewriting approved copy, or testing minor variations can consume additional computing resources without improving the outcome.
Build a small library of approved visuals, prompts, and content assets. Reuse what works, edit existing material when practical, and generate something new when the brief genuinely requires it.
The test: Are you generating a new result because the work needs it, or simply because the tool makes it easy?
QUICK TAKE 06 | RESPONSIBLE AI MARKETING
Your AI Disclosure Shouldn’t Need a Detective
When AI-generated people, places or events appear in marketing, audiences may mistake a fictional scene for something that actually happened.
A short, visible disclosure can prevent that confusion. Explain when imagery is synthetic, particularly when it depicts realistic testimonials, product experiences or future scenarios. Place the disclosure where people encounter the content, not somewhere buried in the terms and conditions.
The test: Could someone reasonably mistake your AI-generated visual for photographic evidence of a real event or experience?
QUICK TAKE 07 | SUSTAINABLE MARKETING
Don’t Confuse Recyclable With Recycled
A product labelled recyclable isn’t necessarily made from recycled materials. And being technically recyclable doesn’t guarantee it will be recycled after use.
When communicating packaging improvements, distinguish between recycled content, recyclability, and actual collection or recovery. Be specific about materials, percentages, and geographical limitations. These details help customers understand what your environmental claim really means.
The test: Does your packaging claim describe what the material contains, what could happen to it, or what actually happens after disposal?
QUICK TAKE 08 | AI & SUSTAINABILITY
Ask AI to Challenge Your Green Ideas
AI can help generate sustainability campaign ideas, but it can also be useful for identifying weaknesses before those ideas reach the public.
Ask it to examine your campaign from the perspective of a sceptical customer. Have it flag vague environmental claims, missing evidence, unintended consequences, and assumptions that need verification. Treat the results as questions to investigate, not as verified findings.
The test: Have you used AI only to make your sustainability message more persuasive, or also to challenge whether it is accurate?
QUICK TAKE 09 | AI & SUSTAINABILITY
AI Can Spot Waste You No Longer Notice
Small inefficiencies often become part of everyday operations because nobody questions them.
AI can help analyse energy records, inventory patterns, delivery schedules, or material usage to identify possible areas of waste. Start with reliable data, ask it to highlight unusual patterns, and investigate the findings before making changes. A useful insight is only the beginning; actual savings need to be measured.
The test: Are you using AI to generate more sustainability reports, or to identify something you can genuinely improve?
QUICK TAKE 10 | SUSTAINABLE MARKETING
The Greenest Product Isn’t Always the Newest
Marketing often encourages customers to replace products with newer, supposedly more sustainable alternatives. But replacing something that still works can create additional environmental impacts through manufacturing, transport, and disposal.
Before promoting an upgrade as an environmental improvement, consider product lifespan, repairability and the impact of replacement. Where appropriate, encourage maintenance, repair or continued use rather than unnecessary purchases.
The test: Does your sustainability message encourage a genuinely better choice, or simply give customers another reason to buy?
QUICK TAKE 11 | AI IN MARKETING
Personalisation Without Permission Isn’t Personal
AI makes it easier to tailor marketing messages using customer behaviour, preferences and purchasing patterns. But greater personalisation doesn’t automatically create a better customer experience.
Use only data you are authorised to process, explain how it informs your communication, and give people meaningful control over their preferences. Avoid using sensitive information or making assumptions that could feel intrusive.
The test: Would your customer be comfortable knowing exactly how you personalised their message?
QUICK TAKE 12 | AI IN MARKETING
AI Agents Can Quietly Change Your Brand Voice
When multiple AI agents create, edit and repurpose marketing content, small changes in language, tone and messaging can accumulate. Over time, your brand may begin sounding different without anyone deliberately changing its voice.
Create a reference library of approved content and define your brand’s vocabulary, tone and writing boundaries. Periodically compare agent-generated material against these examples to identify stylistic drift before publication.
The test: If three different AI agents created content for your brand, would customers recognise the same voice in all three?
QUICK TAKE 13 | AI & SUSTAINABILITY
When You Run AI, It Can Change Its Carbon Footprint
The electricity powering AI systems doesn’t have the same carbon intensity throughout the day. It varies depending on the location, energy sources, and demand on the grid.
For organisations running flexible AI workloads, scheduling non-urgent tasks during periods of cleaner electricity can help reduce associated emissions. However, the benefit depends on where the computing takes place and whether workloads can actually be rescheduled.
The test: Could your organisation shift non-urgent AI processing to lower-carbon periods without affecting delivery or performance?
