Do developer-adoption signals lead revenue?

A research note on six public data-infrastructure companies: Snowflake, Datadog, MongoDB, Confluent, Elastic and Cloudflare.

Kechun Wu · Data as of Nov 1, 2026 · Numbers on this page are regenerated weekly from the latest data · Full methodology

Summary Question Data Method Findings Interpretation Limitations All results

Summary.

HN hiring-post share: No detectable lead. 0 of 10 pooled lag × fixed-effect cells have p < 0.05; 0 of 8 leading cells (k ≥ 1) have a 95% CI entirely above zero. (no cell reaches p < 0.05; strongest cell: lag 1q, company FE, r = +0.11 [-0.05, +0.27], p = 0.180, lag-adjusted p = 0.899, n = 161 quarter-obs across 6 companies)
npm downloads: Weak / inconsistent evidence. 8 of 10 pooled lag × fixed-effect cells have p < 0.05; 7 of 8 leading cells (k ≥ 1) have a 95% CI entirely above zero. (some cells have p < 0.05 but none survive lag adjustment under both FE variants; strongest cell: lag 0q, company FE, r = +0.51 [+0.18, +0.74], p = 0.005, lag-adjusted p = 0.023, n = 142 quarter-obs across 5 companies)
PyPI downloads: not testable yet (not enough overlapping quarters yet).
npm vs baseline (post-hoc): Weak / inconsistent evidence. 7 of 10 pooled lag × fixed-effect cells have p < 0.05; 6 of 8 leading cells (k ≥ 1) have a 95% CI entirely above zero. (some cells have p < 0.05 but none survive lag adjustment under both FE variants; strongest cell: lag 0q, company FE, r = +0.42 [+0.11, +0.66], p = 0.011, lag-adjusted p = 0.056, n = 142 quarter-obs across 5 companies)
Company by company: 0 of 55 pre-registered tests have p < 0.05, against ≈2.8 expected from chance alone; 0 survive a false-discovery-rate correction.
What it means: the interpretation (section 4.5) →
PyPI downloads cannot be tested yet (only ~180 days of history), and every result rests on 6 companies with short, smooth series. Treat the verdicts as “what this data can detect”, not as proof either way.

1. Question

Analysts increasingly use alternative data such as package downloads and job postings to read product adoption ahead of reported numbers. If developers adopting a data-infrastructure product shows up first in what they install and what employers hire for, then growth in those signals should precede growth in reported revenue. This note tests that directly: does year-over-year growth in a company's developer-adoption signals correlate with its year-over-year revenue growth 0–4 quarters later?

2. Data

Company	FY ends	Revenue quarters	# quarters	# revenue YoY	# HN YoY	# npm YoY	npm proxy
Snowflake (SNOW)	01-31	2019-10 → 2026-07	28	24	24	24	`snowflake-sdk`
Datadog (DDOG)	12-31	2018-09 → 2026-06	32	28	28	28	`dd-trace`
MongoDB (MDB)	01-31	2016-10 → 2026-07	40	36	36	36	`mongodb`
Confluent (CFLT)	12-31	2020-06 → 2025-12	23	19	19	3	`@confluentinc/kafka-javascript`
Elastic (ESTC)	04-30	2017-10 → 2026-07	36	32	32	26	`@elastic/elasticsearch`
Cloudflare (NET)	12-31	2018-09 → 2026-06	32	28	28	28	`wrangler`

Revenue: SEC XBRL company facts (10-K/10-Q). Q4 is derived as full-year minus nine-month YTD. Fiscal years ending in January (SNOW, MDB) and April (ESTC) are handled by aggregating every signal over each company's exact fiscal-quarter dates. HN share: share of monthly Who is hiring? job posts mentioning the product, excluding the company's own job posts and customer/investor lists. npm: the company's official client package (proxy rationale). PyPI history only began ~180 days before the first pull, so PyPI cannot be tested yet.

3. Method

Variables: log YoY growth (same fiscal quarter a year earlier) of revenue and of each signal.
Lags: corr(signal_t−k, revenue_t) for k = 0, 1, 2, 3, 4 quarters. All lags are reported; none is selected.
Uncertainty: YoY series are highly autocorrelated, so 30 quarters are worth far fewer independent observations. CIs and p-values use a Bartlett effective sample size (n_eff). A null simulation run before any data was pulled showed that a naive test flags up to 58% of unrelated persistent series as “significant”; the n_eff test keeps that near 5%.
Multiple testing: a Bonferroni p over the five lags (paying for “picking the best lag”) and Benjamini-Hochberg q-values across all company tests.
Pooled panel: the six companies stacked with company fixed effects, and with company + calendar-quarter fixed effects. The second version removes anything common to all six in a quarter (rates, the 2021 boom and 2022–23 cost-cutting, registry-wide download inflation).
Pre-registration: specifications, lags and verdict rules were committed to the repo before the first data pull (commit b02b808, 17 minutes before the first pull in cd5a0b6; both on 2026-11-04, so this guards against tuning to the results, not against prior familiarity with the data). Two signals were added after seeing the data and are labelled post-hoc: npm downloads relative to a basket of mature packages, added because every npm package's downloads inflated together in 2026. Post-hoc results are kept in a separate multiple-testing family.

