The work was there. The system wasn’t.
Attention’s routing was bound to territory and segmentation. It had no way to account for personal fit between a specific buyer and AE: shared hometown, school, interests, or the kind of company and buyer an AE naturally understood. The broader problem was speed: a one-off analysis meant building a table by hand and connecting systems that were not designed to answer the question together.
A sales leader’s pattern, made testable
Jacob Fleisher, Attention’s Head of Sales, noticed that the calls which clicked often had something human underneath them: a shared city, a football team, a university, or deep familiarity with a buyer’s vertical. Territory assignment could not see that signal.
The goal was 0-to-1, not another dashboard
Anis Bennaceur uses Claude Code and APIs to explore questions quickly, then hands successful experiments to the growth team to make them scalable. Before Deepline, connecting CRM, conversation, and external data for that kind of experiment was painful; the team had to decide the path and build the table first.
From scattered judgment to an operating system.
The result was not another dashboard. It was a durable way to find, decide, and act on GTM signals.
Finding the patterns behind fit
Attention joined CRM and conversation data with firmographics, deal amount, stage, win/loss outcomes, and close dates. Deepline added contact and account enrichment, including technographics and public-profile signal. Claude Code could reason across the combined data, testing patterns in geography, interests, past experience, and vertical fluency, not just one guided query.
Putting evidence into the booking flow
After a historical pass across the rep base, Attention built new account-scoring models and embedded the result in the live booking flow. Every new meeting can now be routed to the AE with the strongest evidence of fit. On best-fit assignments, Attention saw a 30% lift in win rate and a 25% reduction in time to close.
A faster way to run ABM
Attention also built an in-house ABM platform to understand awareness, intent, and buying-committee coverage at the person level. Deepline supplies enrichment, technographics, and profile signal into Snowflake, where Attention’s own systems take over. In a broader outbound program, that work contributed to a 3× increase in positive cold-outbound response in roughly three weeks; Attention is clear that the result comes from the full system, not one tool alone.
A question becomes a workflow. Then it keeps running.
- 01
Connect the data that usually stays separate
Bring Salesforce, Attention’s conversation data, and Deepline enrichment into a single analysis surface.
- 02
Let the hypothesis widen
Use Claude Code to investigate the known signals and the unknown ones: buyer background, interests, geography, company profile, and the patterns hidden in deal history.
- 03
Promote what works into a live decision
Re-run the model against historical data, create account scoring, then make it part of routing when a buyer books a meeting.
From a one-time analysis to an operating system for experiments.
Attention now runs live routing and sales experiments with Deepline and Snowflake. The routing model moved from a one-time historical analysis to an always-on booking-workflow step, while the same enrichment layer helps its in-house ABM system understand each buying committee. A question that once required a bespoke table can now be explored in about an hour, at the speed of a prompt.
