Forjance
Book a walkthrough
Questions owners ask

Can AI really help my business?

By Forjance8 min read

Six recent surveys and reports, the questions small business owners keep asking about AI, and what Forjance is doing for businesses right now.

Download the full article (PDF, opens in a new tab)

Short answer: yes, for most owners who have tried it. The longer answer is that it depends on where you start and who stays in charge.

We read six recent surveys of small business owners, and we read them the way an owner reads them: not for the percentages, but for what is going on behind them. A note up front: Forjance sells private AI teams, so we have a stake in the answer. Every number below is attributed, so you can check it yourself. Where we are giving our own read, we say so.

What the research says

  • 70%of owners using AI report a net-positive effect on their financesHeartland Forward
  • 93%of owners using AI say it saves them timeHeartland Forward
  • 89%of AI adopters report a positive business impact (just 0.4% report anything negative)Homebase
  • 76%of owners using AI say their business performed betterBizBuySell, via Small Business Trends

Different surveys count different things, so the numbers don’t line up exactly. They point the same way: most owners who use AI say it helps. In the U.S. Chamber’s report, AI users are at least 7 points more likely to report sales, profit and workforce growth.

Trying AI is easy. Getting it to carry real work is the hard part.

Use AI76%
Have fully embedded it in core operations14%
Goldman Sachs 10,000 Small Businesses Voices survey, of participants in Goldman Sachs’s small business program

Our take: that gap is the real story. Trying AI is easy. Getting it to carry real work, safely, is where owners say they want help: 73% of the owners in the Goldman Sachs survey say more training and implementation resources would help. So the useful question isn’t whether AI works. It’s what is holding owners back from letting it do real work. Six questions come up again and again. Here is what is behind each one, and the answer we would give.

The six questions behind the numbers

Question 1 of 6

“Where do I even start?”

  • 12%of owners said they don’t know where to startBluehost
  • 20%of decision-makers not using AI have looked into specific toolsHomebase
  • 83%of owners who use AI taught themselves by trying itHeartland Forward
  • 12%of owners who use AI learned through a class or training programHeartland Forward
What is really going on (our read)

Owners aren’t stuck on whether AI works. They’re stuck on the first step, because nobody hands them one. The market offers an endless list of tools, and none of them are about your business. Most owners who got going taught themselves, by trial and error. That is a lot to ask of someone already running a company.

The answer

Start with one task, not a platform. Pick something you have been putting off: the follow-up list, a page on your site that is out of date, the invoices nobody has chased. A good first task is small, visible and yours, which means you can judge the result, because you know what good looks like. That is how every Forjance walkthrough begins. You bring one task, and we show you what the team produced from your own material, so you see the work on your own business before you decide anything.

Question 2 of 6

“Is my data safe?”

  • 51%of all respondents named data privacy and security as a concernHomebase
  • 38%of decision-makers not using AI gave it as their top reasonHomebase
  • 37%of those non-users said knowing their data is secure would move themHomebase
  • 16%of owners name it as their top concernBluehost
What is really going on (our read)

Your customer list, your pricing, your invoices and the notes on your difficult accounts are the business. When an owner asks about data safety, the real questions are: where does my information go, and who else can see it or learn from it? Those are fair questions, and they deserve plain answers.

The answer

Ask any provider three things: where does my data live, does it train anyone else’s product, and can I see what was done with it? Here is how we answer them. Each client’s AI team works only for that business, on its own private server. Nothing you feed it trains anyone else’s product, and nothing about your business is shared with another client. Every action it takes lands in a ledger you can open.

Question 3 of 6

“Can I trust what it writes?”

  • 42%of all respondents worry about mistakesHomebase
  • 32%name accuracy and quality as their top concernBluehost
  • 6%of owners highly trust AI with their brand voiceBluehost
  • 31%of non-users said knowing a person stays in control and can override it would move themHomebase
What is really going on (our read)

Notice that nobody is asking for perfection. They are asking for control. Your name goes on whatever gets sent. A wrong price or an off-tone email to a customer costs you something, and you are the one who answers for it. The worry isn’t that AI will be bad at writing. It’s that something goes out in your name that you never saw.

The answer

Treat it the way you would treat a capable new hire in their first month: it learns your work, and you review what leaves the building. In practice that means you set what the team handles on its own, what it drafts for your approval, and what it only flags for a person. Nothing gets published, sent to a customer, or paid without the level of approval you chose. Other AI systems in the team check the work before it reaches your approval queue, so what you review has already been looked at once. And because the team learns your services, your tone, your tools and your rules, drafts start from your voice rather than a generic one.

Question 4 of 6

“What will it cost?”

