Picture a gift shop with one owner and a part-time helper on the Friday after Thanksgiving. The big retailers will be taking up to 30% off listed prices that week, according to Adobe's forecast, and a shop that size cannot match that without losing money on every order. What it can do is post: a photo of the window, a two-line story about the woman who pours this year's best-selling candle, a reminder of the last day for local delivery.
Small businesses trust their own posts more than Black Friday ads this year, and they can't keep up with making them. A new survey says so in plain numbers, and it matters whether you run a small shop, make content for one or build tools that small businesses pay for.
What the survey found
On 19 November, Verizon Business published a survey of 600 US small business owners and decision-makers, fielded online by RepData between 30 October and 9 November. The businesses ranged from one to 500 employees, and solo entrepreneurs made up 35% of the sample, so this is a fair picture of the very small end of the market.
Asked what will lift holiday sales, 45% named their own social media posts, ahead of word-of-mouth referrals at 42% and website or delivery improvements at 31%. Verizon's own summary says owners are prioritising direct customer engagement "rather than relying on Black Friday, Cyber Monday or other major retail tentpole moments". In other words, owners have decided that the audience they built themselves is worth more to them in December than the national shopping calendar.
The same owners say the work behind that bet is heavy. 34% would welcome help creating social media content, which made it the most wanted area of relief in the survey, ahead of administrative work (30%) and website updates (26%). Put those two findings side by side and you can see the problem plainly, because the channel owners expect the most from is also the job they most want taken off their plate.
They are reaching for AI to close the gap. 45% plan to use AI tools to simplify holiday operations, second only to social media itself at 53%, and 56% will expand AI use for internal tasks such as research, marketing and inventory. Close to half the sample plans to hand part of December to software, which is a large share for a group many vendors still treat as late adopters.
The mood behind all this is not relaxed. 51% are stressed about rising costs and tariffs, 34% have pulled an all-nighter to keep the business running, and 18% say they have cried over holiday business demands. Even so, 44% expect higher holiday sales than last year. Owners are hopeful and exhausted at the same time, which is about right for anyone running a shop in December.
Why owners are betting on their own posts
Three pressures push in the same direction. The first is that a small shop cannot win the discount fight. Adobe expects US online holiday sales of $275.1 billion in November and December, up 6.7%, with discounts of up to 30% expected in the five days from Thanksgiving to Cyber Monday. A shop with thin margins can't give away 30% of the price, and a small seller who joins in trades margin for attention and often gets neither.
The second is cost. In the 2026 Bank of America Business Owner Report, 91% of small and mid-sized business owners said their costs rose over the past 12 months, up from 77% a year earlier, and 79% raised prices, by nearly 13% on average among those who did. After a price rise like that, a deep holiday discount hurts twice. A post that reminds existing customers why they like you costs time instead of margin, and owners clearly prefer to spend time.
The third is that AI has made the first draft cheap. The same Bank of America report found that 84% of owners use AI in some capacity, and the most common use among them is marketing and advertising, at 52%. AI tools were also the top planned digital investment for the year ahead (50%), followed by marketing technology (43%). Owners have already worked out that writing a caption, resizing a photo for three platforms or turning one product story into five posts is the sort of job these tools do tolerably well.
There is a quieter fourth reason. Word of mouth came second in the Verizon survey, and a good post is word of mouth with a longer reach: a customer shares it, tags a friend, or sends it to the person they are buying for. People rarely forward a Black Friday banner to a friend.
Two caveats are worth holding on to. Verizon sells phone and internet services to small businesses, so it has an interest in a story about busy owners needing help, and the survey measures what owners expect to drive sales, not what actually did. The figures are also American, and Black Friday matters more there than in Australia or the UK. Still, the direction matches the cost and AI numbers from Bank of America, which Ipsos collected separately in August, and that makes the finding more than a vendor's talking point.
What it means for you
If you run a small business, the survey describes your competitors' December. Nearly half of them will post more and lean on AI to do it, so feeds will be busier and full of similar-sounding captions. Our estimate is that the posts that sell will be the ones only you could have made, with your own photos, your own customers and specific dates, while AI handles the drafting, resizing and scheduling around them.