4. Findings

4.1 Pooled across the six companies

Dots: correlation r at each lag; bars: 95% CI using the effective sample size. Blue: company fixed effects. Orange: company + calendar-quarter fixed effects. A real lead would show up as a positive r at k ≥ 1 that stays clear of zero in both versions.

Signal	Fixed effects	lag 0	lag 1	lag 2	lag 3	lag 4	Verdict
HN hiring-post share	company	+0.09 [-0.07, +0.24]	+0.11 [-0.05, +0.27]	+0.09 [-0.07, +0.25]	+0.07 [-0.10, +0.24]	+0.07 [-0.11, +0.25]	null
HN hiring-post share	company + quarter	+0.07 [-0.09, +0.23]	+0.08 [-0.09, +0.24]	+0.06 [-0.11, +0.23]	+0.02 [-0.15, +0.19]	-0.00 [-0.17, +0.17]
npm downloads	company	+0.51* [+0.18, +0.74]	+0.51* [+0.16, +0.75]	+0.51* [+0.15, +0.76]	+0.49* [+0.11, +0.74]	+0.49* [+0.11, +0.75]	mixed/weak
npm downloads	company + quarter	+0.19 [-0.07, +0.42]	+0.24 [-0.02, +0.47]	+0.30* [+0.04, +0.52]	+0.31* [+0.05, +0.52]	+0.29* [+0.04, +0.51]
npm vs baseline post-hoc	company	+0.42* [+0.11, +0.66]	+0.40* [+0.07, +0.65]	+0.39* [+0.04, +0.65]	+0.37* [+0.00, +0.64]	+0.37 [-0.00, +0.65]	mixed/weak
npm vs baseline post-hoc	company + quarter	+0.19 [-0.07, +0.42]	+0.24 [-0.02, +0.47]	+0.30* [+0.04, +0.52]	+0.31* [+0.06, +0.52]	+0.29* [+0.04, +0.51]

4.2 Company by company

Signal

Cell = r between the signal's YoY growth k quarters earlier and revenue YoY growth. * marks unadjusted p < 0.05. With 5 lags × several signals × 6 companies, some * cells are expected by chance alone (2.8 expected among 55 pre-registered tests).

Survive the lag adjustment (Bonferroni over 5 lags, p < 0.05): none. Survive false-discovery-rate control (q < 0.10): 0 of 55. Positive correlations: 91% of all cells and 89% of leading cells (k ≥ 1); about 50% is expected if there is no relationship, but the cells are not independent (same company, adjacent lags), so this is descriptive only.

4.3 Robustness: acceleration instead of growth

Using the quarter-on-quarter change in YoY growth (acceleration) on both sides strips out slow shared trends. In the pooled panel, 0 of 20 cells have p < 0.05 and 0 survive the lag adjustment. Strongest cell: HN hiring-post share, lag 1, company FE, r = +0.13 [-0.03, +0.29], p = 0.120.

4.4 Look at the raw series

Company

Log YoY growth by fiscal quarter (0.4 ≈ +49%). Seeing the series explains most of the results above: these are short, smooth, trending series, and a few early hyper-growth quarters can dominate a correlation.

4.5 Interpretation

Written 2026-11-04 on data through the week ending 2026-11-01 (SEC filings through 2026-09). The tables and charts above regenerate weekly; this paragraph is updated by hand when the picture changes.

HN hiring-post share: a clean null. Pooled correlations are small (between −0.01 and +0.11) at every lag, every 95% CI includes zero, and nothing survives the lag adjustment. Faster growth in how often job posts mention Snowflake, Datadog, MongoDB, Kafka, Elastic or Cloudflare did not precede faster revenue growth, at least not by enough to see in 19 to 36 quarters per company.

npm downloads: a real co-movement, and at most a weak hint of a lead. With company fixed effects the pooled correlation is about +0.5 at every lag, including lag 0, and flat across lags. A flat profile is the signature of a shared slow trend: young products start with hyper-growth in both their client library and their revenue, and both decelerate together. Once that trend is stripped out, what is left is small and not statistically detectable:

removing anything common to all companies in a quarter (company + quarter fixed effects) shrinks r to +0.19 at lag 0, but it rises with the lag, to +0.30 and +0.31 at lags 2 and 3, and lags 2 to 4 have p < 0.05 with 95% CIs above zero before any correction. That is the shape a lead would have, and at lags 2 and 3 the CIs clear zero in both fixed-effect versions. It does not survive the correction for trying five lags (best lag-adjusted p = 0.09), so it is a pattern to keep watching, not a finding;
switching to acceleration (change in YoY growth) on both sides removes slow trends, and all 30 pooled cells (including the post-hoc ones) then have p > 0.1;
company by company, r is high (median +0.66) and positive in 25 of 25 cells, but each series has an effective sample of only about 3 to 11 independent observations, so none is significant on its own.

