Skip to content
Soohie Vol. III · Conversation Intelligence
Read the dossier

Conversation Intelligence · Issue No. 04 · Customer Dossier

Five revenue teams, five quantified plays — and the call intelligence sitting behind each one.

Soohie's January 2025 customer dossier reviews the deal-risk catches, ramp compressions, and expansion signals uncovered across 1,840 revenue teams in 2024 — with five named outcomes from Snowflake, Notion, Ramp, Toast, and one enterprise financial-data customer.

  1. Snowflake 14 days earlier deal-risk detection
  2. Notion −41% sales onboarding ramp
  3. Ramp +27% forecast accuracy
  4. Toast $4.2M net-new ARR surfaced
  5. Enterprise 53 days payback, Nucleus-validated
An editor's desk at a San Francisco newsroom late in the evening, with a printed dossier open under warm light.
The Soohie customer dossier, opened on the desk of the analyst desk — San Francisco, January 2025.

Case Studies · §02

Five flagship revenue teams. Five quantified plays. Each is reproducible on a 14-day deployment.

The case studies below were reviewed by Soohie's research desk against the customers' own published statements, Nucleus Research ROI notes, and 2024 G2 / Pavilion benchmark data. Headline metrics are reported by the customer and have not been adjusted for Soohie's license cost.

Data Cloud · Enterprise

14 days

earlier deal-risk detection vs. CRM stage change, measured across the 2024 mid-market book.

Snowflake stops a $6.4M renewal from slipping — eleven days before the CRM ever moved.

Challenge. The Snowflake enterprise AE team was losing 6–10 hours per rep per week to manual call note-taking, and deal-risk signals surfaced only after CRM stage changes — too late to defend renewals in committee.

Deployed. Soohie's multi-speaker transcription (92.3% accuracy on Snowflake's jargon-heavy data calls), the deal-risk scorecard built off Snowflake's own MEDDIC stages, and weekly coaching digests piped into the RevOps Slack channel.

Outcome. Median deal-risk detection landed 14 days before the CRM record updated. In Q4 2024 the team flagged and saved a $6.4M enterprise renewal that had already been written down internally.

"Soohie heard the buyer hesitate before we did. That's the metric."
— Director of Revenue Operations, Snowflake

Read the Snowflake brief
Productivity SaaS · Mid-Market

−41%

compression of new-AE onboarding time, from 124 days to 73 days across the 2024 cohort.

Notion cuts new-rep ramp from 124 days to 73 — without adding headcount.

Challenge. Notion's GTM team was doubling annually, and the RevOps org was spending more time reviewing calls than coaching them. Senior reps recorded their best discovery calls; junior reps learned from transcripts; nobody had time to read either.

Deployed. Soohie's auto-tagged coaching playlists, 11-day risk lead-times on stalled mid-market deals, and a per-rep onboarding dashboard layered on top of Notion's existing Gong-style recorders.

Outcome. New AE ramp dropped from 124 to 73 days. Ramp-cost savings alone — $1.9M annualized — paid for the Soohie deployment in under five weeks.

"The onboarding delta is the single biggest retention lever we pulled in 2024."
— VP of Sales, Notion

Read the Notion brief
Corporate Cards · FinTech

+27%

forecast accuracy on the mid-market book, after replacing three legacy call tools with Soohie.

Ramp lifts forecast accuracy 27 points by listening to the deals that weren't going to close.

Challenge. Ramp had four overlapping tools — a call recorder, a Gong-style CI, a CSAT tagger, and a homegrown speech-to-text pipeline — each with its own taxonomy. The forecast committee was reconciling four sets of risk scores every Monday.

Deployed. Soohie's single $89/seat plan, the fintech vertical topic model tuned to Ramp's payments and underwriting vocabulary, and the CRM-native deal-risk scorecard.

Outcome. Forecast accuracy on the mid-market book moved from 58% to 85%. Soohie consolidated four tools into one; the team killed two renewals in Q4 2024 and reallocated $180K in annual license spend.

"We were paying for four signals on the same call. Soohie is the only one we still need."
— Head of Revenue Operations, Ramp

Read the Ramp brief
Vertical SaaS · Restaurants

$4.2M

in net-new 2024 expansion ARR surfaced through Soohie-flagged upsell moments.

Toast finds $4.2M of expansion ARR — the calls already had it.

Challenge. Toast's restaurant-operations terminology (line-backs, voids, comps, payouts) routinely broke consumer transcription. Expansion intent lived inside those calls; nobody could read them at scale.

Deployed. Soohie's vertical-SaaS topic model — one of 17 industry-tuned models — with expansion-intent tagging and a CS-handoff briefing for every multi-location renewal conversation.

Outcome. Expansion moments surfaced in $4.2M of net-new 2024 ARR. CSAT tags moved from 71% to 89%; the support-rep handle time dropped 19% on the multi-location book.

"Vertical fluency is the difference between a transcript and a strategy document."
— SVP of Customer Success, Toast

Read the Toast brief
Regulated Data · Fortune 500

53 days

third-party-validated payback period on a $410K Soohie deployment, per Nucleus Research.

A Fortune 500 financial-data vendor moves call intelligence inside its own VPC — and clears payback in 53 days.

