Essential Steps to Initiate Data and AI Modernisation for SMEs: A Comprehensive Guide

Initiating data and AI modernisation for an SME can be a very daunting process, given a lack of skills, resources, and technologies, as well as the investment needed. Some of these initiatives will yield benefits in the medium term, and if not implemented or prioritised correctly, can result in a big waste of money and time.

One of the biggest mistakes that companies make when they go through data and AI transformation journeys is that they do not have a plan and governance structure in place to measure success. We've seen a lot of our customers engaging with us after spending thousands of pounds and a lot of time on huge data and AI projects that went dead, or they ran out of money before seeing any benefits, or they worked with consulting firms and lost all the visibility on what was being created and the deliverables.

The following steps can be taken by any business owner who's looking to innovate, and describe the groundwork needed before engaging with a consulting firm or hiring dedicated internal resources.

Step 1: Define your objectives and list all the use cases

Identify your business goals

You may not know what data or AI can do for you, or the possibilities that can be opened up with new technology. Start by defining strategic goals that new data and AI capabilities could support you with. Let’s start with some simple steps, in a document note down the following:

Core activity

What is your business's core activity? This might sound trivial, but if you can't explain what your business does in a few words, it will be difficult to articulate and define what your strategy is with data and AI, the objectives you hope to achieve, and the reasons why to internal and external stakeholders.

This will help you identify the areas where data and AI can add value. For example, if you are a retail company, your core activity might include inventory management, customer service, or sales.

Base customers

Who are your clients and why do they want your product? Understanding your customer base and their needs is essential for improving your product or services.

Knowing who your clients are helps identify data sources and analytics that will help improve customer experience and retention. For example, if you serve young professionals, data analytics can help you segment your user data and tailor marketing strategies to their preferences and behaviours.

Inefficient processes

What are your inefficient processes and how are they impacting your business and top line? A good way of mapping inefficiencies is to speak with your team and note time consuming tasks in their role or general complaints. You might not like what you hear but try to stay and be objective and open to feedback.

These necessary but resource intensive processes are often the best starting points to implement data and AI solutions. For example, if your accounts payable team is spending hours manually processing invoices, an AI-powered document processing solution could dramatically reduce processing time and errors. This not only frees up staff time for more strategic tasks but also improves cash flow management and supplier relationships.

Select the use cases

Once you have identified potential use cases in the previous step, it's time to evaluate and choose which ones to prioritise. This depends on what your business does, who your customers are and where you’ve identified inefficiencies. We’ve added 2 examples below:

Operational efficiency

Try to identify processes that can be optimised using data and AI. This might include inventory management, supply chain optimization, or workflow automation.

For example, a manufacturing company might use data to predict maintenance to reduce equipment downtime and increase productivity.

Product development

Data can be used to drive innovation and develop new products or services that meet your customer's needs. This can involve analysing market trends, customer feedback, and competitors.

For example, a tech startup expanding in new markets might use data to identify gaps among their competitors' offerings and differentiate their products to address those needs.

Prioritise use cases

After you list all your use cases, try to put them in a priority order, and try to evaluate the potential impact on the business and feasibility of each use case. When you prioritise try to understand those that align closely with the business goals you've identified and have the potential to deliver quick wins.

Quick wins

Prioritise use cases that can be implemented fairly quickly and show immediate benefits. These quick wins will build momentum and support for further data and AI projects.

For example, implementing a basic customer segmentation model can give you quick insights for targeted marketing campaigns, so that you allocate your budget in a more informed way, avoiding wasting money and time.

Strategic alignment

Ensure that selected use cases align with your long term strategy. This will ensure that your data and AI initiatives are contributing to the overall business objectives and provide sustainable value.

For example, if you're looking to expand internationally, you might want to prioritise data analytics for entering the new market and for mapping competition.

Step 2: Conduct a readiness assessment

Another key step in the journey to modernise with data and AI is to understand your current state and capabilities. In step 2 you need to conduct a readiness assessment to identify gaps and areas for improvement, ensuring that you've got a solid foundation for any data and AI initiative. This step can be broken down into 2 stages: 1) Evaluate your current capabilities and 2) Understand your data sources.

