Every AI use case, mapped to the business.

32 use cases owners are being sold this year, sorted by where they act in a business of 5 to 100 people. Only 3 are proven in production.

Status
Pick a room to see its use cases
How we rate

Four statuses, one rule: rate the evidence, not the pitch.

Each entry gets a status for how proven it is in businesses of 5 to 100 people, and a separate rating for how strong the evidence behind that status is.

Evidence strength

StrongSeveral independent studies or large independent surveys.
ModerateAt least one independent study or credible survey, plus consistent case evidence.
ThinA handful of independent data points.
Vendor claimsOnly what sellers publish about their own products.

How an entry is made

Each entry is drafted from public sources: academic studies, regulators, independent surveys and reputable reporting. Every figure carries a numbered citation you can open. Anything a vendor says about its own product is labelled as a vendor claim.

Review and updates

Every source in this edition was opened and checked against the text in October 2026. Each entry is then signed off by a named operator and re-checked when the evidence moves. Until that sign-off, the entry is marked as a draft at the top of the page.

The full guide

All use cases at a glance

Field GuideBrand, marketing & customersAI ad creative generation
In pilotsBrand, marketing & customers

AI ad creative generation

Also called AI ad images, generative ad copy, AI banner ads, AI ad variations

EvidenceModerate
ClassGenerative

AI-made ad images and copy can lift click-through in studies, but disclosure hurts and clicks are not sales. Evidence, costs and risks.

ReviewDraft, awaiting operator review
Drafted27 October 2026
Fact-checkedEvery source opened, October 2026
Key takeaway

Independent studies suggest AI-generated ad images can match or beat human work on clicks. In one field test of more than 173,000 impressions, the best AI image drew up to 50% more clicks than professional stock photography⁠. NYU Stern and Emory researchers found fully AI-made ads lifted click-through by up to 19%⁠. But telling viewers an ad was AI-made cut click-through by 31.5%⁠. Clicks are not sales, and none of this evidence looks at small firms specifically.

At a glance

StatusIn pilots
What it doesGenerates ad images, backgrounds, headlines and copy variants for testing on Meta, Google and other ad platforms.
Typical costCanva Free at £0 includes up to 20 AI uses; Canva Pro is £100 per person per year (UK list prices, checked Oct 2026)⁠. Generative features inside ad platforms vary by platform.
Data it needsProduct photos, brand assets, approved claims and offers, and past ad results joined to sales so you can judge what worked.
Setup effortLow to start, Medium to do properly: generating is easy, but clean tests that tie clicks to sales take discipline.
EvidenceModerate. Two independent academic studies on clicks plus one large vendor-run test; no independent evidence on sales or on small firms.

What is AI ad creative generation?

AI ad creative generation uses image and text models to produce the parts of an advertisement: product shots on new backgrounds, lifestyle scenes, headlines, body copy and dozens of variants to test against each other. It is offered inside design tools and, increasingly, inside the ad platforms themselves.

For a small business the appeal is simple. A professional shoot or an agency round costs money and weeks. An AI tool can produce twenty variants in an afternoon, and the ad platform's own testing can pick the winner.

Use is spreading from a low base. A US Census Bureau survey found the share of US firms using AI rose from 3.7% to 5.4% between September 2023 and February 2024, with marketing automation among the most common applications⁠. Use was higher in young firms, and generally in large ones⁠.

How does it work?

Image models turn a text prompt, or a product photo plus a prompt, into new pictures. Text models write headlines and descriptions. Ad platforms then show different versions to different people and shift spend towards whichever earns more clicks or conversions.

The newest systems learn from ad performance directly. Meta's researchers trained a text model on historical ad results and tested it with nearly 35,000 US advertisers over ten weeks. They report a 6.7% higher click-through rate, from about 3.1% to 3.3%, than their earlier generator⁠. That is a vendor study of its own product, and it compares two AI models, not AI against human copywriters⁠.

The point for an owner: the machine is good at producing options. It still needs a person to decide what is true, what fits the brand and what the business can deliver.

What it needs to work

AI creative only earns its keep if each ad variant can be tied to sales rather than clicks, and if every claim and image is something the business can stand behind.

  1. Clean product photos and brand assets (logo, colours, fonts) stored in one shared folder
  2. A written list of approved claims, prices and offers, checked against what you can actually deliver
  3. At least three months of past ad results exported from Meta Ads or Google Ads, with spend, clicks and conversions per ad
  4. Conversion tracking that joins ad clicks to orders in Shopify or WooCommerce, or to won deals in your CRM
  5. Gross margin by product, so ads are judged on profit per sale rather than cost per click
  6. A sign-off step where a person checks every image and claim before launch

Most use cases in this guide succeed or fail here, on the data underneath them, before any AI is involved.

What does the evidence show?

  • Quality ratings: in a study of 10,320 AI images judged in 254,400 human evaluations, AI images could surpass human-made ones on quality, realism and aesthetics⁠.
  • Field clicks: across more than 173,000 banner impressions, the best AI image drew up to 50% more clicks than a professional stock photo⁠. That is one campaign, not a general rule.
  • Experiments: NYU Stern and Emory researchers found fully AI-generated visual ads raised click-through by up to 19% against ads by human experts, while human ads merely touched up with AI showed no significant gain⁠.
  • Disclosure penalty: when viewers were told an ad was made with AI, click-through fell 31.5%⁠.
  • Vendor evidence: Meta reports a 6.7% click-through gain from its own performance-trained text model⁠.

What is missing: independent evidence on sales, margin or cost per acquisition, and any study of businesses with 5 to 100 staff.

