-->
Login Sign up ?

A practical guide to using AI in business

Tools & Resources

Key learnings

  • Artificial intelligence (AI) is intelligence demonstrated by machines, as opposed to intelligence demonstrated by animals and humans.
  • AI is not perfect; you’ll get the best results when you treat it as a starting point to improve efficiency and fuel new ideas, then add your human judgement.
  • The best way to learn AI is to experiment and try it yourself.

In today’s fast-paced business landscape, staying ahead of the curve is paramount for success. One revolutionary tool that has the potential to transform the way companies operate is artificial intelligence (AI). With so much on offer, it’s understandable that you might feel lost when exploring the world of AI. In this guide, we’ll delve into the many ways AI can benefit your business, providing advice tailored to match your stage of digital maturity. 

What is artificial intelligence (AI)?


First, let’s get an understanding of what AI is, as well as some of the commonly used jargon surrounding it.

AI jargon-buster

  • Artificial Intelligence (AI) – Intelligence (the ability to learn, reason, generalise and infer meaning from data and information) demonstrated by machines, as opposed to intelligence demonstrated by animals and humans. Common everyday examples include Google search, Netflix recommendations, and Siri. For instance, Netflix’s AI suggests movies and TV programmes you might like based on your previous watching habits and ratings.
  • Artificial General Intelligence (AGI) – Machines that can learn to accomplish any intellectual task that humans or animals can perform. This doesn’t exist yet, but is being worked on by companies such as OpenAI and DeepMind.
  • Machine Learning (ML) – Teaching computers to learn patterns from data and apply those patterns to things they haven’t seen before.
  • Deep Learning/Neural Networks – A method where layers of “nodes” transform inputs step by step, leading to a final output. Think of it as multiple AI layers stacked to produce more sophisticated outcomes.
  • Prompt – Any text, question, information, or code that communicates to AI the response you are seeking. This is how tools like ChatGPT interact with users.
  • Hallucination – A confident response by an AI that’s not supported by its training data – it gets something wrong.
1

How can AI benefit your business?

Using AI can have a range of benefits and greatly improve your efficiency in completing tasks. However, it’s important to remember that despite its amazing capabilities, AI is still early in its journey and the content it generates isn’t perfect, so human logic and review are critical.

If you use AI as a tool to help you and your team work more efficiently, your business can reap meaningful rewards.

Potential benefits include:

Become your own specialist – AI tools like ChatGPT offer an affordable way to begin work that you would otherwise need a specialist to do – for example conducting market research, drafting HR policies, and creating social media calendars. With the right prompts (covered later), AI can generate a strong base to refine.

Learn new skills quickly – Use AI to deep dive into topics and get concise explanations. For instance, ChatGPT can help you grasp basic accountancy concepts so you’re better equipped for conversations with professionals.

Create content faster – Draft blog posts, social media content, product descriptions, and email campaigns in a fraction of the time. Tools like Anyword.ai can help integrate relevant SEO keywords into copy.

Generate images and visuals – Tools like Midjourney can create images from your descriptions. Even if you ultimately hire a designer, AI can help you storyboard and reduce revisions, saving you time and money in the process.

Analyse data for better decisions – AI is excellent at processing large datasets to identify patterns, market trends, and customer behaviours. ChatGPT can work with spreadsheets, while platforms like Shopify include AI assistants to query store data directly.

These benefits apply at every stage of digital maturity, and they deepen as you progress and optimise your use of AI. In the next section, we’ll share tailored strategies for beginners, intermediate users, and advanced adopters.

2

How to get started with AI

Educate yourself – Ask AI to teach you! Use a tool like ChatGPT to research AI concepts, familiarise yourself with terminology and learn fundamentals.

AI strategist Salvia Mirza suggests speaking to AI like you would a colleague. Try: “I’m not very techy and I’m overwhelmed by AI. Can you help me understand its core concepts?”

Identify your business needs – Assess your business processes and spot areas where AI could make a significant impact (eg, automating repetitive tasks, improving data analysis, enhancing customer interactions).

Consider professional training – If your budget allows, workshops or personalised training can speed up adoption and help you make strategic decisions.

Experiment and iterate – Start with simple prompts in ChatGPT, then gradually explore more complex interactions. Try to ‘break’ it to learn what it can and can’t do. Start small, stay curious, and learn from both successes and challenges.