If you are a creator, the 34% figure is a demand signal you can act on this month. Brands are already moving money towards content: we covered how brands will spend $116 billion on creators in 2026, with paying for content the fastest-growing slice. A local business does not need a big influencer. It needs someone who can shoot a batch of good product photos in an afternoon and hand over a month of captions in the owner's voice. Our estimate is that a fixed-price holiday content pack (one shoot, 14 posts, captions and a schedule) is easier to sell to a small shop in late November than an open-ended retainer, because the owner can see exactly what they get before the rush starts.
If you build tools, judge your product against one question: how many publishable posts can a tired owner make in an hour with it? Business AI money keeps flowing to narrow tools that do one job well, the same pattern we saw when companies were spending 5.4 times more on AI tools than a year earlier. An owner in the middle of December will not learn a general platform. They will pay for something that turns ten phone photos and a price list into a week of posts they can approve with a thumb.
On risk, a separate survey offers a useful warning. In PYMNTS Intelligence's study of 533 US small and medium businesses, 60% of AI users said they produce more without adding staff, and 13% skipped a full-time hire they would otherwise have made, against 3.3% who laid someone off. But 23% gained customers who specifically wanted a person doing the work. Customers notice when a business stops talking to them, and a feed of interchangeable captions is a quick way to lose the warmth that made people buy from a small shop in the first place.
A worked example: two weeks of holiday posts
Imagine you run a small candle and homewares shop with an online store and about 2,000 Instagram followers. You want one post a day for the 14 days from 23 November to 6 December. Every number in this section is illustrative, not survey data.
Done by hand, each post takes roughly 45 minutes: take and edit a photo, write a caption, choose hashtags, post at a sensible time. Fourteen posts come to about 10.5 hours, squeezed into evenings during your busiest fortnight.
With AI in the loop, the work changes shape:
Shoot once (2 hours). Photograph your ten best sellers in natural light, plus three shots of the shop and one of you packing an order.
Write the brief (20 minutes). List each product, its price, one true detail (who makes it, what it smells like, how long it burns), your delivery cut-off dates and any offer. Paste in three past captions that sounded like you.
Draft in one sitting (30 minutes). Ask your AI assistant for 14 captions from the brief, each under 80 words, each tied to one product or one date, with no claims that are not in the brief.
Edit and approve (1 hour). Rewrite anything that does not sound like you, check every price and date, and cut any caption you would be embarrassed to read aloud to a customer.
Schedule (30 minutes). Load the posts into your platform's scheduler, with a separate discount code or link for each post so you can see which ones sold.
That comes to about 4 hours 20 minutes against 10.5 hours by hand, a saving of roughly six hours (illustrative). If you value your time at US$50 an hour, that is about US$300 of time back in the fortnight when you have least of it. If the tracked posts bring in six extra orders at an average of US$45, they have earned US$270 in sales, which covers a typical monthly AI subscription several times over (illustrative).
Most of the saving comes from batching the work into one shoot and one writing session. The shoot and the edit still take most of the hours, and they should, because those are the parts customers notice.
What to do this week
Use one decision rule for every post: if it does not name a real product, a real date or a real person connected to your shop, do not publish it. AI is very good at producing the other kind, and those are the posts customers scroll past.
Fix your cut-off dates now (last day for shipping, last day for local delivery, holiday opening hours) and plan the posts backwards from them.
Batch the shoot this weekend, before the rush, and keep the photos real. Do not use AI-generated images of your products; if a candle looks different out of the box, the refund costs more than the post earned.
Give the AI a brief made of facts plus three of your own past captions, and ask for drafts only. You approve every word.
Track each post with its own code or link, so in January you know which posts sold and which only collected likes.
If content is the job you most want help with, as it is for 34% of owners in the Verizon survey, consider paying a local creator for one shoot rather than asking AI to fill the gap with images that look like stock.
For builders, there is one more move. Small businesses will want to look back on December in January, and a tool that tells them which posts made money is easier to sell than one that only writes them. We wrote about a related idea, a December recap your users send to their own clients, which works for any product that holds a season's worth of data about its users.
The owners in this survey have already decided that their holiday sales will come from their own posts, so the useful question for the next five weeks is how to make those posts faster without making them worse.
Sources
PYMNTS: 60% of Small Business AI Users Do More Without Adding Staff (13 November 2026)