Bottom line: developer-adoption signals track the long arc of these companies' growth. HN hiring share shows no lead at all. npm downloads show no lead that survives the pre-registered correction, but the stricter specification leaves a modest lead-shaped pattern (r ≈ +0.3 at two to three quarters) that this sample, five companies with an npm proxy, is too small to confirm or rule out. So: no detectable lead, not evidence of no lead. The weekly rebuild will firm the pattern up or erase it. The usual limits apply: small samples, CI and bot noise in downloads. PyPI becomes testable in late 2027.

5. What this can and can't tell you

Can

Whether, over the available history, faster growth in a developer signal was followed by faster reported revenue growth, with honest uncertainty bands.
Whether any apparent relationship survives autocorrelation, the five lags tried, and a common cycle hitting all six companies at once.

Can't

Prove causation or forecast revenue. A shared driver (rates, the 2021 boom, 2022–23 cost-cutting) can move both.
Detect small effects. Each company has roughly 15–35 quarterly YoY points, and an effective sample often under 10. A null result here means “not detectable with this data”, not “no relationship”.
Isolate paid usage. Proxies include open-source and free-tier use (community MongoDB, Apache Kafka, OpenSearch users of Elastic clients) and CI or bot downloads. HN skews toward US startups.
Use PyPI yet. pypistats keeps only ~180 days; PyPI YoY becomes testable after about a year of weekly appends.

This is a research exercise, not investment advice.

