Auditing 67,000 Emails for a $3.5M Answer

ABC Field Services Company Turns Its Inbox Into a Sales System

Client:

A field-service company with multiple trade divisions and 125+ staff. The business runs on an established on-premise ERP with a job-management module, and quoting is high-volume and low-ticket: thousands of service quotes a year, most under $10,000, owned by a small business development team. Leadership suspected quoted work was slipping through the cracks, but had no way to prove where.

Challenges:

Leadership knew revenue was leaking but could not prove where. Quote outcomes lived in the ERP, the conversations that decided them lived in individual inboxes, and no report connected the two. Reps worked from memory, and the weekly sales report was assembled by hand, going stale the moment it was sent.

Because no one could say which quotes were genuinely dead, every quote stayed technically open, which meant none of them got chased. The follow-up list did not exist anywhere except in people's heads.

Other Considerations & Issues:

• Leadership was doing the follow-up. The owner had personally re-pinged the top producer eight times in five months, and the team's coordinator did so 115 times, chasing answers that had already been asked for.
• The two systems disagreed. Of the quotes the email trail suggested were dead, a third were already approved in the CRM, and 67% of quotes sent never received an explicit yes or no by email at all.
• The team was software-shy. Earlier rollouts had stalled, so anything new had to live inside the tools the team already opened every morning, not add another login to ignore.

Solution:

We treated the inbox as a data source, then built the system the findings called for:

• Audited six mailboxes, read-only. A scoped, read-only connection pulled 67,158 messages and classified every thread by who spoke last, how long it had been silent, and what it was worth.
• Cross-referenced every quote against the CRM. Each quote number found in email was matched to its real status, so the findings rested on confirmed outcomes rather than inferred tone.
• Built a read-only data mirror. 145 tables from the on-premise ERP now sync to a cloud database hourly, so reporting never touches, or slows, the system the technicians depend on.
• Shipped a live console and an AI assistant. One secure page with the pipeline ranked by value and age, an hourly rescan of the inbox for silent money and unhappy language, and an assistant that drafts follow-ups but always waits for a human to hit send.

The Reporting They Get Now:

The first deliverable was the audit itself, recreated below with generic labels. The follow-up view put a dollar value on silence: of $5.45 million in open quotes, only $85,234 was on a healthy follow-up rhythm, and $3.5 million, across 324 quotes, had never been followed up at all.

The CRM cross-reference kept the story honest. Sales ability was never the issue: 522 quotes won in five months, a 78.7% close rate by count. The gap was memory, not skill, and that distinction shaped everything we built next.

Impact:

The audit gave leadership a number it could act on, and the system turned that number into a habit. What used to take a rep an afternoon of memory and searching now loads in a browser tab, current to the hour, with the oldest and largest silent quotes surfaced first. On its first day of live monitoring, the console flagged two silent threads worth $306,000 and $137,000, plus two unhappy-client emails that had never been answered.

Where the Open Pipeline Stood at Audit

Follow-up status	Quotes	Value	Share of value
On cadence, under 15 days	28	$175,498	3.2%
Gone quiet, 15 days or more	124	$1,734,620	31.9%
Never followed up	324	$3,535,814	64.9%

What Changed in Day-to-Day Operation

Measure	Before	After	Change
Sales reporting	Weekly, by hand	Hourly, automatic	Always current
Follow-up list	Memory	Ranked by value and age	476 quotes visible
Unhappy-client alerts	Word of mouth	Hourly inbox scan	2 caught on day one
Job records audit	None	Monthly, systematic	1,613 jobs checked

---

Conclusion:

Nothing about this company's ability to sell was broken, and the audit proved it with a 79% close rate. What was broken was memory. Six mailboxes held the answer the whole time; it just took reading all of them at once, and checking every claim against the CRM before believing it.

In its first six weeks of live operation the console has become the Monday-morning starting point: the pipeline ranked by value and age, silence flagged before it costs money, and a job-records audit that now runs monthly instead of never. Every business has this gap somewhere, between the system of record and the conversations that actually decide outcomes. The engagement is ongoing, and the follow-up number is now measured weekly instead of discovered once.