Setting AI goals: how to pick the right outcomes for your business

Setting AI goals: how to pick the right outcomes for your business

Artificial intelligence, in its many forms, is increasingly accessible to small and medium businesses. Tools like Microsoft Copilot are moving AI out of the realm of abstract possibility and into everyday operations. However, the availability of powerful technology doesn't automatically translate into tangible business benefits. The real challenge, and the real opportunity, lies in defining what you want AI to *do* for your business. Without clear goals, AI initiatives risk becoming costly experiments rather than strategic investments.

This isn't about chasing the latest tech trend. It's about applying a sensible business lens to a powerful suite of new tools. Just as with any significant investment, you need to articulate expected returns and how you will measure them. This article will guide you through setting meaningful AI goals, ensuring your efforts are focused, impactful, and ultimately, good for your bottom line.

Start with business problems, not technology

A common pitfall is to become enamoured with the technology itself, leading to a "solution looking for a problem" scenario. Instead, begin by identifying the pain points, inefficiencies, or growth opportunities within your business. Where are you spending too much time? What tasks are repetitive and prone to error? Where are you missing sales opportunities?

Think about specific areas where a well-applied technological solution could make a material difference. For instance:

• **Customer Service:** Are response times slow? Do agents struggle to find information quickly?
• **Marketing:** Is content creation a bottleneck? Is it difficult to personalise communications at scale?
• **Operations:** Are there manual data entry tasks consuming significant staff hours? Are processes inconsistent?
• **Sales:** Is lead qualification time-consuming? Is it difficult to track and follow up on all prospects?

Once you have a clear understanding of your business challenges, you can then consider how AI might offer a viable solution. This top-down approach ensures that any AI deployment is purposeful and directly addresses an identified business need.

Define clear, measurable outcomes

Once you have identified a business problem, the next step is to translate that into a clear, measurable goal for your AI initiative. This is where the SMART framework (Specific, Measurable, Achievable, Relevant, Time-bound) proves invaluable. Simply wanting to "use AI" is not a goal; wanting to "reduce customer service email response times by 25% within six months" is.

Consider the following examples relevant to SMBs exploring tools like Copilot:

• **Instead of:** "Improve marketing."

**Aim for:** "Increase successful email campaign engagement (open rates and click-throughs) by 15% for product announcement emails by Q4, using AI-assisted content generation."

• **Instead of:** "Make staff more productive."

**Aim for:** "Reduce the average time spent drafting internal reports by 20% across the sales team by the end of Q3, through AI-powered summary and drafting assistance."

• **Instead of:** "Better customer support."

**Aim for:** "Decrease the number of basic customer queries requiring human agent intervention by 30% within the next nine months, by deploying an AI-powered FAQ chatbot."

• **Instead of:** "Streamline operations."

**Aim for:** "Automate the categorisation and initial response to 40% of incoming supplier invoices by the end of the year, reducing manual processing time by 15 hours per month."

The key is clarity and quantification. You need to know what you're aiming for and how you'll determine if you've hit the mark.

Prioritise for impact and feasibility

With a list of potential problems and measurable goals, the next step is to prioritise. Most SMBs don't have unlimited resources for AI experimentation. Focus on areas where AI can deliver significant impact with a reasonable level of effort and risk.

Consider these factors when prioritising:

• **Business Impact:** Which goals, if achieved, would have the greatest positive effect on revenue, cost reduction, efficiency, or competitive advantage?
• **Feasibility:** Is the technical solution readily available (e.g., within an existing platform like Microsoft 365, or a well-documented third-party tool)? Do you have the necessary data? Do your staff have the basic digital literacy required for adoption?
• **Cost-Benefit:** What is the estimated cost of implementing the AI solution (software, training, potential integration work) versus the projected benefit?
• **Data Availability and Quality:** AI tools thrive on data. Do you have the necessary data for the AI to learn from or process? Is it organised and clean enough? For example, if you want an AI to summarise meeting notes, are your meeting notes typically well-formatted and digitised?
• **Risk:** What are the potential downsides? While AI is becoming more reliable, it's not foolproof. Consider data privacy, accuracy, and the impact on workflows. Start with lower-risk applications.

Often, starting with a relatively contained project that addresses a clear pain point can be a good way to gain experience and build confidence before tackling more complex initiatives. Think "quick wins" that demonstrate value.

Involve your teams

AI adoption is not just a technological challenge; it's a people challenge. Ignoring the human element can derail even the best-laid plans. Involve the teams who will be directly affected by AI from the outset. They are the ones who understand the nuances of their daily tasks and will be critical to successful implementation.

• **Gather insights:** Ask your staff where they feel their time is wasted or where processes could be improved. Their input can uncover high-impact areas you might have overlooked.
• **Address concerns:** Be open about the potential changes AI might bring. Address fears about job displacement by focusing on how AI can augment roles, automate tedious tasks, and free up time for more strategic or creative work.
• **Foster ownership:** When employees feel they have contributed to the goal-setting process, they are more likely to embrace and champion the new tools.
• **Identify champions:** Find individuals who are curious about new technology and willing to experiment. These individuals can become internal advocates and help others adapt.

Remember, AI tools like Copilot are designed to be assistants, not replacements. Framing it as a way to enhance capabilities, rather than eliminate jobs, is vital for a smooth transition.

Measure, learn, and iterate

Once your AI initiative is underway, don't just "set and forget." Continuously monitor the metrics you defined in your goals. Is the AI delivering the expected outcomes?

• **Track progress:** Regularly review your chosen metrics (e.g., reduced response times, increased engagement rates, fewer manual errors).
• **Gather feedback:** Collect qualitative feedback from the users of the AI tools. What's working well? What's frustrating? Where are there unexpected challenges?
• **Adjust and refine:** Based on your measurements and feedback, be prepared to make adjustments. Perhaps the initial configuration needs tweaking, or user training needs to be enhanced. AI, especially in its early stages of adoption, often requires a degree of experimentation and iteration to get right.

This iterative approach ensures that your AI investments remain aligned with business needs and deliver continuous value. It also allows you to learn from smaller deployments before scaling up.

Your next step: A focused AI roadmap

The journey with AI for SMBs should be deliberate and strategic. By starting with business problems, setting clear and measurable goals, prioritising for impact, involving your teams, and committing to continuous learning, you establish a solid foundation for success.

Your immediate next step should be to convene your leadership team and perhaps a few key operational staff. Dedicate time to a brainstorming session focused purely on "What are our biggest operational challenges or growth blockers right now?" Document these rigorously. Only then should you begin to consider how AI might present viable, measurable solutions. This disciplined approach will maximise your chances of successful AI adoption and deliver genuine competitive advantage.