Follow ZDNET: Add us as a preferred source on Google.
Are you thinking about investing in AI for your small company? The buzz about AI seems to be everywhere, but is there real value, especially for small businesses? For me, the answer is a definite yes, as long as you're careful about how you approach it. Over the past nearly three years, I have found AI to be a force multiplier for my small business.
To help you see how AI might help your business, I'm going to show you many of the ways AI has helped my business. All told, across all three years since ChatGPT hit the market, I've probably spent less than $1,000. And, $200 of that was in one big sprint that netted me four complete new products. So yeah. Worth it.
Our small business
My wife and I own a small content business. We have a variety of different brands and reach a bunch of different target markets, but the common factor is that we produce and sell digital content. The work itself is similar across all of our business areas, and includes writing, web development, coding, video production, marketing, analytics, and relationship management.
Also: The moment I realized ChatGPT Plus was a game-changer for my business
Keep in mind that we do all this as a two-person business, run out of our house. Although we used to have employees, I didn't enjoy being a boss, and we have no plans to hire. That's important because our use of AI isn't taking someone else's job. What it is doing is enabling the two of us to do more in the available time we have.
I'm going to spend a few minutes discussing the areas we're involved in, as that will help set the context for how the AI features we use have been put to work. Odds are, you are involved in one or more of these areas yourself.
For each work activity, I'll enumerate the tasks we've delegated or collaborated with the AIs to accomplish, for a total of 21 ways we've benefited. I'll also point out the tools we've been using for each of these productivity-boosting tasks.
Let's dig in.
Software products
For the past decade, we've produced and sold freemium software products. At one point, we had ten unique core products, including a donation product that facilitated over $40 million worth of donations during our stewardship.
A few years ago, we decided we wanted to focus more on the top-tier product or products. But I didn't really know how users felt. When someone uninstalls one of our products, they're given an option to fill out a sentiment survey form, but we never had a way to process the data. When we exported all that data and gave it to ChatGPT Plus for processing, we had our answers.
Also: I got 4 years of product development done in 4 days for $200, and I'm still stunned
We sold nine of our products in order to focus on our core security product and its add-ons, which protect more than 20,000 active sites.
It's probably fair to say that this is where AI has provided its most astonishing return on investment. This is also where we spent $200 for a month of ChatGPT Pro. But there are caveats. I am a former computer science professor, so I know my way around code.
Also: 10 ChatGPT Codex secrets I only learned after 60 hours of pair programming with it
Do not assume that if you're a newbie to coding, you can just vibe code your way to new products solely through the magic of AI. Coding AIs require intense and focused management. You wouldn't hire a bunch of expensive contractors without knowing something about the work they need to do. Don't do it with code.
Ways we benefited from AI:
1. Sentiment analysis: We fed in all the survey results from product uninstalls and got back a detailed overview of our products. Users generally liked them, but there were standouts we would not have noticed without the AI.
- Tool used: ChatGPT Plus
2. New software products: Read this and this, then this. They'll show you all about my foray into using AI to crank out four new, powerful software products in less than a week. But keep in mind that after the AI was done, I had weeks of old-school product marketing to do myself.
- Tools used: ChatGPT Plus and ChatGPT Pro.
3. Marketing overview white papers: To save time, I linked the AI to my code repository and asked it to create briefing documents that simulated the document a development team would give to a marketing team about features and benefits. This became the white paper "bibles" that defined the new software.
- Tool used: ChatGPT Pro Deep Research
4. Marketing video overviews: Those marketing overviews were then fed into the AI, which produced a narrated video overview for each of the products. I prepended those overviews to a demo I did using screen capture, but the overviews recorded by the AI were incredibly well done, although the imagery could have been better.
- Tool used: Google's NotebookLM
5. Product logo and look: Midjourney and Photoshop were used to create a new, updated look for the support product. It took a bunch of tries, but it turned out quite nice.
- Tools used: Midjourney and Photoshop Generative Expand
YouTube and newsletter
We produce a popular DIY YouTube channel with about 65,000 subscribers, along with a newsletter that reaches readers and viewers.
Also: These 3 AI tools made my two-minute how-to video way more fun and engaging
Although AI helps out in many ways, that's not apparent to viewers and readers. We have a fleet of robot cameras (because, of course, we do) that help take video while doing projects. The projects themselves usually involve some kind of robotic fabrication tool like a 3D printer, CNC, or laser cutter.
Also: How I started my own LinkedIn newsletter for free - in 5 easy steps
But beyond the physical tools, we use some AI features to prepare the videos themselves. Ways we benefited from AI:
6. Dynamic masking: I sometimes separate myself from the background of a video for a particular effect. Doing this by hand is a brutally tedious frame-by-frame task. Last year, Final Cut Pro (my editor of choice) introduced Magnetic Mask, which tracks a subject (well enough) and provides the dynamic masking with only one or two clicks.