Tailored AI strategies for every stage of digital maturity


Below, you’ll find practical steps and quick wins tailored to where you are today, spanning different business functions. 

1

If you’re just starting out

  • Start practical and small: Pick one time-consuming task (invoicing, expense tracking, appointment scheduling, data entry).

  • Leverage tools you already use: Platforms like Xero and QuickBooks now integrate AI features to help automate routine finance workflows.

  • Quick win: Use a simple assistant such as ChatGPT or Copilot to draft invoice emails, create payment reminder templates, or summarise receipts.

  • Adopt a ‘human-in-the-loop’ review: Always check outputs for accuracy and tone before sending or publishing.

  • Use AI as a drafting partner: Create first drafts of blog posts, social captions, product descriptions, and email campaigns, then review and add your voice.

  • Brainstorm ideas faster: Ask AI for campaign angles, seasonal promotions, and headline variations to prevent the dreaded blank page effect.

2

If you’re already exploring and want to go further

  • Target real pain points: Identify two or three high-friction processes that take up a lot of time and energy, for example purchase order reconciliation, monthly reporting, supplier queries.

  • Set success criteria: Define simple metrics such as ‘reduce time spent on invoice processing by 30%,’ or ‘improve conversion rate by 3%.’

  • Integrate with existing platforms: Connect AI tools to systems you already rely on – for example, using Excel Copilot, Google Sheets with ChatGPT, or Shopify Sidekick, you can query data in plain language and automate repetitive tasks. This could be ‘summarise this data and tell me which product had the highest growth rate from January to March’.  

  • Train AI on your brand voice: Create a short style guide (tone, phrases, do/don’t) and paste it into your prompts. You can also store some reusable prompts for consistency when working on set tasks, eg blog posts, case studies.

  • Expand content types: Add image generation and basic video edits into your processes using eg Canva, Veed.io, Capcut to speed up production.

  • Use in customer service: Deploy simple chatbots for common queries and integrate escalation paths to humans to keep satisfaction high, allowing you to focus on more complex queries.

  • Practice prompting: Give context, audience, objective, constraints, and examples of what good looks like. Provide feedback like you would to a colleague to iterate and refine the outputs.

  • Upskill your team: Run short trainings on effective prompting, data privacy, and human review to ensure nobody in your team feels left behind.

  • Measure what matters: Track time saved and performance improvements such as conversions, and see if it justifies further AI adoption.

3

If you’re looking to optimise your use of AI

  • Expand gradually across departments: Prioritise areas of your business with similar data and workflows, for example finance, operations and supplier management.

  • Introduce policies and processes: Get everyone on the same page by outlining a policy and standardised processes covering data handling, prompt best practice, and quality management. For example, you might want to include review checkpoints, tone checks, and legal/privacy checks, as well as maintaining a shared prompt library and performance dashboards. 

  • Review current tools: As the field is moving fast, take some time to regularly audit your use of features and explore advanced capabilities. For example, are you making use of document classification and anomaly detection capabilities in your accounting software?

  • Find and fix new bottlenecks: AI shifts bottlenecks rather than eliminating them, so keep tracking down where they happen. Is it now in approvals, editing, or data clean-up? Mapping the new pain points will allow you to address those next, for example by adopting standardised prompts, or doing a big data clean before starting a new project, so data errors aren’t duplicated every time.

  • Invest in strategic training: Move beyond teaching staff how to use tools to focus on when and in what scenarios to use AI tools – for example you could provide decision-making frameworks, templates for ROI tracking and risk awareness and assessment for new projects.

  • Operationalise your brand voice: Use settings within the tools that let you create custom brand guidelines and apply them automatically to help improve consistency and reduce human review and editing time.

  • Go beyond text: Experiment with AI video (eg Synthesia, Runway or Canva’s AI video features) and AI audio/voiceover (eg ElevenLabs, Speechify) to enrich your content mix.

  • Introduce greater personalisation: Use AI to segment your audiences and tailor messaging. You can also experiment with dynamic content and interactive experiences based on your segmentation.

  • Scale thoughtfully: Expand to new channels or formats gradually and run A/B tests and creative sprints to explore ideas without diluting quality.

The basics of ChatGPT


ChatGPT is one of the most used AI tools, and its possibilities are seemingly endless. When it first launched, it took ChatGPT only five days to reach one million users –something that took a revolutionary service like Netflix over three years to achieve. It’s also simple to use and understand, making it a perfect entry point to test what AI can do for your business.