Appendix: every test run

Show all 300 company-level tests

Ticker	Signal	Spec	Lag	n	n_eff	r	CI low	CI high	p	p (lag-adj)	q (BH)	p (perm)
CFLT	HN hiring-post share	d_yoy	0	18	18	-0.09	-0.53	+0.40	0.735	1.000	0.939	0.849
CFLT	HN hiring-post share	d_yoy	1	17	17	-0.00	-0.48	+0.48	1.000	1.000	1.000	1.000
CFLT	HN hiring-post share	d_yoy	2	16	16	+0.32	-0.21	+0.71	0.223	1.000	0.893	0.118
CFLT	HN hiring-post share	d_yoy	3	15	15	-0.37	-0.74	+0.18	0.176	0.881	0.893	0.130
CFLT	HN hiring-post share	d_yoy	4	14	13	+0.33	-0.27	+0.75	0.274	1.000	0.893	0.394
DDOG	HN hiring-post share	d_yoy	0	27	27	-0.09	-0.46	+0.30	0.646	1.000	0.939	0.629
DDOG	HN hiring-post share	d_yoy	1	26	26	+0.22	-0.18	+0.56	0.284	1.000	0.893	0.219
DDOG	HN hiring-post share	d_yoy	2	25	25	+0.03	-0.37	+0.42	0.891	1.000	1.000	0.821
DDOG	HN hiring-post share	d_yoy	3	24	24	-0.07	-0.46	+0.34	0.739	1.000	0.939	0.732
DDOG	HN hiring-post share	d_yoy	4	23	23	+0.12	-0.30	+0.51	0.576	1.000	0.939	0.549
ESTC	HN hiring-post share	d_yoy	0	31	28	+0.09	-0.29	+0.44	0.660	1.000	0.939	0.644
ESTC	HN hiring-post share	d_yoy	1	30	26	-0.00	-0.39	+0.38	0.983	1.000	1.000	0.982
ESTC	HN hiring-post share	d_yoy	2	29	21	-0.01	-0.43	+0.42	0.982	1.000	1.000	0.977
ESTC	HN hiring-post share	d_yoy	3	28	20	+0.00	-0.44	+0.44	0.999	1.000	1.000	0.999
ESTC	HN hiring-post share	d_yoy	4	27	18	+0.09	-0.40	+0.53	0.734	1.000	0.939	0.691
MDB	HN hiring-post share	d_yoy	0	35	35	+0.06	-0.28	+0.39	0.728	1.000	0.939	0.696
MDB	HN hiring-post share	d_yoy	1	34	34	+0.21	-0.14	+0.51	0.234	1.000	0.893	0.256
MDB	HN hiring-post share	d_yoy	2	33	27	-0.16	-0.51	+0.23	0.413	1.000	0.939	0.355
MDB	HN hiring-post share	d_yoy	3	32	27	-0.09	-0.45	+0.30	0.665	1.000	0.939	0.618
MDB	HN hiring-post share	d_yoy	4	31	25	+0.27	-0.13	+0.60	0.184	0.922	0.893	0.162
NET	HN hiring-post share	d_yoy	0	27	21	-0.08	-0.49	+0.37	0.746	1.000	0.939	0.746
NET	HN hiring-post share	d_yoy	1	26	23	+0.23	-0.20	+0.59	0.282	1.000	0.893	0.290
NET	HN hiring-post share	d_yoy	2	25	22	-0.14	-0.53	+0.29	0.517	1.000	0.939	0.375
NET	HN hiring-post share	d_yoy	3	24	21	-0.10	-0.50	+0.35	0.679	1.000	0.939	0.640
NET	HN hiring-post share	d_yoy	4	23	23	-0.13	-0.51	+0.30	0.564	1.000	0.939	0.572
SNOW	HN hiring-post share	d_yoy	0	23	20	+0.12	-0.33	+0.53	0.600	1.000	0.939	0.708
SNOW	HN hiring-post share	d_yoy	1	22	20	+0.21	-0.26	+0.59	0.379	1.000	0.939	0.310
SNOW	HN hiring-post share	d_yoy	2	21	19	+0.03	-0.43	+0.47	0.909	1.000	1.000	0.890
SNOW	HN hiring-post share	d_yoy	3	20	18	-0.14	-0.57	+0.35	0.587	1.000	0.939	0.618
SNOW	HN hiring-post share	d_yoy	4	19	16	+0.18	-0.34	+0.61	0.505	1.000	0.939	0.580
CFLT	HN hiring-post share	yoy	0	19	7	+0.36	-0.55	+0.88	0.440	1.000	0.715	0.357
CFLT	HN hiring-post share	yoy	1	18	6	+0.33	-0.64	+0.90	0.513	1.000	0.715	0.326
CFLT	HN hiring-post share	yoy	2	17	6	+0.36	-0.65	+0.91	0.491	1.000	0.715	0.331
CFLT	HN hiring-post share	yoy	3	16	6	+0.38	-0.61	+0.91	0.449	1.000	0.715	0.314
CFLT	HN hiring-post share	yoy	4	15	6	+0.44	-0.65	+0.94	0.422	1.000	0.715	0.286
DDOG	HN hiring-post share	yoy	0	28	28	+0.05	-0.33	+0.42	0.792	1.000	0.851	0.849
DDOG	HN hiring-post share	yoy	1	27	27	+0.12	-0.27	+0.48	0.552	1.000	0.730	0.642
DDOG	HN hiring-post share	yoy	2	26	26	+0.14	-0.27	+0.50	0.508	1.000	0.715	0.595
DDOG	HN hiring-post share	yoy	3	25	25	+0.12	-0.29	+0.49	0.568	1.000	0.730	0.621
DDOG	HN hiring-post share	yoy	4	24	24	+0.15	-0.27	+0.52	0.476	1.000	0.715	0.554
ESTC	HN hiring-post share	yoy	0	32	28	+0.13	-0.26	+0.48	0.516	1.000	0.715	0.654
ESTC	HN hiring-post share	yoy	1	31	28	+0.15	-0.24	+0.49	0.452	1.000	0.715	0.598
ESTC	HN hiring-post share	yoy	2	30	30	+0.22	-0.15	+0.54	0.236	1.000	0.715	0.393
ESTC	HN hiring-post share	yoy	3	29	18	+0.20	-0.30	+0.61	0.436	1.000	0.715	0.486
ESTC	HN hiring-post share	yoy	4	28	16	+0.19	-0.33	+0.62	0.479	1.000	0.715	0.506
MDB	HN hiring-post share	yoy	0	36	36	+0.21	-0.12	+0.51	0.211	1.000	0.715	0.272
MDB	HN hiring-post share	yoy	1	35	35	+0.27	-0.07	+0.56	0.113	0.565	0.715	0.167
MDB	HN hiring-post share	yoy	2	34	34	+0.18	-0.17	+0.49	0.320	1.000	0.715	0.343
MDB	HN hiring-post share	yoy	3	33	33	+0.18	-0.18	+0.49	0.322	1.000	0.715	0.358
MDB	HN hiring-post share	yoy	4	32	32	+0.24	-0.12	+0.54	0.195	0.976	0.715	0.244
NET	HN hiring-post share	yoy	0	28	28	+0.06	-0.32	+0.42	0.772	1.000	0.851	0.825