Current Capabilities

a) Assess current infrastructure

• Technology stack: Here you need to map and evaluate your existing technology infrastructure, including hardware, software, cloud and networking capabilities. This is essential to determine whether your current systems can support the load and complexity that new data and AI capability will introduce. Here you also need to map all your data vendors, or third-party companies managing part or the entirety of your tech stack, make sure you have the key contacts written down.

• Data storage and management: Map how your data is currently stored, managed, and accessed by end users. Here’s important to understand the scalability, security, and cost of your data storage solutions.

  • Data warehousing: Is your current data warehousing solution capable of integrating new data sources? Is it capable of supporting advanced analytics?

  • Data governance: Do you have any data governance policies or frameworks in place that ensure data quality, security, and compliance?

• Scalability and flexibility: You also need to understand the ability of your infrastructure to scale up as your data grows. This includes vertical scaling (adding more computational power to existing solutions) and horizontal scaling (adding more storage capacity). How easy or expensive would that be?

b) Evaluate human resources

• Skills and expertise: Map internal and external expertise, via your partners and vendors, and identify any gap in skills you have. Determine the proficiency of all your staff in key areas like data analytics, machine learning, or AI. Try also to identify any employees with a passion for these topics.

• Team structure: Map the current team structure and reporting lines, as well as responsibilities and decision makers. A big blocker to innovative AI or data projects is often the lack of buy-in or involvement from middle management.

c) Evaluate organizational readiness

• Change management: You must assess your organization’s readiness for change and understand the cultural readiness for adopting new technologies and processes. You can do this by focusing on 2 areas:

  • Stakeholder buy-in: Ensure that key stakeholders, execs, heads of departments and end-users, are all aligned and openly supportive of the data and AI initiatives.

  • Communication plans: You need to think about communication plans and make sure all stakeholders are well informed and engaged throughout the journey. Keeping all stakeholders updated on the progress is key for success as well as open session feedback.

• Processes and workflows: Map all existing business processes and workflows to identify areas that can benefit from automation or data-driven decision-making. This will also allow you to spot inefficiencies and opportunities for improvement.

Understand your data sources

d) Identify available data sources

• Internal data: List all internal data sources, including transactional data, customer data, operational data, and any other relevant data generated within the organization.

  • Data inventory: Create an inventory of all your data assets, detailing the type, volume, source, and quality of data available.

  • Data ownership: Next to each data source, identify who owns and manages them within the organisation.

• External data: Identify any external data sources you utilise. This could include market data, social media data, public datasets, and third-party vendor data.

  • Data acquisition: Write down the current process for acquiring new data sources, and how they get stored and integrated into existing systems.

e) Evaluate data quality and usability

• Data Quality Assessment: Do you trust your data? Have you got duplicate records or inaccurate information? Have you got any validation rules in place at the point of entry for your data? Here you need to conduct a data quality assessment to understand the accuracy, completeness, consistency, and timeliness of your data.

  • Data cleaning: Is someone responsible for cleaning your data? Do you have manual or automated processes? Map how your data is being cleaned, the process, and who is responsible for it.

  • Data enrichment: Have you got external data sources enriching your data? If so, map how these are integrated into your system, the cost, the process and the fields being updated.

• Data accessibility: Assess how easily data can be accessed and consumed by end users. Have you got data access policies, and permissions, in place? Which platforms allow your users to access the data?

Step 3: Invest in internal skills, establish governance and leadership

To successfully implement data and AI initiatives, it’s crucial to have the right skills and people within your organisation, as well as establish strong governance and leadership structures. The following step will help you to ensure your organisation is well equipped to handle new technologies and that projects are aligned with business goals.

Form a steering committee

You need to create a steering committee comprising key members from various departments, including IT, operations, marketing, and finance. The primary job of this committee is to provide strategic direction, oversee project progress, and ensure continuous alignment with the business objectives. Try to select the best people for this job, and avoid packing the committee with too many people who have little value to add or are known for being “bureaucracy makers”. We want to create an agile structure that allows for agile work. Remember these types of initiatives require a top-down approach to work.

• Roles and responsibilities: You need to clearly define the roles and responsibilities of each member, including decision-makers who have the authority to provide resources and resolve issues promptly. You need to delegate authority to these members, to capitalise on the momentum and quick wins and resolve problems quickly.

