EPISODE 111: How are small businesses using AI right now?

Lindsay Beaton, editor, Petfood Industry magazine and host, Trending: Pet Food podcast: Hello, and welcome to Trending: Pet Food, the industry podcast where we cover all the latest hot topics and trends in pet food. I'm your host and editor of Petfood Industry magazine, Lindsay Beaton, and I'm here today with Eric Thornton, eTailPet lead for AI and emerging technologies. Hi, Eric, and welcome!

Eric Thornton, eTailPet lead for AI and emerging technologies: Hi, thanks for having me, Lindsay.

Beaton: In case you're unfamiliar with Eric or eTailPet, here's what you need to know. Eric Thornton brings over a decade of experience in retail software, with a career rooted in product management and software development. He has designed and delivered AI-powered tools across multiple retail sectors and has been a featured presenter on the practical applications of AI in retail.

eTailPet is the all-in-one point of sale (POS) solution created by a pet business owner and built specifically for independent pet retailers. This cloud-based technology empowers independent pet businesses with robust software solutions to compete in the growing digital marketplace. eTailPet simplifies inventory management, provides detailed reporting, and boasts online sales and marketing tools that help specialty retailers focus on growing their business.

Eric's extensive knowledge of AI and how businesses are using it in the marketplace makes him the perfect person to bring on today to answer this question: How are small and medium-sized businesses currently using AI in the workplace?

Eric, I want to start with a question that might be more complex than I intended, but I want to know what you think the ratio is of small businesses to medium or larger businesses currently making use of AI, real-world speaking. Who jumped on this first?

Thornton: There are multiple angles to look at this from. There is the traditional approach — open up ChatGPT, open up Claude, ask a question and get an answer, very similar to what we might have done with a Google search previously. Then there's the level of agentic, autonomous AI systems running and doing things for you. On the first approach, everyone — or at least the majority of people — seems to be using it. They're chatting with AI applications, a basic large language model: ask a question, get a response.

The latter is much less adopted, and the ratio is fairly similar across both small and large businesses. Larger businesses may have more motivation because they're paying attention to competitors and have to answer for revenue and P&L, so they might be more aware of the need for autonomous AI agents running in their operation. But because they are larger, it's just harder to move and get those things going.

On the small business front, people are wearing so many hats, and in many cases it's not just one business they're running — they've got multiple operations going. The ratio of people using AI as more than a chatbot is fairly small at this point, simply because it takes work, and things are changing so fast. If you felt like you were really up to speed on AI three months ago, you are missing out on many things that early adopters are finding a ton of advantage with. A pretty similar ratio is using it across both sizes — it really comes down to who is excited about AI and wants to figure out how it can be helpful.

Beaton: With the way it's changing so quickly right now, do early adopters have an advantage just because they got in on the ground floor, or would somebody jumping in right now — without having to unlearn what's already an old way of doing things — have more of an advantage? Who has the bigger challenge?

Thornton: The early adopter has the advantage. When ChatGPT really blew up, how much the model hallucinated is so different from what it can do today — but those who started then have carried a lot of foundational knowledge forward. Can you write a quality prompt? Do you know what you're actually asking? Even though prompting techniques have changed, prompting is still a thing. 

Whenever a new tool comes out, early adopters have learned so much foundational material that it's more additive than starting from scratch. The unlearning is small in comparison to being able to wrap your mind around what AI can actually do. If you say, "It doesn't do exactly what I want, so I'll wait," I think it will be very difficult to catch up.

Beaton: What are some of the common struggles right now that a small or medium-sized business has that AI is uniquely equipped to solve — things that a larger business with more resources might not even be looking at?

Thornton: When you have a smaller business, it's much easier to have a pulse on what's going on. AI has improved on accuracy over time, but it is still not 100% accurate — and neither are human beings working in a business. Small businesses can use AI to get directional help and then fill in the gaps, because they know their business well.

Take reporting as an example. When you run a report, you're generally looking for 100% accuracy — a retail store isn't going to submit a revenue figure to the IRS because it was 95% correct. But as a small business, you can take your reports and use AI to get a directional read on how things are selling across vendors, which products seem to be performing well. You don't have to do a lot of training to get that general direction. Upload a CSV export to an AI tool, ask some general questions, and you can get there. If you're looking for an exact count of how many units of a specific item sold, you might be disappointed — but for directional analysis, AI is genuinely useful.