QUICK TAKE 14 | AI ART & CREATIVITY
The Same AI Image Can Look Different After Uploading
Your AI artwork may look perfect on your screen but appear noticeably different once published. Colour profiles, image compression and automatic resizing can affect colour accuracy, gradients, sharpness and fine details.
Export images in the appropriate colour space, usually sRGB for web use, and check how your publishing platform processes them. Always inspect the final uploaded image on both desktop and mobile, particularly if your artwork contains subtle gradients or precise colour combinations.
The test: Are you approving the original artwork on your computer, or checking the version your audience actually sees?
QUICK TAKE 15 | SUSTAINABILITY & MARKETING
A More Efficient Product Can Still Increase Consumption
Making a product more energy-efficient doesn’t always lead to lower overall environmental impact. When something becomes cheaper or easier to use, people may use it more frequently, offsetting some of the expected savings.
This is known as the rebound effect. It matters when promoting energy-efficient appliances, transport, digital services and even AI tools. Look beyond efficiency per use and consider whether total consumption is actually decreasing.
The test: Does your sustainability claim demonstrate a reduction in overall resource use, or only an improvement in efficiency per unit?
QUICK TAKE 16 | SUSTAINABILITY & MARKETING
Your Social Media Algorithm Can Reward the Wrong Behaviour
Social platforms optimise recommendations around signals such as engagement and viewing time, not necessarily environmental accuracy or positive impact.
AI can help analyse which sustainability messages attract attention, but high engagement doesn’t mean audiences understand the issue or intend to change their behaviour. A sensational environmental claim may outperform a carefully explained one.
Compare engagement with more meaningful indicators, such as informed questions, resource downloads or measurable participation.
The test: Is your sustainability content helping people make better decisions, or simply generating reactions that the algorithm rewards?
QUICK TAKE 17 | AI & SUSTAINABILITY
AI Can Make Waste Look Like Efficiency
AI can optimise delivery routes, inventory levels and production schedules. But improving one part of a system doesn’t necessarily reduce its overall environmental impact.
For example, faster deliveries may encourage more frequent, smaller orders, increasing packaging and transport activity. This is why AI-driven efficiency claims should be assessed across the entire operation, not just the process being optimised.
Measure total resource consumption, emissions and waste before and after introducing AI, including any changes in customer behaviour.
The test: Has AI genuinely reduced your organisation’s environmental impact, or simply made one process more efficient while shifting the burden elsewhere?
QUICK TAKE 18 | AI & SUSTAINABILITY
Your AI Agent Might Be Doing Work Nobody Needs
An AI agent can continue checking information, calling tools and repeating tasks long after those activities stop delivering useful results. Unlike a single prompt, an automated workflow may run hundreds of times without anyone reviewing its necessity.
Set limits on how often agents operate, define clear stopping conditions and monitor unsuccessful or repeated tool calls. Review whether scheduled tasks still serve a purpose. Reducing unnecessary activity can save computing resources and operating costs.
The test: If you switched off your AI agent for a week, which activities would anyone genuinely miss?
QUICK TAKE 19 | DIGITAL MARKETING
Your Customers Can Help You Find Your Next Content Idea
Your most useful marketing ideas may already be hiding in customer questions, support conversations and product reviews.
Use AI to group recurring questions, identify topics customers struggle to understand and uncover gaps in your existing content. Turn these findings into practical explainers, demonstrations or FAQs that answer genuine needs. Remove personal information before analysing customer feedback with AI tools.
The test: Does your next content idea answer something customers have actually asked, or something your marketing team assumes they want to know?
QUICK TAKE 20 | DIGITAL MARKETING
One Good Idea Can Become a Useful Content Journey
Instead of treating every social post, email and website article as separate pieces of content, connect them around one useful subject.
Start with a customer question. Answer it briefly on social media, explore it in a blog post and offer a practical checklist or demonstration for those who want more detail. AI can help adapt the material to each format while preserving the core message.
This creates a more coherent experience without requiring an entirely new idea for every channel.
The test: Can someone move naturally from discovering your content to understanding a subject and taking a useful next step?
QUICK TAKE 21 | SUSTAINABILITY & MARKETING
Your Email’s Carbon Footprint Doesn’t End When You Hit Send
An email campaign involves more than sending a message. Images must be downloaded, tracking systems process activity, and campaign data may remain stored long after it is useful.
Instead of focusing only on email volume, examine the entire campaign. Compress images, remove unnecessary attachments, clean inactive contacts and establish sensible data-retention periods. These practices can reduce unnecessary data processing while improving email performance.
The test: How much of your email campaign’s data is genuinely needed, and how much are you continuing to store or transmit out of habit?