NET	HN hiring-post share	yoy	1	27	27	+0.05	-0.34	+0.42	0.808	1.000	0.851	0.866
NET	HN hiring-post share	yoy	2	26	26	-0.05	-0.43	+0.35	0.825	1.000	0.851	0.864
NET	HN hiring-post share	yoy	3	25	25	-0.08	-0.46	+0.33	0.721	1.000	0.846	0.757
NET	HN hiring-post share	yoy	4	24	24	-0.08	-0.47	+0.33	0.705	1.000	0.846	0.754
SNOW	HN hiring-post share	yoy	0	24	24	+0.06	-0.35	+0.45	0.783	1.000	0.851	0.888
SNOW	HN hiring-post share	yoy	1	23	23	+0.09	-0.34	+0.48	0.694	1.000	0.846	0.857
SNOW	HN hiring-post share	yoy	2	22	22	+0.05	-0.38	+0.46	0.835	1.000	0.851	0.927
SNOW	HN hiring-post share	yoy	3	21	21	-0.02	-0.45	+0.42	0.933	1.000	0.933	0.968
SNOW	HN hiring-post share	yoy	4	20	20	-0.07	-0.50	+0.38	0.763	1.000	0.851	0.908
CFLT	npm downloads	d_yoy	0	2	–	–	–	–	–	–	–	–
CFLT	npm downloads	d_yoy	1	1	–	–	–	–	–	–	–	–
CFLT	npm downloads	d_yoy	2	0	–	–	–	–	–	–	–	–
CFLT	npm downloads	d_yoy	3	0	–	–	–	–	–	–	–	–
CFLT	npm downloads	d_yoy	4	0	–	–	–	–	–	–	–	–
DDOG	npm downloads	d_yoy	0	27	26	-0.02	-0.40	+0.37	0.933	1.000	1.000	0.941
DDOG	npm downloads	d_yoy	1	26	25	+0.13	-0.28	+0.50	0.547	1.000	0.939	0.520
DDOG	npm downloads	d_yoy	2	25	22	+0.35	-0.09	+0.67	0.113	0.566	0.893	0.243
DDOG	npm downloads	d_yoy	3	24	16	+0.37	-0.15	+0.73	0.153	0.765	0.893	0.169
DDOG	npm downloads	d_yoy	4	23	16	+0.21	-0.33	+0.64	0.444	1.000	0.939	0.429
ESTC	npm downloads	d_yoy	0	25	25	+0.31	-0.10	+0.63	0.132	0.660	0.893	0.179
ESTC	npm downloads	d_yoy	1	24	24	+0.04	-0.37	+0.44	0.845	1.000	1.000	0.862
ESTC	npm downloads	d_yoy	2	23	23	+0.03	-0.38	+0.44	0.881	1.000	1.000	0.805
ESTC	npm downloads	d_yoy	3	22	20	-0.39	-0.71	+0.07	0.091	0.455	0.893	0.193
ESTC	npm downloads	d_yoy	4	21	20	-0.25	-0.63	+0.22	0.292	1.000	0.893	0.371
MDB	npm downloads	d_yoy	0	35	35	+0.06	-0.28	+0.38	0.751	1.000	0.939	0.771
MDB	npm downloads	d_yoy	1	34	34	+0.12	-0.22	+0.44	0.490	1.000	0.939	0.459
MDB	npm downloads	d_yoy	2	33	25	+0.14	-0.27	+0.51	0.498	1.000	0.939	0.430
MDB	npm downloads	d_yoy	3	32	25	+0.23	-0.18	+0.57	0.268	1.000	0.893	0.155
MDB	npm downloads	d_yoy	4	31	24	-0.28	-0.61	+0.13	0.175	0.874	0.893	0.163
NET	npm downloads	d_yoy	0	27	27	+0.04	-0.34	+0.41	0.838	1.000	1.000	0.803
NET	npm downloads	d_yoy	1	26	26	-0.11	-0.48	+0.29	0.591	1.000	0.939	0.614
NET	npm downloads	d_yoy	2	25	25	+0.08	-0.32	+0.46	0.689	1.000	0.939	0.719
NET	npm downloads	d_yoy	3	24	24	-0.14	-0.51	+0.28	0.525	1.000	0.939	0.467
NET	npm downloads	d_yoy	4	23	22	-0.28	-0.63	+0.16	0.201	1.000	0.893	0.169
SNOW	npm downloads	d_yoy	0	23	21	+0.28	-0.17	+0.63	0.218	1.000	0.893	0.289
SNOW	npm downloads	d_yoy	1	22	21	+0.27	-0.18	+0.62	0.235	1.000	0.893	0.275
SNOW	npm downloads	d_yoy	2	21	20	+0.08	-0.37	+0.51	0.724	1.000	0.939	0.736
SNOW	npm downloads	d_yoy	3	20	19	+0.35	-0.12	+0.69	0.139	0.695	0.893	0.245
SNOW	npm downloads	d_yoy	4	19	17	-0.03	-0.50	+0.46	0.923	1.000	1.000	0.919
CFLT	npm downloads	yoy	0	3	–	–	–	–	–	–	–	–
CFLT	npm downloads	yoy	1	2	–	–	–	–	–	–	–	–
CFLT	npm downloads	yoy	2	1	–	–	–	–	–	–	–	–
CFLT	npm downloads	yoy	3	0	–	–	–	–	–	–	–	–
CFLT	npm downloads	yoy	4	0	–	–	–	–	–	–	–	–
DDOG	npm downloads	yoy	0	28	4	+0.75	-0.80	+1.00	0.268	1.000	0.715	0.015
DDOG	npm downloads	yoy	1	27	4	+0.78	-0.82	+1.00	0.250	1.000	0.715	0.010
DDOG	npm downloads	yoy	2	26	4	+0.81	-0.84	+1.00	0.241	1.000	0.715	<0.001
DDOG	npm downloads	yoy	3	25	4	+0.78	-0.90	+1.00	0.285	1.000	0.715	0.008
DDOG	npm downloads	yoy	4	24	4	+0.76	-0.80	+1.00	0.259	1.000	0.715	0.029
ESTC	npm downloads	yoy	0	26	3	+0.67	–	–	0.493	1.000	0.715	0.015
ESTC	npm downloads	yoy	1	25	3	+0.63	–	–	0.567	1.000	0.730	0.027
ESTC	npm downloads	yoy	2	24	3	+0.65	–	–	0.473	1.000	0.715	0.006
ESTC	npm downloads	yoy	3	23	3	+0.66	–	–	0.492	1.000	0.715	0.024
ESTC	npm downloads	yoy	4	22	3	+0.77	–	–	–	–	–	0.008
MDB	npm downloads	yoy	0	36	11	+0.32	-0.34	+0.77	0.330	1.000	0.715	0.178
MDB	npm downloads	yoy	1	35	10	+0.53	-0.18	+0.88	0.128	0.641	0.715	0.014
MDB	npm downloads	yoy	2	34	8	+0.56	-0.21	+0.90	0.135	0.674	0.715	0.011
MDB	npm downloads	yoy	3	33	8	+0.55	-0.24	+0.90	0.152	0.759	0.715	0.012
MDB	npm downloads	yoy	4	32	8	+0.48	-0.33	+0.88	0.224	1.000	0.715	0.043
NET	npm downloads	yoy	0	28	7	+0.60	-0.34	+0.94	0.178	0.890	0.715	0.090
NET	npm downloads	yoy	1	27	6	+0.52	-0.53	+0.94	0.300	1.000	0.715	0.150
NET	npm downloads	yoy	2	26	6	+0.45	-0.61	+0.93	0.396	1.000	0.715	0.238