Another area where AI is well-equipped is image recognition. Every retailer deals with purchase and receiving orders from multiple vendors, and every vendor has their own format — one calls it SKU, one calls it manufacturer ID, one just calls it ID. AI has gotten very good at parsing those differences. Being able to upload a PDF or a CSV from a vendor and have AI help you receive inventory and put it into a system is a real time saver.

Beaton: I've had eTailPet on the podcast a few times, and I know that what you deal with is a lot of the minutia of business — inventory management, reporting, all of those things that aren't exactly thrilling to do. When you first started looking at AI, was there anything that made you think, "This is going to change the game" — something that would allow people to take things to the next level or make things easier for your customers?

Thornton: The most straightforward win we saw was around content. We have a messaging assistant within the eTailPet platform that allows you to two-way text your customers. You might get a challenging text message from a customer and, rather than composing a response from scratch, you can click a button to generate an AI response that draws on the conversation context, then tweak and edit it before sending.

Beyond content, we've had a lot of thoughts on what we can do with AI — but our system needs to be very dependable. When we release something, retailers expect it to work. They don't want to be at the register with a customer ready to walk out the door and have numbers that don't add up. That does slow us down on releasing some things, but it's intentional. We want what we give to retailers to help them, not hurt them.

Beaton: I find it refreshing to hear that, because right now there's kind of an all-or-nothing approach to AI — people are either completely in, or they'd rather not touch it at all, and there's not a lot of conversation about the nuance in the middle. It's interesting to hear from a company that has every reason to integrate AI for efficiency, but is still stepping back and saying, "Here's what we want — we're not sure it's fully there yet, so let's keep working on it." Is that something that continues to be intentional on your part?

Thornton: Yes. Going back to the earlier point — chat applications are the main way people consume AI right now. We could throw a chat interface into our software and say, "Great, we've got AI." But is a retailer going to stop and type something out while they've got a customer standing in front of them? We have to figure out ways to make AI work really well for our retailers in an intuitive way that keeps speed — so they can keep moving. We don't want to release things just for the sake of releasing them.

There's also a piece to AI that I think pet retailers are uniquely positioned for. It's a very general tool — it gives generic responses by default. But retailers who know the pet space can guide AI much more effectively. A good example: CBD terms are flagged as prohibited in text messaging platforms. If you're not in the pet space, you'd think there's no CBD in pet retail — but anyone in the industry knows there are CBD dog treats and similar products. When you know the unique things about the space, you can take a good AI response and make it great.

Beaton: As you're introducing customers to AI or helping them with questions, are you seeing any patterns in who's most accepting of integrating it, by demographic, business size, or anything else?

Thornton: The most accepting are the ones with the right attitude, and I've seen that across all demographics. The people who struggle are the ones who try AI, find it doesn't do exactly what they expected, and conclude that it doesn't work. Everyone I've seen turn the corner has the attitude of: "There is something here, and I'll figure out a way to make it work." That's when the light bulb goes on, and it can become genuinely addicting.

One example: I was talking to someone outside of work who runs a small coffee shop chain — older demographic, 30 to 40 years in retail. They started using chat applications and just committed to figuring it out. Now they're using AI to help architect new buildings they're constructing. They're not architecture experts, but they can get the feedback they need to guide their conversations with architects. They're using it to get general legal guidance before talking with their legal team. Attitude is what changes things most.

There's also the challenge of, "I'm buying into AI, but nobody else at my company is. What do I do?" You have to evangelize for it. You have to find something that's meaningful to the person you're trying to bring along — a use case that matters to them — and get them to experience it firsthand. You can be persistent, consistent and clear that you're a believer, but it won't stick until they've had that experience themselves.

Beaton: For people trying to make that happen, how long does it take? Do you have an example of a particularly resistant client you were eventually able to win over? Is it a matter of finding the right use case, or is there a longer warm-up process?

Thornton: There is a warm-up process, and it varies from person to person. A lot of it depends on the day you catch them. You might have a great use case to share, but if they've got 50 other things on their plate, it's not going to stick. If you have regular contact with someone, the turnaround is much quicker. If it's someone you work with once a month, it takes longer.

