Demo modeIllustrative and simulated market data. No live integrations. Metrics about named companies are public-signal examples only.Methodology
Applied researchDemo data

SIGNAL LABS

Each lab starts with a question a market cannot answer, states a measurable baseline, runs a defined experiment and publishes the result — including the results that disappoint the sponsor.

About this sponsorship

MarketPulse funds selected experiments. Baseline, method and result are published unchanged, including where the finding is unflattering to a sponsor.

Lab 001 · Public

Can AI reduce insurance quotation time from 20 minutes to 60 seconds?

Objective

Compress a multi-step quotation journey into a single conversational flow.

Baseline

20m 14s median quotation time (demo panel, n=18)

Experiment

Conversational intake + risk pre-fill + instant indicative premium

Result so far

68s median indicative quote; 41% completion lift

Learning

Pre-fill accuracy mattered more than conversational tone.

Progress
6 weeks · week 4 of 666
Sponsorship

Open for sponsorship

Running
Lab 002 · Public

Can a dealership answer every lead in under 30 seconds?

Objective

Eliminate after-hours lead leakage across a dealer network.

Baseline

8h 14m first useful response (Response Race, n=24)

Experiment

Automated acknowledgement, qualification and morning routing

Result so far

12s median acknowledgement, 97% coverage

Learning

Speed of acknowledgement beat completeness of response on conversion.

Progress
3 weeks · complete100
Sponsorship

Sponsored — automotive group (demo)

Complete — evidence published
Lab 003 · Private pilot

Can AI improve university admission enquiries?

Objective

Reduce unanswered admission questions during peak intake.

Baseline

1,284 unanswered public questions in intake window

Experiment

Answer library + escalation routing + weekly gap review

Result so far

Baseline established; first answer library shipped

Learning

Most unanswered questions cluster into 14 repeatable themes.

Progress
8 weeks · week 2 of 822
Sponsorship

Open for sponsorship

Running
Sponsorship rules

What a sponsor gets and does not get

  • · Gets: named disclosure, early briefing on publication day, the underlying method
  • · Does not get: influence over the result, veto over publication, exclusivity on the finding
  • · A lab that fails is still published, with the failure explained
Propose

Suggest a lab

The strongest labs come from an unanswered customer question with a measurable baseline. If you can state the baseline, we can run the experiment.

See sponsorship options →
SponsoredIndependence disclosure

Signal Labs carries clearly labelled paid placement. This is paid promotional placement. Sponsorship does not affect independent rankings, scores, evidence or methodology.

Independent signal rankingEvery score, rank and evidence record on this surface is calculated from evidence only.