NET	npm downloads	yoy	3	25	6	+0.35	-0.67	+0.91	0.510	1.000	0.715	0.381
NET	npm downloads	yoy	4	24	6	+0.27	-0.71	+0.89	0.617	1.000	0.774	0.509
SNOW	npm downloads	yoy	0	24	4	+0.81	-0.54	+0.99	0.155	0.774	0.715	0.008
SNOW	npm downloads	yoy	1	23	4	+0.89	-0.58	+1.00	0.126	0.628	0.715	0.003
SNOW	npm downloads	yoy	2	22	4	+0.89	-0.69	+1.00	0.139	0.694	0.715	0.001
SNOW	npm downloads	yoy	3	21	4	+0.89	-0.82	+1.00	0.168	0.841	0.715	0.001
SNOW	npm downloads	yoy	4	20	3	+0.83	–	–	0.271	1.000	0.715	0.001
CFLT	PyPI downloads	d_yoy	0	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	d_yoy	1	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	d_yoy	2	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	d_yoy	3	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	d_yoy	4	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	d_yoy	0	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	d_yoy	1	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	d_yoy	2	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	d_yoy	3	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	d_yoy	4	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	d_yoy	0	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	d_yoy	1	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	d_yoy	2	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	d_yoy	3	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	d_yoy	4	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	d_yoy	0	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	d_yoy	1	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	d_yoy	2	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	d_yoy	3	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	d_yoy	4	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	d_yoy	0	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	d_yoy	1	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	d_yoy	2	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	d_yoy	3	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	d_yoy	4	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	d_yoy	0	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	d_yoy	1	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	d_yoy	2	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	d_yoy	3	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	d_yoy	4	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	yoy	0	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	yoy	1	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	yoy	2	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	yoy	3	0	–	–	–	–	–	–	–	–
CFLT	PyPI downloads	yoy	4	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	yoy	0	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	yoy	1	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	yoy	2	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	yoy	3	0	–	–	–	–	–	–	–	–
DDOG	PyPI downloads	yoy	4	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	yoy	0	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	yoy	1	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	yoy	2	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	yoy	3	0	–	–	–	–	–	–	–	–
ESTC	PyPI downloads	yoy	4	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	yoy	0	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	yoy	1	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	yoy	2	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	yoy	3	0	–	–	–	–	–	–	–	–
MDB	PyPI downloads	yoy	4	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	yoy	0	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	yoy	1	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	yoy	2	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	yoy	3	0	–	–	–	–	–	–	–	–
NET	PyPI downloads	yoy	4	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	yoy	0	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	yoy	1	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	yoy	2	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	yoy	3	0	–	–	–	–	–	–	–	–
SNOW	PyPI downloads	yoy	4	0	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	d_yoy	0	2	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	d_yoy	1	1	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	d_yoy	2	0	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	d_yoy	3	0	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	d_yoy	4	0	–	–	–	–	–	–	–	–