• Regular meetings: Schedule regular meetings to review progress, address challenges, and adjust strategies as needed. You must have an agenda for these meetings and discussion points where decisions need to be taken by the group. This will foster accountability and keep the projects on track.

Upskill your existing staff

• Identify skill gaps: Conduct a skills assessment to identify gaps in your team’s knowledge and expertise related to data and AI. This will inform your hiring plans and job descriptions, so everyone knows who you should hire and what skills are missing.

• Training Programs: Identify those people who are keen to learn about data and AI and develop training programs tailored to your team’s needs. Consider online courses, workshops, and certifications from reputable providers. This ensures you can retain the domain knowledge of your existing team while creating new expertise.

• Hands-on learning: The best way to learn new things is by learning by doing. Try to encourage hands-on learning through internal projects, and collaboration with more experienced colleagues.

Hire specialists

• Recruit data experts: If skill gaps cannot be filled internally, put together plans for hiring specialists. Look for candidates with a proven track record, hands-on experience and who are adaptable to change.

• Job Descriptions: Write clear job descriptions outlining the following sections:

  • The company

  • The mission

  • Their responsibilities

  • The experience required

  • The interview process.

  Be clear on compensation and make sure it matches with market standards, otherwise, you won't be able to hire the best talents.

• Interview Process: Data and AI experts are in high demand, so make sure you don't create a very long interview process. We recommend a quick intro chat to describe what the role is and give an overview of the company, as well as understand the experience and cultural fit of the candidate. We then recommend a technical interview followed by a case study to evaluate problem-solving abilities, and a final step to review the exercise with the candidate and meet other senior members of the team. The whole process should be completed within a week. The data and AI market moves fast with good candidates.

Appoint data champions

• Designate project leaders: Appoint data champions across your business who will lead specific data and AI projects. These individuals need to have a strong understanding of the business and the ability to drive initiatives forward, as well as the ability to coordinate and engage with different stakeholders.

• Responsibilities: Your data champions will be responsible for coordinating project tasks, communicating with stakeholders, and ensuring that projects stay aligned with business goals.

• Empower them: Empowering data champions with the authority and resources is essential to make decisions and overcome obstacles. This will ensure they can effectively lead their projects to success. There is no worse thing than a champion without decision-making power.

Step 4: Seek external expertise

Hiring external expertise can provide additional knowledge and experience required to successfully land your data and AI initiatives. However, it’s important to engage with external partners wisely to ensure that they are accountable, effective and aligned with your overarching business goals.

Engage consultants wisely

• Select the right partners: Choose consultants and external partners with a proven track record or hands-on experience in data and AI transformation projects. Look for those who understand your specific challenges and opportunities and are transparent.

• Scope of work: You must define the scope of work, deliverables, success KPIs and timelines in your contract. This will ensure that both parties have a shared understanding of project goals and expectations.

• Knowledge transfer: Ensure that consultants are committed to transferring knowledge to your internal team. It is vital to build internal capabilities and reduce dependency on external resources over time.

Pilot projects

• Start small: Start with small pilot projects to test and validate new approaches before scaling up. You need to select projects with clear objectives and measurable outcomes.

• Learn and iterate: Use the results of pilot projects to refine your strategies and methods. Learn from successes and failures to improve future implementations.

• Demonstrate value: Use pilot projects to demonstrate the value of your new initiatives to stakeholders, and build momentum and support for larger-scale projects.

How We Can Help

• Expert Guidance: Our team offers expert guidance on all aspects of data and AI modernisation, from strategy development to hands-on implementation and training.

• Customised solutions: We work with you to develop tailored solutions that meet your specific needs and goals. Whether you need help with a single project or a comprehensive transformation, we can provide the expertise you need.

• Ongoing support: We offer ongoing support to ensure that your initiatives continue to deliver value over time. This includes troubleshooting issues, optimisation, and updates as needed

Conclusion

Implementing data and AI modernization in an SME is no small feat, but with a clear plan, the right skills, and strategic use of external expertise, it can transform your business operations and set the stage for long-term success. Follow these steps to navigate your data and AI transformation journey effectively, and ensure you achieve tangible benefits without unnecessary pitfalls.