Overview

Invoca is a revenue execution platform for B2C brands. Invoca AI connects the complete buyer journey across digital and conversational touchpoints to maximize revenue, increase conversion rates, and improve contact center sales performance. The company employs roughly 375 people and has been in business for 18 years.

Supernova Award Category

Future of Work: Employee Experience

The Problem

AI usage inside Invoca was widespread but fragmented. Teams were testing tools in isolation, with no shared documentation, workflow standards, or measurable accountability, so individual experiments rarely turned into company-wide capability. Leadership saw a clear risk: high tool usage that looked like progress, when real progress for an AI company means intelligently building AI into employees' day-to-day processes.

The Solution

Invoca's 27-person marketing team launched the AI Launch Lab, an AI-first operating model built on a simple six-step framework: define objectives, identify AI opportunities, execute with purpose-fit tools, document prompts and workflows, measure impact, and embed learnings into future processes. The team chose a real, high-stakes project as its first test case, the annual Call Conversion Benchmark Report, historically a two-month production effort, so the new model would be proven on flagship work rather than a side experiment. Tool selection followed the work rather than leading it, with public tools (ChatGPT, Claude) approved for non-confidential use, Google Gemini for company-confidential workflows, and agentic automation (Gumloop) for employees building production agents, all under clear governance: if you wouldn't publish it on the website, it stays out of open AI platforms.

The results

The Launch Lab proved itself on its first project. The Call Conversion Benchmark Report, which had taken two months to produce and shipped as a single edition, was rebuilt through the AI-first model in two days, with campaign planning falling from eight hours to fifteen minutes and nine verticalized editions launching at once instead of one. That result turned a marketing pilot into a company-wide mandate, and the same governed model began spreading across functions. AI Passion Week was the next evolution of the Lab. Where the Launch Lab embedded AI into one team's workflows, Passion Week gave every employee dedicated time and space to build: a five-day, all-employee sprint with internal meetings cleared, a build platform open to everyone, and hands-on support throughout. Employees from six functions, most of them non-technical, built 379 working AI agents that made 64,453 tool calls into production systems and executed 958,875 lines of code doing real work. The shift moved Invoca from using AI (summarizing meetings, drafting documents) to building with AI (wiring intelligence into the CRM, the data warehouse, the support queue, and the campaign workflows where revenue is won or lost).

Metrics

AI Launch Lab (marketing, within 90 days): - Benchmark report production: two months to two days. - Campaign planning: eight hours to fifteen minutes. - Editions shipped simultaneously: one to nine verticalized editions. - Immediate downloads: up 185 percent. - Earned media coverage: up 275 percent. - AI-driven account analysis enriched 1,400+ accounts, saving 130+ hours. - 100 percent of the marketing team adopted AI weekly under a governed model. AI Passion Week (five days, all employees): - 379 working AI agents built. - 64,453 tool calls into production systems. - 958,875 lines of code executed in the background. - Builders spanned six functions, most non-technical. - A sales pipeline risk reviewer was running in production by that Friday. - Marketing ops agent compressed display ad campaign creation from eight hours to one. - A solo employee compressed a week of keyword-spotting setup into twenty minutes.

The Technology

Agentic workflow automation (Gumloop) for employee-built production agents; Google Gemini for company-confidential workflows within Google Cloud; ChatGPT and Claude for approved non-confidential use; multiple internally developed and actively managed CustomGPT and Gemini Gem systems; and a centralized AI Productivity Kit and AI Resource Library (prompt repositories, workflow documentation, governance guidelines, and curated learning pathways).

Disruptive Factor

Most enterprises measure AI adoption the way they measured software adoption, by logins and active users, which rewards habit rather than capability. Invoca challenged that by measuring who builds something that changes how work gets done, and by treating capability-building as a budget line with a real price: a build platform license, protected calendar time with internal meetings cleared, a vendor partner reachable in Slack, internal AI office hours, and public permission to demo rough work. The hardest part was cultural rather than technical, because it required leaders to give non-engineers the tools, time, and trust to ship against live production systems inside clear governance boundaries. The result is a workforce that turns AI from a personal shortcut into shared operating leverage, which is the gap Invoca sees most B2C enterprises still failing to close.

Shining Moment

The proudest moments came from people no one would have predicted. During AI Passion Week, employees who had never written a line of code (an account manager, a finance lead, people from partner marketing) stood up and demoed working agents wired into real customer records and real Slack channels. The best ideas did not come from leadership planning sessions; they came from the employees closest to the work, who knew exactly where customer friction and revenue leakage hide. That is what Invoca is most proud of: an innovation culture where a great idea carries the same weight whether it comes from the CEO or someone in their first week, and where the company invests real money, time, and trust to help that idea ship.