Why small businesses need an AI mindset and strategy

Getting started with AI can seem intimidating. It is a whole new, rapidly evolving, highly complex world whose implications and consequences no one has fully grasped. So a smart move is to stop thinking about the technology and instead think about your business. Which parts of it take too much time, cause delays or could be done differently? From there you should be able to start formulating your AI strategy.

A well crafted AI strategy will integrate smart tools and technologies into a specific business process to achieve measurable objectives. These could be improving productivity or customer service, or reducing operational costs.

Imagine you decide that you need a new employee. What do you want the person to do? Which parts of the business need an extra pair of hands? How could the new recruit make your life easier? This is the mindset with which AI should be approached.

The AI strategy should have a roadmap that enables you to gauge the effectiveness of the technology. Without it, it is difficult to assess the gains that AI is bringing. For maximum effectiveness business goals should line up with the AI strategy.

The strategy will increase productivity by identifying inefficiencies in the business: typically staff having to spend too long on tasks that are administrative, repetitive and non-revenue generating.

The strategy should take into account risk management and ethics. Issues such as data privacy, bias avoidance and transparency with customers must be considered.

In addition the strategy should anticipate the business growing larger. As it does, more advanced and complex features may be required, and these ought to be able to be integrated with the existing tools.

A good starting point for AI adoption is business process management – this can be narrowed down by listening to your team about which tasks they feel they are taking up too much time.

Then map out how the process works at the moment, where its sticky points are and whether or not AI can step in there. If it can, identify which form of AI is appropriate.

The options include machine learning and deep learning; natural language processing; generative AI and large language models; AI agents and agentic automation; robotic process automation; computer vision and intelligent document processing.

AI is likely to be able to add the most value to activities where there is a need to interpret, classify, generate or extract information. Repetitive tasks and customer service are the other key areas for AI adoption.

Start with a small test running alongside the existing process. This should reveal any faults in the AI process, which can then be ironed out before it is fully deployed.

Keep the documentation relating to the AI build. This will help with integrating AI into your company’s process infrastructures.

Choosing your first project

Ideally the first AI implementation should be in an area where it can make a significant impact without disrupting operations. If you are struggling to keep up with your customers’ demands then you could look at training AI chatbots or autonomous agents.

Data readiness

AI can only work with what you give it so if your customer data is spread over different systems, not up to date or incomplete, your shiny new tools will not be as effective as they could be.

Ideally, customer relationship management software should be able to integrate structured data, such as sales figures, and unstructured data from emails and chat transcripts.

Coding made simple

Low code and no code software mean non-developers can get started with building AI tools. Some come with pre-built workflows for ease of adoption.

AI literacy

Staff may well be alarmed about being replaced by AI, which can lead to resistance. Create an AI literacy programme for them, with regular training sessions, and listen to their feedback about the tools. A chat area where AI adopters can share their experience and tips is a good idea.

FOMO : The hype and constant media attention that AI is receiving mean some business owners may feel they have to get onboard right away in case they miss out. This risks wasting money on software that does not add meaningful benefits and using up staff time in adapting to tools that are not helping them.

Not getting the team onside: staff will need proper training and a sense of how the tools will help them. Not bringing them along can lead to resistance and the new software not being used to its best effect as staff ignore it in favour of tried and tested workflows.

Short timescales: You wouldn’t expect a new employee to be as capable as an experienced one so give the AI tools the time to bed in. Every new tool will need to be integrated and optimised, and there are bound to be things that need to be adjusted for maximum effectiveness.

Wrong size strategies: What works for medium-sized or large companies probably won’t be right for you – they will have budgets and in-house technical expertise that smaller businesses simply don’t have.

Safety

Fundamental to any AI strategy is data security. Access to customer data should be limited to those people that need to see it, and your data privacy policies should be publicly available and should explain how data is stored and protected. Be willing to share your security documentation, incident response procedures and how your AI models use data.

Data breaches can be seriously damaging for companies and for customer trust. In addition any AI implementation should be compliant with relevant regulation.

Getting started

Developing an AI mindset will be a work in progress; just as AI is constantly learning so company owners will be constantly learning about what AI can do for them. Likewise an AI strategy will evolve over time as the tools are implemented and new possibilities for use emerge.

Careful documentation and paying attention to staff feedback will support further development of the AI projects that are successful and identify those that are not.

This article is brought to you by Finsbury Robinson. We offer a full suite of tax, accounting and business advisory services, and are based in south-east London. Our friendly and highly experienced team is available on 020 8858 4303 or via email at info@finsburyrobinson.co.uk