  • 5%of owners named cost as their top concernBluehost
  • 6%of owners not using AI cited cost as a reason (another 6% cited complexity)Heartland Forward
  • $41 to $150the most common monthly AI spend among adoptersHomebase
  • 13%pay nothing extra, using AI features already inside tools they haveHomebase
What is really going on (our read)

Those dollar amounts are real, and they are low. Here is what they buy, in our read. At $41 to $150 a month, or nothing at all, an owner is typically using a shared AI chat assistant, or the AI features built into software they already pay for. That is a useful tool. It is also a different thing from an AI team set up around your business, with your rules, your files and your tools, doing the work itself. So don’t be surprised that a private team is priced in a different range. The fair comparison isn’t tool to tool. It’s what the work costs you today. And cost is rarely what stops an owner. What stops them is not knowing what they would get for the money, so any price feels like a guess. Owners who use AI are not describing a cost story: in Heartland’s survey, 93% say it saves them time, and more than half save five or more hours a week.

The answer

Judge any price per task, not per tool. A chat assistant gives you answers that you still have to turn into finished work. A team gives you the finished work, checked, and waiting for your approval. So write down what one task costs you in hours today, and that is the number to beat. A first task you can judge in one sitting settles the question faster than any price list, which is another reason to start small.

Question 5 of 6

“Will it replace my people?”

  • 87%say AI augments rather than replaces employeesGoldman Sachs 10,000 Small Businesses survey (program participants)
  • 81%of AI users say it lets them do more with the same staffHeartland Forward
  • 51%of owners using AI say it has not changed their staffing levelsBizBuySell, via Small Business Trends
  • 49%expect AI to help their business growHomebase
  • 22%expect it to reduce the number of employees they need for certain tasksHomebase
  • 1 in 5say AI-driven growth helped them afford to hireHeartland Forward
What is really going on (our read)

This is the question owners ask quietly, because it is about people they know. The numbers lean toward helping, not replacing. The work that moves to AI is the work nobody was hired to enjoy: answering the same lead questions, updating the site, chasing unpaid invoices.

The answer

Build it for a business with staff, and let everyone on your team use it. When the routine work comes off your people’s plates, they spend that time on the work only they can do, such as the customer who needs a phone call, the job that needs judgment, the idea that needs a person. The owner of one software company that runs on the team puts it this way: “I approve the work. I stopped having to do the work.”

Question 6 of 6

“Is this for a business like mine?”

  • 52%of owners not using AI say they haven’t seen a need for itHeartland Forward
  • 54%use AI among businesses under $500,000 in annual revenueHomebase
  • 81%use AI among businesses at $5 million and upHomebase
What is really going on (our read)

That 52% is not saying AI is useless. They are saying nobody has shown them their own business in it. Smaller shops trail, we suspect, because the examples they see are written for bigger companies with IT departments.

The answer

See it on a task from your own business. A trade company, a consulting practice and a cleaning company have different tasks, but the pattern is the same: one task, kept private, with a person approving. The next section shows what that looks like in different kinds of business.

Real work, by the kind of business

Here are some examples of where Forjance AI teams are already at work.

  • A software company

    The team carries a product that ships to two app stores, a live web app, the product’s marketing, and its support desk. It wrote and published a complete illustrated help center, 22 articles and 27 screenshots, overnight. When a screen freeze was reported overnight, it was root-caused, fixed and verified on a real device within the hour.

  • A provider-data software company

    The team took its app through a major marketplace’s formal review, fixing every reviewer item across two rounds, and had the sales materials and meeting briefs ready.

  • A new product launch

    The team went from an idea to a live site with outreach running in days, not months. It researched the market, wrote the copy and built the site.

  • A consulting firm

    The team redesigned the firm’s full website and built its library of articles into it, each with its own page and download.

  • A home-technology company

    The team designed two complete versions of the company’s website. It also read the architectural, electrical and lighting drawings for a residence and prepared a systems proposal, with every item’s room noted, covering lighting control, motorized shading, network and security cameras, and voice control. The proposal has itemized equipment tables with quantities and part numbers, labels where every quantity came from (the scope of work, the drawings, the manufacturer’s own materials, or trade convention), and lists what still needs confirming before pricing.

  • Small cleaning companies

    For one of them, the team built a fifteen-page website from the owners’ own brief and flyer, with prices shown only once the owners confirmed them. There, a customer’s text becomes a drafted reply, and a plain reply goes out once the owners approve it with one tap. The team keeps the accounting records in step and sends the owners a weekly activity report. The two companies are kept apart, with nothing shared between them.

Our take: none of this means AI is magic or that every business gets the same result. It means the pattern in the research, start small, keep it private, keep a person in charge, is something we do every day, and you can see it working on your own task before you commit to anything.

Sources