Everyone I've seen come around to loving it — including my wife — had to experience it in their own context. For her, it was gardening. Now she's asking questions about seasonal changes, about what adjustments to make for her region. You can't just read about AI and get it. You have to use it.

Beaton: You mention gardening — I have friends who used AI to help plan out a large garden plot in a way that maximized how all their crops worked together.

Thornton: Very cool.

Beaton: What are the challenges right now — not with getting people to use AI, but with companies that are already using it? Are there cases where people are fully bought in but not sure they're using it correctly? Are there control issues where they feel they should hold on to something even though AI could help? What are the challenges for people who are already in?

Thornton: I think a lot about Clayton Christensen and disruptive innovation, and I think small businesses are well-positioned here. Let's say a small business figures out how to run their ordering process more efficiently using AI — they take a CSV export of sales data, run it through an analysis, and get what they need to place orders. A small company can identify how that works and have it implemented the same day. They walk into the next room and say, "Hey, we're going to do this. Here's how. Upload the CSV here and do this."

A large company has processes that take much longer to change. The regional manager has to meet with store managers — and that meeting is once a month. Then the store managers have to train their teams. It's the same principle as a large ship taking much longer to change course than a small one.

There are also data and compliance considerations that grow with company size. Larger companies can't rely on informal trust — they need systems that prevent problems, like employees uploading sensitive CSVs to outside tools. Small businesses aren't exempt from those risks, but they can move much faster.

Beaton: It's interesting that you bring up how nimble small businesses can be, because that message comes up constantly when I cover business strategy. Yes, small businesses have fewer resources — but in the early years especially, that nimbleness is a real advantage. A scrappy team of three can make a leap forward that takes a larger company years to execute.

Thornton: Absolutely. And going back to an earlier point — the unique knowledge small businesses have in the pet space is extremely valuable. These AI tools are generic. A large company has to figure out how to make AI work within their organization and then layer in pet industry knowledge on top of all their institutional knowledge. A small pet retailer who knows their customers and knows the pet space? That specialized knowledge is what they're working with. There's much less to untangle.

Beaton: Regular listeners know I like to wrap things up by looking to the future, and AI is particularly interesting in that regard given how much is in flux. Based on what you're seeing right now, what do you think is the most likely path forward for AI in business — especially in the pet space?

Thornton: It is not going away, that's for certain, and it will keep making major gains. When I think back to other major technology disruptions — like personal computers — it's hard to imagine how business was run before them. Now it's hard to hold almost any job without using a computer or a phone in some way.

AI will be very similar. The analogy I keep coming back to is Excel and accounting. Before spreadsheet software, people were handwriting general ledger statements — something nobody wants to do now. But new opportunities also emerged: who can build a model in Excel? It didn't just do everything automatically. Someone still had to have the specialized knowledge to use it well.

I think AI will follow that same path. It won't be a matter of simply telling AI to do something and walking away. Someone still needs to have the specialized knowledge to direct it, connect pieces, and make good decisions with what it produces. There will be a lot of cool things it can do, but there will also be real opportunities for the people who know how to work with it. It's going to keep moving forward like a train — and you definitely need to embrace it.

Beaton: Thank you so much for coming on today to talk about this, Eric. Being able to look at the different facets of how AI is being implemented in business is genuinely valuable, especially given how quickly everything is moving — whatever was new three months ago might as well be ancient history. That makes it even more important to keep having these conversations, so people can at least try to stay on top of things and understand how a tool that is so generic can be applied in so many different ways. Thank you for that.

Thornton: Thanks for having me. There are so many cool things happening in this space. Try something rather than sitting and waiting — that's the message.

Beaton: Before we go, where can people find more information about you and eTailPet?

Thornton: The best places are our website, etailpet.com, and LinkedIn — both the eTailPet company page and the personal profiles of our team, including myself.

Beaton: That's it for this episode of Trending: Pet Food. You can find us on petfoodindustry.com, SoundCloud or your favorite podcast platform. You can also follow us on Instagram at Trending Pet Food Podcast. If you want to chat or have any feedback, feel free to drop me an email: [email protected]. Once again, I'm Lindsay Beaton, your host and editor of Petfood Industry magazine, and we'll talk to you next time. Thanks for tuning in!