Where AI Actually Helps Small Teams Work (And Where It Doesn’t)

In the past few years, AI has transformed the corporate world. Most companies are using it as a feedback mechanism, to complete menial tasks or, in some extreme cases, to replace entire departments. And no matter where our personal feelings align on this ever-emerging technology, it appears that AI is here for the long haul. 

So, what does that mean for entrepreneurs and small businesses? 

We understand the idea of using AI at work may lead to mixed feelings and questions. Is it safe? Will it make mistakes? Where do I even start?

As part of a small business team, it also took us a while to figure out where we might need AI’s input and where our brains and humanity should be prioritized. And truth be told, we are still working to find the right balance for us. 

So, while I’m still wrapping my head around some of its functionalities and toying the line between using too much AI or not leveraging it enough, I wanted to share some of the insights my team and I have gathered so far. Things that have been working well, and some that have gone wrong. 

What AI is actually doing inside small teams right now

The role of AI in an organization depends largely on how comfortable their leader is with using the technology. However, we have observed that most companies start the same way: getting a basic subscription to test out one of the main platforms (e.g. ChatGPT, Claude, or Gemini). 

(Stay tuned for our breakdown on which platform might be best for your unique needs; they each have their own specialties.)

The most straightforward way many companies are integrating AI into their workflow (maybe without even realizing it) is through automated meeting notes and summaries. There are a myriad of note-taking apps that are AI-powered. Using these wisely can help teams focus more on the actual conversation and how to respond appropriately. Using automated meeting note tools can also save  time and resources, because now, no one needs to join solely for the purpose of taking notes. That time is now freed up for more important tasks.

Another common task that companies are handing off to AI is content generation: using AI to draft articles, social media posts, or any written content the organization needs to create. To an extent, we do this too. We have tested using Claude in the past to draft communications or improve templates. But AI-produced content should never go straight out into the world without careful human editing. Once AI supports with the first pass, it’s important to dive in and make your changes, so it matches your voice, your brand, and the personal message you want to convey.

To clarify, we don’t use AI for all of our written communication. Our team is made up of writers and creatives, and to fully hand off this task would stifle our unique voices and skillsets. For example, our Insights and newsletters are fully generated by us, because we believe that this is a space where it’s important to communicate directly from our experience. (Plus, many of us find that there is an entire thought process that occurs while writing; we often think by writing, and this is lost when we lean on AI.) This is a line we have drawn as an organization after much thought and experimentation. 

And this is sort of the point of this message: It’s up to you and your organization to determine where the lines will be drawn when it comes to your company’s AI usage. But that can only happen if you’re informed on its capabilities and have experimented enough to form an opinion. 

Since we’re talking about drawing the line…

When is it a bad idea to use AI? 

As a rule of thumb, anything related to building human connection and relationships is probably best handled without the intervention of a computer. In the business world, the only thing that can truly differentiate us from other similar organizations is ourselves - our personalities, our wit, our experiences, and the authenticity we use to treat others. These are things that should never be replaced by a basic template or a fabricated interaction. 

Another important category to avoid is: judgement calls. AI shouldn’t be the one making the decisions for us or our company. We set the pace and invite it to follow, not the other way around. (However, I’ve found that if you’re feeling stuck, AI can be helpful with providing extensive research and organizing your pro-con lists.)

Everything else is murky territory, but it’s important to keep in mind what your company’s value proposition is and how it could risk getting lost by using artificial intelligence. For example, if the differentiator is being a consulting firm that values relationships and connecting human interaction to operational strategy that real humans will use, it’s important to keep that approach intact. 

Now that we’ve gotten the good, the bad, and the ugly out of the way, the question is… 

If I haven’t used AI for work yet, where can I get started? 

As an extremely organized individual, I have always found it easy to start by following a list. A simple sequence of steps guiding what I need to do next, so that I might achieve the final goal. In this case, that goal is to start experimenting with AI and integrating it into the daily processes of a company to maximize resources and growth. 

Here’s the list to get there:

• Pick one recurring or time-consuming task to hand off first (starting small is important to avoid getting overwhelmed). This might look like: asking AI to summarize a lengthy report; creating automated AI search for daily news; or having AI create a template for an Excel model that you don’t have time for.

• Treat AI output as a first draft, not a final answer (always double check every output, and tweak as necessary until it understands what it needs to do better; don’t be shy about telling your AI tool what it got wrong, as it learns your goals and tone, future outputs will improve).

• Write down what worked well (and what didn’t) and share it with your team, so everyone can benefit from your experience(maybe hold sessions where everyone shares a new discovery and consider implementing it in your workflows).

• Set an explicit boundary on what stays human (judgement calls, sensitive data, client-facing materials). 

The bottom line here is… 

AI should do the draining, repetitive work, so your team can do the work that matters

To clarify, this doesn’t mean throwing out how your team already works; it means being intentional about what gets handed off and what doesn’t. 

If you need a hand figuring that out, or want to have a conversation about your operational goals, , schedule a free 20-minute conversation or send us a message. We look forward to connecting human to human.