Gladius Research
Original research from our pilot rooftops, plus curated public dealer industry data. We publish what we can defend — sourced, dated, and honest about sample size.
Category 01 · Pilot Data
We have three rooftops in pilot across three verticals and about ninety days of data. That is a research-grade sample of very nearly nothing. We publish what we see anyway, framed as directional — the patterns are real even when the magnitudes are not yet defensible.
n=3 rooftops · ~90 days of data · statistically thin · treat as directional, not authoritative
Directional read on first-touch latency across three pilot rooftops in the auto, detailing, and landscaping verticals. We see the patterns we expected — the long tail of after-hours leads is the leak — but the sample is too thin to publish a curve.
Read summary →How service-package attach rates move across auto F&I, recon detail upsells, and landscaping maintenance plans. Directional only — the comparison is interesting; the absolute numbers are too dealer-specific to generalize.
Read summary →What lapsed customers do when re-engaged via cross-vertical signals (a service customer pinged about a trade window, a detail customer pinged about delivery prep). ETA Q3 2026 — collecting baseline now.
Summary · Pilot speed-to-lead
Across the three pilot rooftops, the lead-arrival distribution is heavily back-loaded toward the evening and weekend. The shape we observe matches what a 20-year operator would predict — the long after-hours tail is where qualified buyers fall through. We do not yet publish a specific first-touch curve; the sample is too small to be load-bearing. Pattern direction: clear. Curve shape: forthcoming.
Sample: 3 rooftops · Window: Q1 2026 · Source: Gladius pilot platform telemetry
Summary · F&I / upsell attach variance
Across the three pilot verticals (auto F&I, detail recon upsells, landscaping maintenance plans) the variance in attach rate between rooftops dwarfs the variance between product categories. The implication is the one operators already know: attach is about coaching and process discipline, not about the product menu. We will not publish absolute numbers until n is larger; the directional observation stands on its own.
Sample: 3 rooftops · Window: Q1 2026 · Source: Gladius pilot platform telemetry
More pilot data published as we accumulate it. Email research@gladiuscrm.com to be notified when the next release drops.
Category 02 · Public-Source Digests
These digests pull together what is already on the public record — Cox Automotive annual surveys, NADA data, dealer-trade reporting, 10-K filings, breach disclosures. Every figure is sourced. Nothing extrapolated. If a digest isn’t live yet, it’s flagged as in-progress with an ETA, not faked into a tile.
A single-page synthesis of what Cox Automotive, NADA, and the dealer-trade press have said publicly about the CRM category in the last 12 months. Every figure cited; nothing extrapolated.
The two-vendor DMS oligopoly (CDK and Reynolds), the smaller-vendor long tail, and a timeline of public-record incidents — including the June 2024 CDK ransomware event that took ~15,000 rooftops offline for ten days. Sourced to filings and reporting.
A read on what dealer groups have published about their post-2023-amendment Safeguards posture (written security program, qualified individual, MFA, encryption). Built from public statements and 10-K filings — not from any private channel.
“The CDK June 2024 incident is the most expensive cybersecurity lesson the auto industry has on the public record. It belongs in any honest digest of the DMS market.”
Public-record reference · reporting collated from Reuters, the Wall Street Journal, Automotive News
Category 03 · Live Operational Metrics
Uptime, attacks blocked, latency, federation events — the operational telemetry that runs the Gladius platform is available to anyone with the URL. No login. No redaction. These are not research papers; they are live feeds. Cross-link only.
Live feed
Live uptime across all five Gladius verticals (auto, detailing, turf, stone, BDC). 30-day rolling window.
Open feed →
Live feed
Last-24-hour attack log — every blocked request, every rule that fired, every signature that matched. The opposite of how most CRMs handle security disclosure.
Open feed →
Live feed
SVG status badges any partner, dealer, or journalist can embed. Refreshes on every deploy.
Open feed →
Methodology
We anonymize and aggregate at the cross-tenant level. We do not re-identify a dealer, a rep, or a customer in anything we publish. A pilot finding is always either a cross-rooftop pattern or a cohort-level shape — never a single dealer’s number with the name scrubbed off.
Public sources only. Cox Automotive Annual Dealer Survey, NADA Data, 10-K filings, breach disclosures, mainstream reporting. Every figure cited inline. Nothing proprietary leaked into a digest, ever.
Large-sample statistical power. We’re early. Treat findings as directional. We will publish confidence intervals when n is large enough to justify them. Until then, the magnitudes you read here are shape, not score.
We will not put a percentage on a slide we cannot source. We will not extrapolate from three rooftops to an industry. We will not republish a vendor white paper as research. If a digest is unfinished, the tile says “in progress, ETA” — we do not fake the finding to fill the grid.
Participate
If you operate a rooftop and would like to contribute anonymized data to our research — with full control over what is shared — reach out. Contributors get full access to the aggregated findings before public release.
We’re especially interested in
Direct contact
Press kit, brand assets, and on-the-record commentary are also linked from /press. Live platform telemetry lives at /status, /awais/incidents, and /status/badges.