- Tool used: Final Cut Magnetic Mask
7. Extending photos: Creating click-worthy thumbnails is everything on YouTube when it comes to getting views. In order to accommodate the visual elements of the thumbnail, and still show an image, we often move the photo to the right. That's often past the edge of the frame. Adobe's Generative Fill comes in clutch, filling in the empty space and making the thumbnail complete.
- Tool used: Photoshop Generative Fill
8. Headline and SEO optimization for YouTube: We use the AI in the Mac's Voice Memo app to transcribe videos and feed the results into ChatGPT. ChatGPT analyzes the content and helps narrow down headline options, taking SEO considerations in mind.
- Tool used: Voice Memo and ChatGPT
Downloadable plans and patterns, and video tutorials
My wife runs the branch of the business that focuses on a very popular craft-related hobby. That business sells plans and patterns for folks in the hobby, and every download is accompanied by a detailed video tutorial that only buyers can see. There's also a very active Facebook community that revolves around these product offerings.
Also: I asked ChatGPT to write a WordPress plugin I needed. It did it in less than 5 minutes
This was one of our first uses of modern generative AI for productivity. My wife runs a monthly giveaway where she uses a randomized list of names, but instead of de-duping the names, names intentionally get repeated if users contribute craft photos for more than one project. She wanted a tool that would randomize a list so that multiple instances of the same person's name wouldn't be clumped together in one location on the wheel.
On a whim (this was way back in February 2023), I decided to see if ChatGPT could do it. We now know how well AIs can code, but back then, it was astonishing.
We also use AI graphics tools to create an image that's evocative of the month's project theme. When I did this by hand with Photoshop, it could take me a few hours to mock up a mediocre image. With AI, it takes me about 10 minutes to produce three or four images for her to choose from.
Ways we benefited from AI:
9. Custom programming: Creating a custom WordPress plugin to solve a specific business need.
- Tool used: ChatGPT free tier
10. Image generation: Create a monthly project theme image evocative of the craft focus of my wife's e-commerce site.
- Tools used: Midjourney and ChatGPT Plus
Print on demand
We're running an Etsy shop that sells print-on-demand products. We're still tinkering with it, and it's not a major push, but it's part of our portfolio.
Also: How I used ChatGPT and AI art tools to launch my Etsy business fast
We licensed a set of images from a human artist for some of our merch. I created some of the text-only designs myself. But we also enlisted AI to create an extensive array of images for the shop's logo, design some of the merch, and even used an AI mockup tool to place some of our images on people for the store photos.
We also created a set of characters and identities for both the images and the store itself. While it was a lot of fun, I'm not sure it was worth the effort, even if the AI helped a lot by brainstorming character names and stories.
Ways we benefited from AI:
11. Character creation: AI was used to create images of characters for our merch, brainstorm names, and write backstories.
- Tools used: Midjourney and ChatGPT free tier
12. Mockup creation: AI was used to composite our designs on photos of real human models, and make the T-shirt mockups look real.
- Tool used: Placeit
Music production and distribution
My music is distributed through Spotify, Amazon, Apple Music, and all the other big music distributors. Some months, the number of listens is low. But we've reached as high as 16,000 listeners in one month.
Also: How AI helped get my music on all the major streaming services
I used AI extensively in preparing to distribute my two EPs. That said, and I post this on the music's website, "No AI was used in the creation or mixing of this music." I never use AI for the actual content. But I do use it a lot in support of that content.
Ways we benefited from AI:
13. Images and graphics: Album cover art, artist profile image for Spotify, various supporting artist images.
- Tools used: Midjourney, Adobe Photoshop
14. Promotional prose: Descriptions of the music and artist music bio for the streaming services.
- Tool used: ChatGPT free tier
Strategy and advice
I write primarily for ZDNET, although I also provide strategic advice to members of the military and government agencies on topics primarily related to counterterrorism, cybersecurity, and AI policy and practice.
Because I write about AI for ZDNET, I pretty much use everything. To be clear, though, I don't use AI to write any of my articles, unless I'm specifically showcasing what an AI produces for a given article.
Ways we benefited from AI:
15. AI testing: I test chatbots, image generators, video generators, AI detectors, coding tools, and so many other AI-related products. This isn't something most small business owners would do, but almost all of these tools have a free trial, so if you're curious about a capability, give it a test.
- Tools used: All of them
Publishing
For years, we produced online magazines, but retired from that a few years back. That operation was actually fueled by early generative AI code I wrote to help manage the business, a decade before today's large language models like ChatGPT. Today, we publish a few books distributed through Amazon.
