AI Use Cases Index 2026

None of these 170 examples reach grade A yet, so the verified share is 0%. 67 (39.4%) state a specific time or money result. Verified means grade A on this site: independently verified.

Verified (grade A): 0%
State a time or money result: 39.4%

How many real-world business AI use cases show verified results?

Of the 170 business AI use cases we've published as of Oct 27, 2026, none yet reach grade A (independently verified). 44 (25.9%) are detailed first-hand accounts, 67 come from vendors or secondhand write-ups, and 59 are light evidence. 67 (39.4%) state a specific time or money result.

Grades, sources, skill and setup

Counted across all 170 published use cases. Grade B is a detailed first-hand account. The verified grade is A.

Each one comes from a named source, and each carries an evidence label. 0 are verified, 44 are first-hand accounts with details, 67 come from a vendor or a secondhand write-up, and 59 are light evidence. Treat light-evidence examples as ideas to test, not proof. Evidence grades

Grade A, Independently verified: 0 (0%)
Grade B, First-hand account: 44 (25.9%)
Grade C, Vendor or secondhand: 67 (39.4%)
Grade D, Anecdotal: 59 (34.7%)
Vendor case studies: 52 (30.6%)
Setup in an afternoon or less: 97 (57.1%)

Where the write-up comes from

First-hand: 88
Vendor or promotional: 59
Official: 9
Secondhand: 8
Press: 3
Guide: 3

Skill the example needs

No code: 92
Some tech skills: 47
Developer skills: 31

How the AI is used

Assisted: 80
Approve-to-send: 50
Autopilot: 40

Setup time

Under an hour: 55
An afternoon: 42
A few days: 40
A few weeks: 33

Grades and cost by department

Median start cost is $0 when the tool's pricing model is free, a free plan, open source or self-hosted. Otherwise it is the tool's listed USD price per month. 38 examples name no cost tool and are left out of the cost medians, along with missing, non-USD and unparseable prices. The count next to a median is how many examples entered it.

Evidence grade, stated results, median start cost, median paid plan and median setup by department
Department	Examples	Grades A/B/C/D	With a stated result	Median start cost	Median paid plan	Median setup	Sample
Operations	57	0/14/21/22	23	$0 (44)	$20/mo (37)	An afternoon
Sales	34	0/10/17/7	16	$0 (24)	$20/mo (23)	A few days
Marketing	26	0/5/8/13	8	$0 (22)	$20/mo (21)	An afternoon
Content	18	0/7/10/1	8	$0 (17)	$16/mo (17)	An afternoon
Finance	17	0/4/5/8	8	$0 (13)	$20/mo (11)	A few days
Customer Support	14	0/2/6/6	2	$0 (7)	$20/mo (4)	A few days
Product	4	0/2/0/2	2	$20 (3)	$20/mo (3)	An afternoon	Small sample (under 5 examples)

A yearly USD price is divided by 12 before the median: Canva $144/year counts as $12/mo (checked 2026-10-15). A price that is not in US dollars is left out of the medians: n8n €20/month (EUR, checked 2026-10-15).

Tools named most often

Counted from the tools list on each published use case. The table keeps the ten highest counts, and every tool tied for tenth place (here, 3 or more).

Tools named most often in the published use cases
Tool	Use cases	Note
1. Claude	43
2. Claude Code	28
3. ChatGPT	24
4. Grok Bot	22	House section
5. Gemini	6
6. Codex	4
7. Make	4
8. Photoroom	4
9. Zapier	4
10. Canva	3
11. Clay	3
12. Hermes Agent	3
13. Microsoft 365 Copilot	3
14. n8n	3
15. OpenClaw	3
16. Opus Clip	3

Grok Bot is in this table because 22 published use cases name it. Grok Bot is this site's house section (/grok-bot/). The count uses the same rule as every other tool. This is not an independent ranking and not a paid placement. Grok Bot setups

Free dataset

One row per published use case, 170 rows, the same rows the tables above are counted from. 25 income automations are left out: they are a different record.

CSV of the 170 published use cases. The same rows as JSON.

No public license is set on this dataset. Do not treat the file as CC BY 4.0 until the site owner confirms a license. Summaries are ours; sources belong to their authors.

How these figures are computed

Inclusion: published use cases only (status other than hidden). Each one links a named source. Income automations are excluded.
Grading: A to D, the same labels as the department pages. Each one comes from a named source, and each carries an evidence label. 0 are verified, 44 are first-hand accounts with details, 67 come from a vendor or a secondhand write-up, and 59 are light evidence. Treat light-evidence examples as ideas to test, not proof. Evidence grades
Results: a use case counts as stating a result only when it has a row in the results table. Every number in that row also appears in the use case outcome. The build checks that before this page is written.
Cost: the figure is the tool's own pricing page (the checked date is on the tool), not what the business spent. Start cost is $0 for a free, free-plan, open-source or self-hosted model, and otherwise the listed USD price per month. Non-USD prices, missing prices and prices that do not parse are excluded, and the median shows how many prices remained. A yearly USD price is divided by 12 before the median: Canva $144/year counts as $12/mo (checked 2026-10-15). A price that is not in US dollars is left out of the medians: n8n €20/month (EUR, checked 2026-10-15). 38 examples name no cost tool.
Setup time: an editor's estimate from the source, on the scale under an hour, an afternoon, a few days, a few weeks. The median of an even count is the lower of the two middle values.
Limitations: these are public examples the editors chose to publish, so they are not a random sample of businesses. Vendor case studies are 52 of 170 (30.6%). No example has independent verification yet. Product has fewer than 5 published examples, so that row is marked small sample.
As of Oct 27, 2026, the content commit 8739c74. That is the newest commit that changed the use cases, the results table, the tool pages or the department list. A later edit to this page does not move the date.

Revisions

v3.22.0 (Oct 27, 2026, content 8739c74): first published index. Every count is computed from the published use cases at build time.