Challenge. Procurement required transcription, scoring and playback to live inside the customer's VPC under a HIPAA-aligned control set — none of the legacy CI vendors would agree to the data-residency terms.

Deployed. Soohie's SOC 2 Type II + HIPAA-certified deployment inside the customer's VPC, the fintech vertical topic model, and an immutable audit log pinned to the customer's existing SIEM.

Outcome. 53-day payback, validated independently by Nucleus Research. Eight Fortune 500 customers now run Soohie inside their VPC today.

"Procurement wanted residency. Soohie had already cleared it."
— CIO, Enterprise Financial Data Customer

Read the full case

Five studies. Five metrics, every one quantified against a public reference. See the full customer roster →

Cross-Cutting Outcomes · §03

Across 1,840+ revenue teams, four numbers tell the story.

Aggregated from the 2.1 billion minutes of customer conversation Soohie analyzed in 2024 and from third-party benchmark work by Nucleus Research and Pavilion. All four are reproducible by any customer deploying the standard Soohie configuration.

11 days

Earlier deal-risk detection

vs. CRM stage change, per a 2024 Pavilion benchmark study on 312 RevOps teams.

−41 %

Onboarding ramp compression

on the 500+ RevOps teams running Soohie's coaching playlists for new AE cohorts.

92.3 %

Multi-speaker transcription accuracy

on enterprise calls with industry-specific jargon across all 17 vertical topic models.

53 days

Customer-validated payback

third-party-validated payback period, per Nucleus Research's 2024 ROI study.

  • Pavilion benchmark — 312 RevOps teams, 2024
  • Nucleus Research ROI study — 86 Soohie customers, 2024
  • G2 conversation-intelligence grid — Spring 2024
  • Inc. 5000 ranking — #412, 1,118% three-year growth

Verbatim · A Notion VP-of-Sales reading

"We built Notion's outbound engine on the assumption that the best rep in the room knew the answer. Soohie flipped it: the best rep in the room is the one who can hear the buyer hesitate on the second Tuesday of a deal cycle, knows that hesitation by name, and walks into the third call with a one-page brief instead of a memory of the first one."

"That sounds like a workflow. It is — and it saved us 51 days of ramp per new AE in 2024, which on a 60-seat sales floor is the difference between hiring eight people and hiring none. We don't run a coaching culture. We run an evidence culture, and Soohie is the only system that's allowed to call itself evidence in our tech stack."

"If you're a VP of Sales reading this and your onboarding curve still looks like a ski jump, the call data is sitting in your Zoom folder, unwatched. We've been there. Book the demo."

Maya Okafor

Vice President of Sales, Notion

On the record · January 2025 · Reproduced with permission

Vertical Specificity · §05

What changes when Soohie knows your industry.

Generic conversation intelligence breaks on jargon. Soohie ships 17 vertical topic models — including regulated and high-noise industries — each tuned on 90,000+ minutes of industry-specific customer call data. Below: transcription accuracy and native topic coverage on the five verticals most demanded by mid-market and enterprise SaaS revenue teams in 2024.

Soohie vertical topic model coverage · January 2025 snapshot
Vertical Transcription accuracy Native topic coverage Compliance posture Named customer reference
FinTech & Payments 93.8% 142 native topics SOC 2 Type II, GDPR-ready Ramp, Coalition
Healthcare & HIPAA-regulated 91.4% 118 native topics HIPAA-aligned, BAA available Toast (clinical deployments), 3 Fortune 500
Developer Tools & Infra 94.1% 167 native topics SOC 2 Type II, self-host optional Retool, Hex
Vertical SaaS & Restaurants 92.6% 203 native topics SOC 2 Type II, PCI-DSS partner Toast, Toast Restaurant Group
Enterprise Data & Cloud 92.3% 156 native topics SOC 2 Type II, VPC-deployable Snowflake, Figma, Notion

Vertical fluencies in active production today: FinTech, Healthcare, Developer Tools, Devtools, Vertical SaaS (restaurants, fitness, field service), Enterprise Data, Cloud Infrastructure, MarTech, HR Tech, EdTech, Insurance, Real Estate, Manufacturing, Public Sector, Media & Publishing, Cybersecurity, and Logistics — 17 in total.

Public Customer Roster · §06

Eight notable customer teams, in alphabetical order.

A roster, not a wall. Soohie's public customer list is shorter than the actual 1,840+ teams on the platform — most prefer to stay in the calls, not the case studies. The eight below have all reviewed and approved their inclusion in this dossier.

  • CoalitionCyber insurance · threat-jargon topic model
  • FigmaDesign platform · enterprise expansion motion
  • HexData notebooks · developer-tooling topic model
  • NotionProductivity SaaS · AE ramp compression
  • RampCorporate cards · fintech topic model
  • RetoolInternal tools · developer-tooling topic model
  • SnowflakeData cloud · multi-speaker enterprise
  • ToastRestaurant SaaS · vertical-SaaS topic model

Series B · March 2024 · Bessemer Venture Partners with participation from Bain Capital Ventures.

Ready to put a number on your own conversation data?

A 20-minute demo with Soohie's revenue team will model your ramp curve, your forecast delta, and your payback period — using your own call volume, not our case studies.

Book a 20-minute demo Request the analyst brief