New to Attio?Get 10% off when you sign up through Craftt.Try Attio free →
← All case studies

69% of the open pipeline was already past its own closing date, and the dashboard said so.

KNR Agency is a Paris e-commerce agency that builds and runs online stores for fashion, cosmetics and luxury brands. They moved to Attio from Zoho and had already done more than most: web forms and bookings creating records through n8n, an AI lead scorer, three dashboards. The audit found a clean database with a pipeline nobody was acting on, a finance tool that could never talk to Attio the way they had tried, and no motion at all for existing clients.

Free audit

Engagement

595

Deals audited

64 of 91

Open deals past closing date

€1.22M

Expired pipeline

71.9% vs 24.8%

Up-sell vs first-deal win rate

0

Duplicate emails in 1,886 people

13

Fixes prioritised

The problem.

On the call, Gabriel Kaam named four worries: reports nobody used, quotes in PennyLane that never reached Attio, no customer-success motion after a deal closed, and a database he believed was full of junk and duplicates.

The last one was wrong, and that mattered. Across 1,886 people there was not one duplicate email address. The data was clean. The problem was the opposite: fields that did not exist. The deal object had no account manager, no lead source, no next step, no quote or invoice number, and no won date.

The pipeline showed what that costs. 64 of 91 open deals, worth €1,220,075, had a closing date in the past. Their own Expired Deals tile showed exactly that number. We rebuilt it deal by deal and it matched to the euro. Nothing in the workspace acted on it, and two deals had expired sixteen months earlier.

The automation had the same pattern. The AI lead scorer produced genuinely good reasoning on 218 people, then stopped in March because its trigger was manual: someone had to right-click each record. None of the 279 people created since April had a score. The Slack alert for new leads had fired 11 times against 292 new people.

Objectives.

  • Reproduce every number on their dashboards from the raw records, so the audit could say which tiles were right and which were not.
  • Answer the four worries from the call with data, including the ones where the answer was no.
  • Find the honest date on the workspace, since the migration had stamped most records with the same creation day.
  • Give the deal object the fields the sales process already used out loud.
  • Design the PennyLane link and the customer-success motion Gabriel asked for, step by step, so the team could build them without us.

What we found.

Thirteen fixes in total, ranked by priority and effort. The findings that shaped them:

Up-sells close at 71.9%, first deals at 24.8%

An existing client was almost three times as likely to buy again as a new logo. Yet 57 of 94 clients had no open deal and no win in over six months, and nobody owned them. There was no customer-success pipeline, no account manager on the deal, and no check-in reminder.

The creation date on 441 deals was one day

The Zoho migration created 441 of 595 deals on 8 December 2025. Any age or cycle-time report built on created date would lie. Closing Date, filled on 589 of 595 deals and reaching back to 2018, was the only honest date, so every analysis in the audit used it.

Six webhooks, no filters

Every webhook in the workspace pointed at n8n, three on record created and three on record updated, and none had a filter. Editing one phone number fired three executions. We recommended cutting six to two filtered ones, and moving anything that stays inside Attio into native workflows.

One dashboard tile could not be reproduced

Expired Deals was right. Pipe Value showed €331,500, which matched neither the total open pipeline (€1,758,825) nor the future-dated part of it. Five yearly charts were hard-coded per year. Nothing reported on source, cycle time, activity or client health.

What we were honest about.

On the call we said the PennyLane problem sounded like an Attio bug. It was not, and the audit said so on the first page.

Why BCC could never work

Attio's forwarding address only accepts mail sent from an address that belongs to a workspace user. PennyLane sends quotes from its own servers, so Attio rejects them. That is documented behaviour, not a bug, and no setting changes it.

The route that does work

PennyLane's API. A right-click on a deal creates the quote and writes its number and link back. PennyLane has no webhook for a signed quote, so n8n polls its change feed every 15 minutes, moves the deal, and turns the signed quote into an invoice. About 96 API calls a day.

The fixes that carry the most weight.

Give the deal a memory

About a dozen fields the process already used verbally: account manager, lead source, next step and its due date, proposal-sent and won dates, payment status, quote and invoice numbers. Backfill only the 91 open deals, which takes under an hour.

A Client Success list on Companies

Customer success happens to an account, not a deal, so it lives on a Companies list with its own stages, a health flag and an up-sell signal. Two workflows cover what Gabriel asked for: a six-month check-in, and an alert when no one has spoken to a client in 14 days, read from Attio's own last-interaction field.

Nine workflows, built in Attio

A stale-deal task at 21 days, a task back to the owner when a deal is lost without a reason, a won-deal cascade that updates the company and starts onboarding, and the lead scorer switched to run on every new person instead of a right-click.

Results.

  • Every dashboard tile checked against the raw records. One confirmed to the euro, one shown not to reconcile.
  • The junk-and-duplicates worry answered with data: zero duplicate emails in 1,886 people. The real gap was missing fields.
  • €1.22M of expired pipeline turned into a one-sitting triage list, oldest and biggest first.
  • A clear case for a customer-success motion, built on the team's own win rates, with the two alerts they asked for specified exactly.
  • A working PennyLane design, after correcting our own first guess from the call.
  • Thirteen fixes with priority, effort and the cost of skipping each, plus a build sheet the team can work from without us.
George and the team at Craftt provided an incredibly thorough audit of our Attio workspace. They identified key bottlenecks in our data model and workflows, delivering a clear, actionable roadmap that gave us total clarity on how to optimize our CRM. Highly recommend them for anyone looking to unlock the full potential of their Attio setup.

Gabriel Kaam

Co-founder & CEO, KNR Agency

Ready when you are.

Book a call and we will tell you honestly whether this is worth doing, or start with the free 48-hour audit and decide afterwards.