Talkdesk identified high-value accounts using a predictive ICP model with advanced insights

Founded in 2011, Talkdesk is a cloud-based contact center software provider, which empowers companies to continuously improve their customer experience.

We ultimately embarked on this ICP project to identify and prioritize likely to purchase prospect accounts. With enriched company data and a predictive model backed by data science, we were able to uncover new opportunities and align our sales and marketing efforts on key accounts.

Brian AdamDirector of Marketing Operations, Talkdesk

Business challenges

Looking to identify high-value companies willing to invest in their contact center, Talkdesk needed to precisely define their Ideal Customer Profile (ICP) criteria and quickly prioritize the highest valued accounts.

Talkdesk wanted to align focus on their top prospects to fuel rapid growth - the challenge was operationalizing an ICP identification strategy that was both scalable and repeatable.

With Advanced Insights, we’re able to quickly surface companies indicative of our top customers and think strategically about significant attributes that will allow us to move upmarket.

Brian AdamDirector of Marketing Operations, Talkdesk

Why Talkdesk chose Oracle

  • Talkdesk needed a strategic partner to help them implement a robust ICP modeling process, in order to surface statistically significant ICP traits evident of high-value accounts.
  • Account Data Management  - Talkdesk enriched their account data by integrating human-verified and AI-powered company data directly into their CRM to provide a deeper and more accurate understanding of an account’s data scope.
  • Advanced Insights - A predictive ICP modeling process evaluated the statistical significance of account characteristics by running standard deviation analysis on over 4,000 attributes to surface influential traits evident of Talkdesk’s best customers.
  • Account Scoring - The Talkdesk team then incorporated these traits into an account scoring model, assigning each trait a score based on its significance. With nearly 40 key traits being evaluated for every account, Talkdesk was easily able to identify and prioritize their high-scoring accounts.


Talkdesk successfully implemented a predictive ICP identification process to align company resources on high-value accounts. Now, 55% of Talkdesk’s annual recurring revenue comes from the top 3.3% of their highest-scoring accounts.

Augmenting their intuitions with data science, Talkdesk is able to gain clarity on which attributes indicate a high-value account. This heightened understanding of their ICP definition, paired with an account scoring model, are the key drivers of Talkdesk’s ability to find and prioritize companies motivated to buy.

With enriched company data on every account and data science that analyzes the statistical significance of attributes across millions of companies, Talkdesk can position these insights to:

  • Eliminate wasted prospecting time by aligning sales and marketing efforts on key accounts
  • Improve ABM campaigns
  • Now, the highest scored 3.3% of accounts provide 55% of Talkdesk’s Annual Recurring Revenue
Published:March 12, 2019