Store velocity analysis spots volume jumps

Drop in applications with each store and date to compare this period with the previous one and the same period last year against how peers moved. See which stores have jumps the test says chance does not explain, in a minute.

Applications file

Drop a CSV file here or pick one

33,337 applications from 31 stores between 2024-09 and 2026-08 were read, and 0 rows without a store or a readable date were left out.

The file is read in this browser and never uploaded, and only counts are sent if you email the result.

Store, dealer or agent

Application date

Period length

Period analysed

Warning multiple, ×

A store is flagged when it grew at least this many times faster than its peers and the test says chance does not explain it.

Minimum applications

Confidence, %

As more stores are tested, more may jump by chance alone, so the test demands more from each store. Higher confidence flags fewer stores.

These are generated demo applications from 30 stores over two years, with a few jumps and one new store planted in them.

Stores flagged

Book change+11.2%

Year-on-year change+33.0%

Share from flagged stores20.2%

3 stores grew at least 1.5× as fast as their peers by more than chance explains at 95% confidence, or are new. Together, they account for 20.2% of this period’s applications, and a jump merits review but does not establish fraud.

Applications over time

S05

The dashed line shows the store at its share of the book over the twelve periods before this one. Click a store in the table to draw it.

Stores by excess

Rank	Store	This period	Previous	Change	Peer index	Last year	Year peer index	Excess	Flag
1	S05	150	165	−9.1%	0.8×	60	2.0×	74	Year on year
2	S31	72	0	–	–	0	–	72	New
3	S17	105	40	+162.5%	2.5×	28	2.9×	62	Both
4	S02	48	26	+84.6%	1.7×	28	1.3×	19	With peers
5	S12	58	37	+56.8%	1.4×	45	1.0×	17	With peers
6	S13	37	25	+48.0%	1.3×	27	1.0×	9	With peers
7	S23	34	24	+41.7%	1.3×	19	1.4×	7	With peers
8	S19	39	29	+34.5%	1.2×	27	1.1×	7	With peers
9	S26	61	49	+24.5%	1.1×	61	0.7×	7	With peers
10	S10	44	34	+29.4%	1.2×	36	0.9×	6	With peers
11	S22	65	53	+22.6%	1.1×	51	1.0×	6	With peers
12	S18	75	63	+19.0%	1.1×	56	1.0×	5	With peers
13	S20	42	36	+16.7%	1.1×	29	1.1×	2	With peers
14	S15	64	57	+12.3%	1.0×	40	1.2×	1	With peers
15	S14	18	16	+12.5%	1.0×	23	0.6×	0	With peers
16	S21	19	17	+11.8%	1.0×	25	0.6×	0	With peers
17	S25	29	27	+7.4%	1.0×	26	0.8×	-1	With peers
18	S08	63	59	+6.8%	1.0×	70	0.7×	-3	With peers
19	S04	60	57	+5.3%	0.9×	43	1.1×	-4	With peers
20	S30	38	38	+0.0%	0.9×	27	1.1×	-4	With peers
21	S09	35	36	−2.8%	0.9×	38	0.7×	-5	With peers
22	S16	21	27	−22.2%	0.7×	15	1.1×	-9	With peers
23	S28	75	76	−1.3%	0.9×	60	0.9×	-10	With peers
24	S24	74	76	−2.6%	0.9×	66	0.8×	-11	With peers
25	S06	19	27	−29.6%	0.6×	27	0.5×	-11	With peers

6 more stores are in the download.

Get the result by email

Run the same check on every application at the moment of the decision, using the store’s recent volume and its peers’ volume. Start free

Comparison method

Peer growth = (This period’s book − This period) ÷ (Previous book − Previous)
Peer index = (This period ÷ Previous) ÷ Peer growth
Year peer index = (This period ÷ Last year) ÷ Year peer growth
Excess = This period − Previous × Peer growth
z = (This period − n × p₀) ÷ √(n × p₀ × (1 − p₀)),  n = This period + Previous,  p₀ = Peer growth ÷ (1 + Peer growth)
Flag = p ≤ (1 − Confidence) ÷ Tests run  &  Peer index ≥ Warning multiple

Peers are every other store, so a season or campaign that lifts every store does not flag one. Confidence is shared across all tests, and a store with no applications in either earlier period that reaches the minimum is shown as new.

Store volume against peers

Velocity analysis compares a store’s application volume with its previous period and the same period last year. It measures each store’s growth against the growth of its peers, then tests whether a jump could be due to chance.

A broad rise across stores can come from seasonality or a campaign, so the peer comparison helps show when one store stands apart. Use the results to decide which jumps deserve a closer look, and review new stores that reach the minimum volume.

Common questions

What is velocity analysis in fraud monitoring?

Velocity analysis compares application volume across periods. This tool checks each store’s growth against its peers and tests whether a jump could be due to chance.

How can I spot a store with a sudden application spike?

Compare its volume with the previous period and with the same period last year. This tool flags stores whose growth against peers passes the test and warning multiple.

How do I compare store volume with peers?

For each store, compare its growth with the growth of every other store. The tool does this for the previous period and the same period last year.

How can I tell a real jump from seasonal growth?

Compare a store’s growth with its peers, since a season or campaign can lift every store. The test checks whether the store’s jump is more than chance explains.

Why is high volume at a new store a warning sign?

A new store has no applications in either earlier period to compare against. If it reaches the minimum volume this period, the tool shows it as new for review.