Marketing
16 columns
60 rows
8 formats

A/B Testing Results Dataset

This dataset provides detailed results from A/B testing experiments, including variant-level conversion metrics, statistical significance, and audience segmentation. It enables marketing teams to rigorously compare strategies, identify winning variants, and optimize campaigns for higher conversion rates. The data structure supports robust analysis, reporting, and decision-making.

last updated 12 Jul 2025
by GoMask

The brief that made it

Evaluating the effectiveness of different marketing strategies or website changes

Sample rows

preview · 8 of 60 rows · all 16 columns

experiment_idstring	variant_idstring	upliftfloat	is_controlboolean	experiment_namestring	variant_namestring	start_datedate	end_datedate	audience_segmentstring	sample_sizeinteger	conversionsinteger	conversion_ratefloat	statistical_significancefloat	primary_metricstring	secondary_metricsstring	notesstring
EXP1001	control	0	true	Homepage Banner CTA Test	Current Banner	2024-01-05	2024-01-20	all visitors	750	54	0.072	0.46	click-through rate	bounce rate,time on page	Baseline for homepage CTA
EXP1001	A	0.213	false	Homepage Banner CTA Test	Red Banner	2024-01-05	2024-01-20	all visitors	755	66	0.0874	0.12	click-through rate	bounce rate,time on page	Higher CTR observed
EXP1001	B	0.085	false	Homepage Banner CTA Test	Green Banner	2024-01-05	2024-01-20	all visitors	742	58	0.0782	0.34	click-through rate	bounce rate,time on page	Not statistically significant
EXP1002	control	0	true	Signup Button Placement	Bottom of Page	2024-02-10	2024-02-28	returning customers	540	31	0.0574	0.62	signup rate	page scroll depth,engagement time	blank
EXP1002	A	0.295	false	Signup Button Placement	Top of Page	2024-02-10	2024-02-28	returning customers	525	39	0.0743	0.21	signup rate	page scroll depth,engagement time	Improved signup rate
EXP1003	control	0	true	Email Subject Line Test	Default Subject	2023-12-01	2023-12-15	newsletter subscribers	1800	149	0.0828	0.41	open rate	click rate,unsubscribe rate	blank
EXP1003	A	0.172	false	Email Subject Line Test	Curiosity Subject	2023-12-01	2023-12-15	newsletter subscribers	1812	176	0.0971	0.09	open rate	click rate,unsubscribe rate	Statistically significant
EXP1003	B	0.104	false	Email Subject Line Test	Urgency Subject	2023-12-01	2023-12-15	newsletter subscribers	1795	164	0.0914	0.19	open rate	click rate,unsubscribe rate	blank
EXP1004	control	0	true	Checkout Flow Simplification	Current Checkout	2024-03-05	2024-03-19	mobile users	1120	98	0.0875	0.56	purchase rate	cart abandonment rate,avg order value	blank
EXP1004	A	0.11	false	Checkout Flow Simplification	One-Page Checkout	2024-03-05	2024-03-19	mobile users	1132	110	0.0971	0.13	purchase rate	cart abandonment rate,avg order value	Slightly higher purchase rate
EXP1004	B	0.063	false	Checkout Flow Simplification	Express Checkout	2024-03-05	2024-03-19	mobile users	1097	102	0.093	0.22	purchase rate	cart abandonment rate,avg order value	blank
EXP1005	control	0	true	App Push Frequency	1x Daily	2024-04-01	2024-04-15	app users	900	50	0.0556	0.79	engagement rate	session length,retention rate	blank
EXP1005	A	0.142	false	App Push Frequency	2x Daily	2024-04-01	2024-04-15	app users	913	58	0.0635	0.59	engagement rate	session length,retention rate	blank
EXP1005	B	-0.099	false	App Push Frequency	Every 3 Days	2024-04-01	2024-04-15	app users	899	45	0.0501	0.87	engagement rate	session length,retention rate	Lower engagement
EXP1006	control	0	true	Pricing Page Layout	Grid Layout	2024-05-10	2024-05-25	new users	780	57	0.0731	0.48	trial signup rate	bounce rate,cta clicks	blank
EXP1006	A	0.088	false	Pricing Page Layout	List Layout	2024-05-10	2024-05-25	new users	793	63	0.0795	0.31	trial signup rate	bounce rate,cta clicks	blank
EXP1006	B	-0.016	false	Pricing Page Layout	Tabbed Layout	2024-05-10	2024-05-25	new users	765	55	0.0719	0.55	trial signup rate	bounce rate,cta clicks	blank
EXP1007	control	0	true	Product Recommendation Algorithm	Legacy Algorithm	2024-01-20	2024-02-04	high spenders	670	62	0.0925	0.72	add-to-cart rate	order value,conversion rate	blank
EXP1007	A	0.207	false	Product Recommendation Algorithm	AI Model V1	2024-01-20	2024-02-04	high spenders	663	74	0.1116	0.07	add-to-cart rate	order value,conversion rate	Statistically significant
EXP1008	control	0	true	Onboarding Tutorial Copy	Standard Copy	2024-03-16	2024-03-30	trial users	585	34	0.0581	0.77	tutorial completion rate	churn rate,feature adoption	blank