What is ChatGPT?

ChatGPT is a large language model (a very smart assistant) trained on a mixture of internet content, books, and articles. It is predictive, meaning its answers are based on statistical patterns learned from data rather than a live database of facts. It will be slightly out of date, and its outputs depend heavily on your prompt and its training cutoff.

What makes a good prompt?

  • Be clear and specific, as ambiguity increases the chance of incorrect answers.
  • Provide context (audience, objective, tone) and any constraints (word count, format).
  • Give it a persona, for example, “You are a highly successful and experienced sales professional.”
  • Iterate and give feedback: Tell it what worked and what didn’t. Build on initial answers and ask for alternatives.
  • If you can give an example of something you’ve already produced or some sample text to use for tone, it will help. If you do this, be clear whether you want the tone or style replicating or used only for inspiration.

Example prompt (customer research)

“You are a highly successful and experienced sales professional. You understand audiences and what drives them to purchase a product. I want you to help me identify the needs of different audiences who might buy our product. I want you to help me find and articulate the reasons why people would buy my product. Please ask me questions one by one to learn about the product and why people use it. Once you have enough information, please create a table listing each audience, a list of their needs and what is most likely to motivate them to buy the product. If you understand, give me the first question.”

AI tools you will find useful


Now you’ve seen what AI can do, it’s time to take the plunge and try tools for yourself. Try a few to find what works best for your style.

1

Create copy
2

Image generation
3

Video generation
4

Podcast creation
  • Headliner – this tool acts as a podcast assistant.
  • Descript – this tool transcribes audio and allows you to edit recordings like you would edit texts.
  • Wondercraft – this tool can convert articles into podcasts.
5

Translation
  • HeyGen – this tool can translate video into other languages.
6

Integrations with other platforms
  • Shopify Sidekick - helps Shopify users move along their entrepreneurial journey. You can ask it questions regarding data you would like to see for your business and it will generate it for you. For example, ‘I had a drop off in Sales in March 2023, why?’
  • Canva’s AI image generator – through your Canva account, it can create images from text prompts which can be added to your designs.
7

For your own curiosity
  • There’s an AI for that - a website that tells you if there is any AI available for jobs/tasks you would like to perform.
  • WAIFinder – an interactive digital map developed by UKRI that shows companies, funders, incubators, and academic institutions that are involved in creating artificial intelligence (AI) products, services, processes and research.

Common pitfalls to avoid 


Publishing unedited AI content
Pitfall: AI outputs can be generic or contain errors.
Fix: Always apply human review for accuracy, tone, and brand voice.

Trying too many tools at once
Pitfall: Fragmented workflows and low adoption.
Fix: Start small, choosing one or two tools per use case and standardise prompts/templates.

Skipping prompt hygiene
Pitfall: Vague prompts produce weak outputs.
Fix: Give clear context, objectives, constraints, and examples, then iterate with feedback.

Ignoring data privacy
Pitfall: Sensitive information accidentally shared in prompts.
Fix: Avoid pasting confidential data – use only anonymised examples and check each tool’s privacy settings.

Not having a definition of success
Pitfall: It’s hard to prove value and sustain momentum.
Fix: Set simple metrics, for example time saved, errors reduced, or response times improved, and review monthly.

Over-automating customer interactions
Pitfall: Robotic experiences can frustrate customers and lower satisfaction.
Fix: Keep humans in the loop by offering easy escalation to a person if initial interactions don’t resolve the customer’s query.

Neglecting team training
Pitfall: Inconsistent results, low confidence in using tools, risks with data.
Fix: Run short sessions on prompting, brand voice, review standards, and governance.

The integration of AI into business processes opens up a world of opportunities. While AI is not without its limitations, when viewed as a tool to enhance efficiency and productivity, the benefits are clear. Whether it's automating tasks, learning new skills, or analysing data, AI can be a game-changer for businesses looking to thrive in a competitive landscape. As we stand at the forefront of the AI revolution, embracing this technology intelligently can pave the way for a more agile, efficient, and successful future for your business.

Next steps…

Tags:

Did you find this article useful?

Like what you see?

Join our mailing list to get the best information, expertise and finance directly to your inbox. Get a bulletin full of practical and inspirational content to help you and your business do more.

You are subscribed.

Thank you for subscribing to UMi Sat Nav for business. You can now view all are content.