AI-generated creative vs agency or in-house creative

Compared onAI-generatedAgency or in-house
Time to first variantsMinutes to hoursDays to weeks
Number of variantsDozens, cheaplyA handful, each costing time or fees
Click performanceMatched or beat human work in two independent studies⁠The benchmark AI is tested against
Main riskMisleading images, invented claims, disclosure penalty⁠Cost and slower testing
Best forTesting many angles on products you already sellLaunches and claims that need careful judgement

Status and evidence

StatusIn pilots
Evidence strengthModerate

We rate this pilots: independent studies show AI creative can match or beat human creative on clicks⁠. But they measure clicks rather than sales, come from experiments or single campaigns, and show a large penalty when AI use is disclosed⁠. Evidence is moderate: two independent academic studies plus a large but vendor-run test⁠.

Where does it go wrong?

  • Misleading images. UK advertising rules apply however an ad was made, and the advertiser stays responsible. The ASA warns that AI images used to show what a product does can mislead if they overstate results⁠. A too-perfect AI photo of a meal, a room or a renovated kitchen invites complaints.
  • Fake testimonials. The FTC's 2024 rule bans fake reviews and testimonials, including AI-generated ones⁠. AI avatars are not banned outright, but an avatar testimonial can still be deceptive⁠.
  • Disclosure backlash. 76% of US adults say it matters to know if content is AI-made⁠, and labelled AI ads lost about a third of their clicks in testing⁠.
  • Political and social issue ads. Meta requires disclosure in certain cases when such ads are digitally created or altered⁠.
  • Sameness. AI adopters' work became less novel on average over time⁠, so competitors using the same tools converge.
  • Chasing clicks. Cheap variants make it easy to optimise for click-through while cost per sale quietly rises.

Numbers it should move

Quote conversionThe share of quotes that turn into orders.
Gross marginSales minus the cost of the goods sold, as a share of sales.

Ask the demo business

Northgate Supply is a fictional wholesale distributor of catering and packaging supplies: 28 staff, R46.6m in sales over the last year. Its data is generated for this demo, to 30 September 2026.

Fictional data
What share of our quotes turn into orders?
38%of quotes won in September

We won 38% of 1 909 quotes over the year. By rep over the last six months, the range runs from 30% to 46%.

OctNovDecJanFebMarAprMayJunJulAugSep38%
Thandi46%
Pieter35%
Aisha42%
Marco30%
How this was worked out

Quotes marked as won divided by all quotes issued, by month and by rep.

Answered in 2 msScanned 1 909 rowsFrom quotes
Want this on your own numbers, in the AI tools you already use?Book a demo

Where the data usually lives

Meta AdsGoogle AdsShopifyWooCommerceHubSpotExcel and Google Sheets

Where the records this use case needs are usually kept in a small business. It is not a list of recommended tools.

Frequently asked questions

How much does AI ad creative cost for a small business?

Design tools with AI can start free: Canva Free includes up to 20 AI uses and Canva Pro costs £100 per person per year (UK list prices, checked Oct 2026)⁠. Generative features inside ad platforms vary by platform. The larger cost is ad spend on variants that win clicks but not sales, so budget for a proper test, not just the tool.

Do I need a developer or agency to make AI ads?

No developer is needed to generate images or copy. Technical help pays off in tracking: making sure each ad's clicks are joined to orders or won deals, so you can see which variant actually made money. Without that, you are judging ads on clicks, which is what the research measured⁠, not on profit.

Are AI-generated ads legal in the UK and US?

Yes, but normal rules still apply. The UK's ASA says its code covers ads however they were made and the advertiser remains responsible⁠. In the US, the FTC's rule bans fake reviews and testimonials, including AI-generated ones⁠, and avatar testimonials can still be deceptive⁠. We did not review South African advertising rules for this entry. This is general information, not legal advice.

Should I tell customers an ad was made with AI?

It is a trade-off: disclosure costs clicks, but audiences want to know. In NYU experiments, disclosing AI use cut click-through by 31.5%⁠. Yet 76% of US adults say knowing matters to them⁠, and Meta requires disclosure on some digitally created or altered political and social issue ads⁠. Whatever you decide, never use AI to show results a product cannot deliver⁠.

Do AI ads perform better than ads made by people?

Sometimes, on clicks. Independent studies found AI images matched or beat human ones, with the best AI banner getting up to 50% more clicks than stock photography in one field test⁠. Human ads merely touched up with AI did no better⁠. We found no independent evidence on sales or profit for small businesses.

More in brand, marketing & customers

AI review responsesIn pilotsEvidence: ThinCustomer-service chatbotsIn pilotsEvidence: StrongAI-written social contentCosmeticEvidence: Thin

Sources

  1. 1
    The power of generative marketing: Can generative AI create superhuman visual marketing content?
    SSRN (Hartmann, Exner and Domdey; International Journal of Research in Marketing, forthcoming), 2024Study
  2. 2
    The AI Advertising Paradox
    NYU Stern School of Business (Ghose, Lee, Todri and Adamopoulos), 2025Study
  3. 3
  4. 4
    Generative AI & Advertising: Decoding AI Regulation
    Advertising Standards Authority / CAP, 2023Government
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
    Canva pricing
    Canva, 2026Vendor
  10. 10
  11. 11

This entry is educational. It is not legal, financial or professional advice, and it does not recommend any vendor. Figures attributed to vendors are their own claims.

Ask your own business. Get the true number in seconds.

Veld connects your accounting, sales and operations data to the AI tools you already use, and keeps it running, so every answer comes from your numbers.

Book a demo