But even though that generative AI was custom-built, it could give you some inspiration for how you might use today's tools in your business. Ways we benefited from AI:
16. Industry-relevant news summaries: One of the AI systems I built tracked news related to the subject of each of our journals, and then displayed a summary of the relevant news daily on each journal's web site. You can do the same thing using AI agents feeding data to your site or sites.
- Tools you can use: AI agents like ChatGPT Agent
17. Content analysis: I also built a tool that would analyze each full article we published for topics and subject matter. It would then feed that information into the next process (described in the bullet below) for affiliate revenue.
- Tools you can use: You could use NotebookLM to ingest an article or group of articles and give back data, which you could then use with a tool like Zapier or Hubspot to pass along to additional AI systems.
18. Affiliate and cross-sell marketing: I used the data pulled from each article to guide my product picker AI. This tool picked complementary products from Amazon that could be cross-sold in each article, which resulted in a small stream of affiliate revenue.
IT activities
One of the reasons we're able to participate in all of these business areas is that actual work done is similar across them.
Also: I got ChatGPT Agent to deliver a truly useful report - and then NotebookLM took it to the next level
For example, the websites look different from business unit to business unit, but the actual website management uses the same skills. Likewise, we have one file server that handles all our important storage and video backups, whether it's for the crafting e-commerce site's video tutorials or the YouTube channel.
Here, too, we benefited from AI. Ways we benefited from AI:
19. Researching storage services: When Google once again pushed back on the huge amount of storage we used on their services, I used ChatGPT Agent to go out and research alternative storage options. I got back a very in-depth analysis, which I then fed to Google's NotebookLM to quickly summarize and give me an audio briefing. This took something overwhelming and made it digestible.
- Tools used: ChatGPT Pro Agent and Google's NotebookLM
20. Replacing an obsolete software component: One of our sites gets a constant stream of fake user registrations. For years, I used a software tool that blocked that, but recently it got flagged by my hosting provider as vulnerable. I couldn't find another tool that seemed to do as well without tons of added junk I'd have to bypass. Since the tool was open source but no longer maintained I fed it into ChatGPT, telling it to identify its feature set and give me a report. In another session, I fed that report into Codex, and told that to produce secure code that provided those features. End-to-end, it was only two hours to a fully working solution that's been active for the last six months.
- Tools used: ChatGPT Deep Research and ChatGPT Codex
21. Configuring computer systems: When Pocket discontinued, I wanted to set up a local server on my home lab. I used ChatGPT a lot to get the proper Linux configuration set up and working. Saved a ton of time, even if it was fussy.
- Tool used: ChatGPT Plus
So, what tool should I use?
AI tools are just that: tools. Some are more general in nature, while others serve a very specific need.
Also: The best AI chatbots of 2025: I tested ChatGPT, Copilot, and others to find the top tools now
If you're just getting started, I recommend you check out general-purpose chatbots like ChatGPT, Microsoft's Copilot, or Google Gemini. They each have free tiers that have proven to be quite impressive.
For my small business, we only pay for four tools: ChatGPT, Midjourney, Photoshop, and Notion AI. I've been paying for Photoshop since way before the generative AI boom, so it doesn't even really count. I only bought Notion AI for a specific project, but they insist you buy a year of service.
Also: I'm an AI tools expert, and these are the 4 I pay for now (plus 2 I'm eyeing)
By the way, that's my trick with AIs. If you don't think you'll use the tool every day, just buy a month's worth of service. I spent $200 on ChatGPT Pro for one month, then downgraded it to the $20/month plan. So far, that's all I've needed to knock out some major upgrades.
A quick thought on time management
If you've never had your own small business, you probably don't realize just how many types of tasks small business owners have to do, and how much time juggling we do to get it all done. My wife and I are involved in a whole bunch of endeavors, but some are more side projects and some are our primary income sources.
The key is ruthlessly managing priorities, but also being super-consistent about giving everything even just a few minutes on a regular basis. You'd be amazed how much you can get done on a project if you just give it ten minutes a day.
In any case, beyond the fact that I test AIs for ZDNET, the reason we've used (and stuck with) AI in all these areas is because (at least mostly) using the various AIs has saved time. And any regular hour-long task we can trade for part of a $20/mo fee in order to get back 55 minutes is well worth the expense, especially when those expenses are split across tasks and months.
Given that we spend less on AI each month, across all the tools, than we do for our QuickBooks subscription, it seems well worth it.
What about you? Have you tried using AI to boost your own productivity or streamline your business? If you run a small company, where do you think AI could save you the most time? Is it coding, marketing, content creation, or day-to-day operations? Do you see tools like ChatGPT or Midjourney as worth the cost for small teams, or do you prefer sticking with traditional methods? How do you decide which tools are worth paying for and which are better to skip? Let us know in the comments below.
You can follow my day-to-day project updates on social media. Be sure to subscribe to my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.
English (US) · 