DDOG	npm vs baseline (post-hoc)	d_yoy	0	27	26	-0.12	-0.49	+0.28	0.544	1.000	0.993	0.567
DDOG	npm vs baseline (post-hoc)	d_yoy	1	26	25	+0.07	-0.33	+0.45	0.738	1.000	0.993	0.737
DDOG	npm vs baseline (post-hoc)	d_yoy	2	25	25	+0.27	-0.14	+0.60	0.197	0.984	0.827	0.358
DDOG	npm vs baseline (post-hoc)	d_yoy	3	24	19	+0.35	-0.12	+0.69	0.143	0.716	0.827	0.201
DDOG	npm vs baseline (post-hoc)	d_yoy	4	23	19	+0.14	-0.34	+0.56	0.576	1.000	0.993	0.606
ESTC	npm vs baseline (post-hoc)	d_yoy	0	25	25	+0.27	-0.15	+0.60	0.199	0.993	0.827	0.279
ESTC	npm vs baseline (post-hoc)	d_yoy	1	24	22	+0.01	-0.41	+0.43	0.957	1.000	0.993	0.963
ESTC	npm vs baseline (post-hoc)	d_yoy	2	23	21	+0.02	-0.41	+0.45	0.918	1.000	0.993	0.869
ESTC	npm vs baseline (post-hoc)	d_yoy	3	22	17	-0.39	-0.73	+0.10	0.112	0.562	0.827	0.168
ESTC	npm vs baseline (post-hoc)	d_yoy	4	21	18	-0.23	-0.63	+0.27	0.362	1.000	0.919	0.421
MDB	npm vs baseline (post-hoc)	d_yoy	0	35	35	-0.10	-0.42	+0.24	0.557	1.000	0.993	0.564
MDB	npm vs baseline (post-hoc)	d_yoy	1	34	34	-0.00	-0.34	+0.34	0.988	1.000	0.993	0.989
MDB	npm vs baseline (post-hoc)	d_yoy	2	33	29	+0.00	-0.37	+0.37	0.991	1.000	0.993	0.990
MDB	npm vs baseline (post-hoc)	d_yoy	3	32	28	+0.12	-0.26	+0.47	0.537	1.000	0.993	0.462
MDB	npm vs baseline (post-hoc)	d_yoy	4	31	27	-0.33	-0.63	+0.05	0.090	0.452	0.827	0.086
NET	npm vs baseline (post-hoc)	d_yoy	0	27	27	+0.00	-0.38	+0.38	0.993	1.000	0.993	0.992
NET	npm vs baseline (post-hoc)	d_yoy	1	26	26	-0.18	-0.53	+0.22	0.367	1.000	0.919	0.401
NET	npm vs baseline (post-hoc)	d_yoy	2	25	25	+0.06	-0.34	+0.45	0.770	1.000	0.993	0.806
NET	npm vs baseline (post-hoc)	d_yoy	3	24	24	-0.10	-0.49	+0.31	0.630	1.000	0.993	0.547
NET	npm vs baseline (post-hoc)	d_yoy	4	23	22	-0.32	-0.65	+0.11	0.143	0.715	0.827	0.097
SNOW	npm vs baseline (post-hoc)	d_yoy	0	23	23	+0.10	-0.33	+0.49	0.662	1.000	0.993	0.685
SNOW	npm vs baseline (post-hoc)	d_yoy	1	22	22	+0.22	-0.23	+0.58	0.334	1.000	0.919	0.392
SNOW	npm vs baseline (post-hoc)	d_yoy	2	21	21	-0.06	-0.48	+0.38	0.783	1.000	0.993	0.809
SNOW	npm vs baseline (post-hoc)	d_yoy	3	20	20	+0.27	-0.20	+0.64	0.251	1.000	0.897	0.368
SNOW	npm vs baseline (post-hoc)	d_yoy	4	19	19	-0.00	-0.46	+0.45	0.988	1.000	0.993	0.986
CFLT	npm vs baseline (post-hoc)	yoy	0	3	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	yoy	1	2	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	yoy	2	1	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	yoy	3	0	–	–	–	–	–	–	–	–
CFLT	npm vs baseline (post-hoc)	yoy	4	0	–	–	–	–	–	–	–	–
DDOG	npm vs baseline (post-hoc)	yoy	0	28	4	+0.68	-0.83	+0.99	0.324	1.000	0.516	0.030
DDOG	npm vs baseline (post-hoc)	yoy	1	27	4	+0.71	-0.83	+0.99	0.307	1.000	0.516	0.026
DDOG	npm vs baseline (post-hoc)	yoy	2	26	4	+0.74	-0.85	+1.00	0.295	1.000	0.516	0.009
DDOG	npm vs baseline (post-hoc)	yoy	3	25	4	+0.72	-0.89	+1.00	0.330	1.000	0.516	0.014
DDOG	npm vs baseline (post-hoc)	yoy	4	24	4	+0.71	-0.80	+0.99	0.293	1.000	0.516	0.041
ESTC	npm vs baseline (post-hoc)	yoy	0	26	4	+0.63	-0.97	+1.00	0.457	1.000	0.546	0.013
ESTC	npm vs baseline (post-hoc)	yoy	1	25	3	+0.57	–	–	0.525	1.000	0.570	0.056
ESTC	npm vs baseline (post-hoc)	yoy	2	24	4	+0.59	-0.94	+1.00	0.465	1.000	0.546	0.012
ESTC	npm vs baseline (post-hoc)	yoy	3	23	4	+0.60	-0.96	+1.00	0.475	1.000	0.546	0.031
ESTC	npm vs baseline (post-hoc)	yoy	4	22	3	+0.73	–	–	0.400	1.000	0.546	0.017
MDB	npm vs baseline (post-hoc)	yoy	0	36	16	-0.40	-0.75	+0.13	0.130	0.649	0.361	0.093
MDB	npm vs baseline (post-hoc)	yoy	1	35	15	-0.41	-0.76	+0.13	0.129	0.643	0.361	0.107
MDB	npm vs baseline (post-hoc)	yoy	2	34	12	-0.50	-0.83	+0.09	0.089	0.445	0.361	0.041
MDB	npm vs baseline (post-hoc)	yoy	3	33	10	-0.53	-0.86	+0.13	0.104	0.521	0.361	0.023
MDB	npm vs baseline (post-hoc)	yoy	4	32	9	-0.57	-0.89	+0.13	0.101	0.507	0.361	0.017
NET	npm vs baseline (post-hoc)	yoy	0	28	7	+0.53	-0.40	+0.92	0.237	1.000	0.516	0.139
NET	npm vs baseline (post-hoc)	yoy	1	27	6	+0.42	-0.56	+0.91	0.386	1.000	0.546	0.259
NET	npm vs baseline (post-hoc)	yoy	2	26	6	+0.34	-0.60	+0.89	0.480	1.000	0.546	0.370
NET	npm vs baseline (post-hoc)	yoy	3	25	7	+0.24	-0.64	+0.85	0.607	1.000	0.632	0.541
NET	npm vs baseline (post-hoc)	yoy	4	24	7	+0.16	-0.69	+0.82	0.739	1.000	0.739	0.695
SNOW	npm vs baseline (post-hoc)	yoy	0	24	5	+0.78	-0.29	+0.98	0.112	0.562	0.361	0.022