What the 60 rows show

from the 60-row sample

Control (variant id) stands out: mean uplift is 0.0, against 0.17 for the rest.

43%is_control = true
0.05median uplift
13primary metrics
990median sample_size
113.5median conversions
0.10median conversion_rate

Mean uplift by variant_id60 rows

uplift60 rows, in bands of 0.1

Median 0.05, from -0.12 to 0.44.

experiment_id60 rows · top 10 of 26 values

EXP10013
EXP10033
EXP10043
EXP10053
EXP10063
EXP10093
exp_2024_413
exp_2024_433
exp_2024_453
EXP10022

16 columns by typefrom the column list below

string 8
integer 2
float 3
date 2
boolean 1

Columns

16 columns in four groups

blueprint · 16 columns
column	type	description	example
Text 8 columns
`experiment_id`	string	Unique identifier for the A/B test experiment.	EXP1001
`experiment_name`	string	Descriptive name of the A/B test experiment.	Homepage Banner CTA Test
`variant_id`	string	Unique identifier for the test variant (e.g., control, variant A, variant B).7 variants	control
`variant_name`	string	Human-readable name for the test variant.	Current Banner
`audience_segment`	string	Description or identifier for the targeted audience segment (e.g., 'new users', 'returning customers').optional	all visitors
`primary_metric`	string	The main metric used to evaluate the test (e.g., 'click-through rate', 'purchase rate').	click-through rate
`secondary_metrics`	string	Comma-separated list of secondary metrics tracked in the experiment.optional	bounce rate,time on page
`notes`	string	Additional notes or observations about the experiment or variant.optional	Higher CTR observed
Numbers 5 columns
`sample_size`	integer	Number of users or sessions included in this variant.1 or more	750
`conversions`	integer	Number of successful conversions for this variant.0 or more	54
`conversion_rate`	float	Conversion rate for this variant (conversions/sample_size).0 to 1	0.072
`statistical_significance`	float	P-value or confidence level indicating the statistical significance of the result.0 to 1 · optional	0.46
`uplift`	float	Percentage increase or decrease in conversion rate compared to the control variant (expressed as a decimal, e.g., 0.05 for 5%).optional	0
Dates and times 2 columns
`start_date`	date	Date when the A/B test or variant started.	2024-01-05
`end_date`	date	Date when the A/B test or variant ended.	2024-01-20
True or false 1 column
`is_control`	boolean	Indicates whether this variant is the control group.	true

Use it for

A marketing dashboard
The is_control rate, uplift by variant_id and a breakdown of experiment_id. Excel, Power BI or Tableau.
Why do the 26 control rows have a mean uplift of 0.0?
A root-cause class exercise
Hand out the rows and one question. The answer is in the data, not in the brief.
Experiments60
EXP10010control
EXP10010.213A
EXP10020control
A software demo
Believable experiments with experiment_name, variant_id and variant_name to fill a screen in front of a buyer.

This dataset60 rows16 columns

Yours1,000 rows16 columns

blueprint · ab-testing-results-dataset

Behind this dataset

Same schema. As many rows as you need.

These 60 rows came out of a blueprint — 16 columns with generation rules behind each one. Open it in Data Factory to retune a column, add your own, wire in foreign keys, and run it at the size you actually need.

Rules it was built with

Each group must have minimum 30 records
Conversion events timestamped within test window
Significance tests run for each variant
Invalid test data excluded from results

Rows1,000 rows10,000 rows100,000 rows1,000,000 rows

Open the blueprint in Data Factory

1 credit per row. Preview 20 rows free. Buy rows for the full run.

Exports: CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
Licence: yours to use, including commercially
API slug: ab-testing-results-dataset

What should your data show?

No signup. No card.

Sample rows

What the 60 rows show

Columns

Use it for

A marketing dashboard

A root-cause class exercise

A software demo

Make it yours.

Same schema. As many rows as you need.

More in Marketing

What should your data show?