GPT for SMEs: Mapping Impact and Ways to Benefit From Generative AI

By Emil Bjerg, journalist end editor of The European Business Review

Leveraging AI effectively might soon become a deciding factor in determining which small and medium-sized enterprises thrive and which don’t. This article examines the impact of the hyped GPT-4 on SMEs and advises on practical applications of the revolutionary software.

ChatGPT, and the subsequent release, GPT-4, have been greeted with equal measures of hype and horror. Tech optimists visualize a future where generative AI takes care of dull and time-consuming tasks through automation. Critics fear that there are far too many unexplored ethical implications, including mass unemployment.

Many smaller companies, which often struggle with recourses to carry out all the necessary operations, may embrace the optimistic perspective on generative AI. Reaping the fruits of generative AI will give many SMEs a massive competitive advantage. However, navigating the disruptive challenges can also prove to be critical for all SMEs.

This article explores generative AI’s positive and negative impact on SMEs and advises on how to use GPT-4 successfully.

Generative AI and the possibilities in automation

Generative AI has the potential to be a game-changer for businesses. That’s at least what ChatGPT replied when McKinsey asked it to write about the transformational effects of generative AI for businesses. ChatGPT continued that generative AI “has the power to revolutionize industries and transform the way companies operate”.

Four months after McKinsey’s prompt, with GPT-4 and Google’s Bard on the market, many SME workers and owners will argue that generative AI already has transformational effects.

An area where GPT-4 and other generative AIs already have an impact is within a broad specter of automation. GPT-4’s capabilities extend beyond text generation, making it a versatile tool for automating various business processes.

For instance, it can streamline customer service by handling inquiries, enabling companies to help their clients more effectively. Additionally, GPT-4 can aid in marketing copy and social media content creation, reducing the time and effort of human teams.

Further, by automating data analysis and insights generation, GPT-4 can empower decision-makers with valuable information. By enabling the automation of many tasks that humans previously did, generative AI has the potential to increase efficiency and productivity, reduce costs, and open up new growth opportunities. A mix of advantages that many business owners dream of. Combined, effective automation can free up economic and human recourses that pave the way for innovations.

Innovation and AI

Freeing up resources give space for innovation – here, GPT-4 can also play a role. The chatbot can help businesses identify new market opportunities, develop new products, and enhance existing services by providing valuable insights and streamlining decision-making processes.

If you doubt GPT-4’s ability to innovate, just ask it to give you a new business idea. If you want to be impressed, try to ask it for a business idea within a very particular field. If you want GPT-4 to be helpful for your SME, tell it about your company and ask which products you should add to your product line.

The generative AI revolution holds exciting potential for SMEs, but there are also reasons to be cautious and take necessary measures to mitigate potential risks.

New technologies, new challenges

When ChatGPT was released in late November 2022, it became the software with the fastest global adoption rate ever. Commentators have compared its impact with other technological breakthroughs like the web browser and the smartphone. Generative AI is a highly disruptive force. Some companies win on disruption; some see the negative consequences before they even start transitioning to new opportunities.

McKinsey writes that businesses “able to effectively leverage the technology are likely to gain a significant competitive advantage”. Here larger companies may hold an advantage over SMEs. The European Economic and Social Committee (EESC) warns that AI can result in growing inequality between micro, small, and medium-sized enterprises versus larger companies, which often have more recourses to invest in new technology and to recruit the necessary talent.

According to the EESC, AI-driven business models have a “winner takes all dynamic” that’s highly disruptive and allows the winners to acquire competitors quickly. Knowing how to make AI a competitive advantage – and how competitors do the same – might define which SMEs will prosper in a world impacted by AI. Here are four practical pieces of advice on how SMEs can benefit from GPT-4.

How to benefit from GPT-4 and other generative AIs

There are several benefits to those who successfully implement an ‘AI colleague’. They are cost-efficient, don’t need breaks, and they have a lot of skills. They are also very dependent on good and ethical use by their human collaborators. Here are four smart rules to leverage the potential of GPT-4 and other generative AIs.

Learn how to prompt

Learning how to prompt is key to benefiting from generative AI. To those new to GPT-4, a prompt is the input you give the chatbot. The answer you get is its interpretation of your prompt. If the answer is irrelevant to your query, it could be due to the teething problems that novel technology has, but a better prompt would have likely resulted in a better response.

According to Zappier, good prompts are given by 1) Offering context, 2) Including helpful information upfront, 3) Giving examples, 4) Telling it [GPT-4] the length of response you want, and 5) Defining the expected formats. 

Additionally, using temperature is a great variable to play around with. Temperature defines how creative GPT-4 is, with a low temperature being less creative but often more accurate and a higher temperature giving more creative responses, oftentimes with less accuracy.

As Zappier recommends, giving context is essential to getting good results.

Learn how to give context

If you want to use GPT-4 for automation, teach it as much as possible about the processes that need to be automated. It’s beneficial to give knowledge about the company, its employees, its products, and its challenges.

Not sure what context to give? A great way to give context is to ask GPT-4 to ask 20 questions it needs to know about your business to help you automate or write content for you. Once you’ve answered them, the chatbot’s responses will be much closer to the reality of your business.

Using the same conversation session – the same tab on the left – on GPT-4 will improve the chatbot’s recall of the context. As the levels of information increase, giving a recall might be necessary.

GPT-4 is also great at summarizing, so after giving it the necessary context, a smart trick is to make GPT-4 summarize the context you’ve given it, save it in a document, and add it again when needed. With the relevant context in mind, GPT-4 will provide advice tailored to the specific needs of a business.

Often, the imagination of the applicability of regenerative AI is more limited than the possibilities of applying it. After giving context, it’s a great sparring partner for ideas and innovation.

Use GPT-4 to idea generate and draft

While GPT-4 is great for many things, humans beat it in quite some areas too. One of them is editing. While you can give it an understanding of your brand’s identity and make it try to communicate as your brand would, it might not get it just right.

Think of generative AI as a sparring partner. Like any other sparring partner, you might not take any advice it hands you. Instead, take inspiration and then give things your own spin.

The ideas, drafts, or input it generates will need edits to ensure it doesn’t feel robotic and stiff. An approach that is equally curious and critical is great. And don’t forget to fact-check.

Use GPT-4 to automate routine tasks

Routine tasks are a part of the daily tasks for all SMEs. It’s also something very few employees or owners appreciate doing. Here, generative AI can free considerable recourses.

Begin by evaluating your business processes to pinpoint tasks that are time-consuming, repetitive, or prone to human error. Examples include customer support, marketing content creation, data analysis, and administrative tasks like email management and scheduling.

After a use-case evaluation, explore the various GPT-4 APIs and tools available to integrate into your existing systems. Because GPT-4 is open source, the number of services integrating the AI is almost limitless. If there’s a demand, the integration likely exists.

It’s a good idea to run tests to explore how well a specific area can be automated before scaling up. If GPT-4 is used for customer support, automation shouldn’t come at the expense of good customer relations. After gathering feedback from stakeholders and employees, a large-scale implementation can be made to reap the benefits.

Continuous monitoring and staying up to date with developments in AI are essential for good automation in the long run.

Great use cases of GPT-4 differ from SME to SME. However, exploring and mapping how generative AI can benefit your business is the first good step for any SME. The recommendation to start experimenting with GPT-4 can hardly be overstated.