Independent product / Sales intelligenceLugue.
Finding the gold
in the CRM.
Which of the thousands of leads in our CRM needed us today?
I started Lugue with a question from my own sales work. It evolved into a tool for researching prospects and finding a relevant way to start the conversation.
- Role
- Founder & Product Builder
- Started
- February 2026
- Stage
- Part-time · In development
Screens captured from the current app code in demonstration mode. Accounts, signals and messages are sample data.
01 / From the problem to the product
Knowing who to contact
was only the start.
In my work at MOIC, prioritizing a lead still left another question: what would make the first message relevant to that person?
Lugue’s focus became more specific: research a cold prospect, find a credible angle and help a BD write the opener. Account research and prioritization support that conversation.
Decisions that shaped the scope
Start with a CSV.Bring leads from the existing CRM. Native sync was deferred, with demand from the first user cohort intended to guide that decision.
Build around the first message.Research, enrichment and the dossier earn their place by helping the BD understand the person and choose an approach.
Keep sending with the person.Lugue prepares a draft. The BD reviews it, decides when to send and handles the relationship.
02 / Product experience
Put the reasoning
beside the draft.
The dossier brings the research signal, account fit and writing guidance into the same view as the message. Draft, Follow-up, Reply and History organize the work around a conversation.
A reply changes the next move.
When the user supplies a reply, Lugue examines its tone, the request being made, the useful hook and what to watch. That reading supports the next response.
03 / The difficult part
Would I actually
send this?
A fluent message can still have the wrong angle, sound unlike the sender or make a claim the research cannot support.
I needed a way to evaluate the draft as a piece of sales work. Three questions became useful criteria:
Can I support the claim?
A reference to a post, event or company update needs a basis in the research.
Why would this person care?
The angle should connect the prospect’s situation to a reason for talking.
Does this sound like the sender?
The BD’s writing samples guide the draft. The person still needs to judge the wording.
04 / Testing & validation
Testing the interface
wasn’t enough.
Interface testing could tell me whether someone could move through the app. Message quality needed its own evaluation, and usefulness in a BD’s day needed external users.
Can someone complete the workflow?
An internal test drive covered onboarding, the pipeline, dossiers and drafting. Mobile layout issues and reply/draft crashes were found and fixes recorded.
Internal QA completed · July 2026Did the change improve the message?
The project includes a fixed bench of 40 leads for comparing prompt and model changes. Its checks cover evidence, voice and draft quality; fabricated claims count as failures.
Evaluation harness built · live run pending in the reviewed notesDoes a BD choose to keep using it?
A small external cohort is planned to test the product in real prospecting. Those outcomes are still to be documented.
External validation planned05 / My contribution
My idea.
Built with Claude.
I lead product discovery, UX, AI-assisted development and quality evaluation. Claude helps me implement; I make the product decisions and review the result.
The next step is to test the prototype with external BDs and learn whether it helps them prepare a first message they would use. The public website has a waitlist.
Explore Lugue ↗