SNOW	npm vs baseline (post-hoc)	yoy	1	23	5	+0.80	-0.28	+0.99	0.106	0.529	0.361	0.017
SNOW	npm vs baseline (post-hoc)	yoy	2	22	5	+0.80	-0.28	+0.99	0.105	0.527	0.361	0.021
SNOW	npm vs baseline (post-hoc)	yoy	3	21	5	+0.81	-0.31	+0.99	0.108	0.542	0.361	0.028
SNOW	npm vs baseline (post-hoc)	yoy	4	20	5	+0.76	-0.47	+0.99	0.161	0.804	0.402	0.027
CFLT	PyPI vs baseline (post-hoc)	d_yoy	0	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	d_yoy	1	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	d_yoy	2	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	d_yoy	3	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	d_yoy	4	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	d_yoy	0	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	d_yoy	1	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	d_yoy	2	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	d_yoy	3	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	d_yoy	4	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	d_yoy	0	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	d_yoy	1	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	d_yoy	2	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	d_yoy	3	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	d_yoy	4	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	d_yoy	0	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	d_yoy	1	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	d_yoy	2	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	d_yoy	3	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	d_yoy	4	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	d_yoy	0	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	d_yoy	1	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	d_yoy	2	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	d_yoy	3	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	d_yoy	4	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	d_yoy	0	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	d_yoy	1	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	d_yoy	2	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	d_yoy	3	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	d_yoy	4	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	yoy	0	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	yoy	1	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	yoy	2	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	yoy	3	0	–	–	–	–	–	–	–	–
CFLT	PyPI vs baseline (post-hoc)	yoy	4	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	yoy	0	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	yoy	1	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	yoy	2	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	yoy	3	0	–	–	–	–	–	–	–	–
DDOG	PyPI vs baseline (post-hoc)	yoy	4	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	yoy	0	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	yoy	1	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	yoy	2	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	yoy	3	0	–	–	–	–	–	–	–	–
ESTC	PyPI vs baseline (post-hoc)	yoy	4	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	yoy	0	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	yoy	1	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	yoy	2	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	yoy	3	0	–	–	–	–	–	–	–	–
MDB	PyPI vs baseline (post-hoc)	yoy	4	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	yoy	0	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	yoy	1	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	yoy	2	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	yoy	3	0	–	–	–	–	–	–	–	–
NET	PyPI vs baseline (post-hoc)	yoy	4	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	yoy	0	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	yoy	1	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	yoy	2	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	yoy	3	0	–	–	–	–	–	–	–	–
SNOW	PyPI vs baseline (post-hoc)	yoy	4	0	–	–	–	–	–	–	–	–

Downloads: leadlag_company.csv · leadlag_pooled.csv · quarterly_panel.csv · revenue_quarterly.csv. Columns: p = effective-n t-test; p_lag_adj = Bonferroni over lags; q_bh = Benjamini-Hochberg; p_perm = block permutation (secondary; anti-conservative on persistent series).