18 competitors, one intelligence system. From battle cards to paid displacement.

At the start of 2025, competitive intelligence at a B2B communications platform was three stale battlecards in a Drive folder. Thirteen months later it was an 18-competitor repo, a CI Pricing Tracker, a published State of CI report, and a paid Aircall displacement campaign with deal-level attribution. The losses are on the record. That is the proof.

What it was

A communications platform competing against Aircall, RingCentral, Dialpad, CloudTalk, OpenPhone and a long tail of AI voice agents. The competitive surface was unowned. Sellers wrote their own battlecards in slack threads. Comparison LPs ran without anyone tracking them.

The diff

Before, feb 2025: three out-of-date battlecards in a Drive folder. No central repo. No measurement layer. No paid competitive channel.

After, mar 2026: 18 competitors tracked. Six battlecards live by Jun 2025. Twelve plus shipped in Jul 2025. A CI Pricing Tracker via Relay. A State of Competitive Intelligence report. Win/Loss automation via Context DB. A $48.6K paid Aircall campaign across five geos with weekly iteration.

What I actually did

1. Built the repo. Standing cadence, single source for every competitor file. By Jun 2025: 18 competitors tracked, six battlecards live, first attributed win at $550 MRR via the JustCall vs Aircall sheet.
2. Crossed competitive into measurable CRO. Jul 2025: JustCall vs Aircall LP +107.7% demos. JustCall vs Kixie LP +123.1% demo signups, +39.7% signup CR. Power Dialer +184.9% signups. Bulk SMS +123.2% demos. Same report: JustCall vs Dialpad LP -83.6% conversions, logged as the honest loss it was.
3. Built the measurement system. Aug-Dec 2025: CI Pricing Tracker live, State of CI report published (first internal), customGPT for sales-call CI extraction, baseline win rates established (Aircall 33%, RingCentral 10%, Dialpad 21%, CloudTalk 26%, Sales Impact Report Q3 2025).
4. Shipped the Aircall depositioning slides in Oct 2025. Built from G2 sentiment plus win/loss insight. Refreshed and turned into a takeout campaign by Dec 2025.
5. Ran competitive as a paid acquisition channel. Mar 2026: $48,600 spend, 25.3 conversions, geo-level budget control. AU at $1,099/conv. UK and Alt US paused on the same report that reported the wins. Twelve Aircall-influenced closed-won deals that month.

What stayed honest

The JustCall vs Dialpad -83.6% result is in this case because it was in the monthly report. The Mar 2026 campaign missed three of four targets and the report says so. Spend tripled, conversions fell 3%, and the root cause (Quality Score 1/10 on top keywords inflating every click) is named in the same document that names the 12 closed-won deals. Numbers are self-reported, sourced to monthly PMM reports and the Q3 Sales Impact Report. Anonymized while still employed; the flip switch restores company names on resignation.

What it became, in substrate

• skills/competitive-repository. single-repo model, standing cadence, monthly shipping rhythm.
• skills/comparison-lp-cro. value-first vs feature-first split rule from the Jul 2025 insight.
• skills/honest-loss-logging. every LP test reported with the win or the loss, both at the same place.
• patterns/competitive-as-paid-channel. geo-budgeted displacement with weekly iteration and a kill rule.

What the work looks like

Two slides from the Apr 2026 competitive convergence read. Full deck on request under hiring-stage NDA.