**Fictional worked example.** The company, records, counts, timings, and costs below are invented for this illustration. This is a completed example of the report format, not a client engagement or evidence of results. All estimates need checking in a real assessment.

## The decision

Start by standardizing the quote handoff and assigning a follow-up owner in the existing CRM. Test that change before building an AI assistant. In this fictional case, missing information and unclear ownership deserve attention first.

## Business and scope

The example company is a fictional US industrial parts distributor with 45 employees. A sales coordinator receives quote requests in a shared inbox. Account managers check requirements and pricing. An operations lead confirms unusual delivery commitments.

This assessment covers receipt of a quote request through the first follow-up after the quote is sent. It excludes prospect research, pricing strategy, inventory planning, order fulfillment, and implementation. The question is whether the team can move complete requests through the existing inbox and CRM with less chasing.

## Evidence and confidence

These are simulated records, not observations of a real company.

- **Fictional review set:** 20 recent quote requests, an example CRM export, and a walkthrough with the sales coordinator. This set is too small to represent every product or season.
- **Fictional record counts:** 6 of 20 requests lacked a part specification or quantity; 8 waited for a pricing or delivery decision; 9 had no next-step date in the CRM. These groups overlap and must not be added together.
- **Assumed monthly volume:** 120 requests, supplied by the fictional manager. Not checked against a complete inbox export.
- **Estimated handling time:** 18 active minutes per request, based on the fictional walkthrough. No stopwatch study was performed, even within this scenario.
- **Unknown:** CRM plan capabilities, API access, data retention rules, seasonal demand, and the actual causes of lost quotes. No revenue impact is inferred.

## How the work runs today

**1. Receive and read.** The coordinator reads the email and records the customer and request in the CRM. Estimated active time: 4 minutes. Exception: an attachment is unreadable or a part number is unclear.

**2. Clarify requirements.** The coordinator asks for missing information and checks the account history. Estimated active time averaged across requests: 6 minutes. Waiting for the customer is elapsed time, not staff handling time.

**3. Prepare and approve.** The account manager prepares the quote. The operations lead reviews delivery exceptions. Estimated active time averaged across requests: 5 minutes. Pricing and delivery commitments remain human decisions.

**4. Send and follow up.** The coordinator sends the approved quote. The account manager owns the next step, but the fictional CRM records often omit its date. Estimated active time: 3 minutes. A customer reply should stop any scheduled reminder.

## Starting baseline

**Estimated active work:** 4 + 6 + 5 + 3 = 18 minutes per request. At the assumed 120 requests per month, that is 120 × 18 ÷ 60 = **36 staff hours per month**. This is an illustrative workload estimate, not hours available to save.

**Fictional record measures:** missing information in 6/20 requests; no recorded next-step date in 9/20. The 8/20 approval waits identify a question to investigate, not evidence that approvals are unnecessary.

**Separate elapsed-time measure:** record the time from complete request to approved quote. We cannot calculate it from the fictional evidence supplied. Do not substitute active handling time for customer waiting time.

## Options considered

**A. Clearer handoff using current tools. Recommended first.** Agree on required quote fields, the person who owns the next action, and when an exception needs approval. Use a shared checklist and the existing CRM task feature if the current plan supports it. Estimated internal setup effort: 4 to 8 staff hours. Assumed extra license cost: $0, subject to checking the plan. No Syndesi implementation price is implied.

**B. Fixed-rule reminders. Investigate after A.** Create a dated task when an approved quote is sent and close or update it when a customer replies. Estimated setup effort: 6 to 12 hours, plus review and maintenance. Subscription and integration costs are unknown. Do not activate until reply detection and duplicate prevention work reliably.

**C. AI-assisted request drafts. Defer.** Extract requested parts and quantities into a draft for staff review. Estimated pilot setup effort: 16 to 32 hours, plus staff checking time. Model, integration, and hosting costs require a separate estimate. Access, attachment quality, and correction rates are untested. There is not enough evidence to choose this now.

All effort ranges are illustrative estimates, not quotes. Add training, review, exception handling, running costs, and maintenance before comparing options.

## Human decisions and exceptions

- The account manager approves pricing and the final quote. The operations lead approves unusual delivery commitments.
- Missing or conflicting part numbers go back to a person. Software must not guess a substitute.
- A customer reply pauses any automated reminder. If reply detection fails, return to manual follow-up.
- Use sanitized examples during an initial test. Confirm access, storage, and permitted data use before connecting customer records.
- The sales coordinator records failures. The sales manager can stop the test and return to the existing process.

## A small test and a decision rule

**Before changing the workflow:** the coordinator records the next 20 consecutive requests over at least two working weeks. Capture request completeness, active handling minutes, approval wait, corrections, and whether each open quote has an owner and next-step date. Label estimates. Include difficult requests.

**Then test option A:** use the same measures for the next 20 consecutive requests over at least two working weeks. Extend the period if volume is lower. Compare similar request types and record changes in product mix or staffing.

**Proposed operating targets, not forecasts:** every open quote has an owner and dated next step; no unapproved price or delivery commitment leaves the business; median active handling time falls without more corrections or missed replies. The team must agree these criteria before the test.

**Decision:** keep the change if staff use it, quality holds, and measured effort including review and maintenance justifies the setup cost. Revise it if chasing moves to another person. Stop if approvals are bypassed or customer replies are missed. Two small groups do not establish statistical certainty or prove the change caused an improvement.

## Walkthrough and next steps

The report walkthrough would resolve three questions: does the workflow map reflect the team's work, which assumptions can be checked, and who owns the first test? For this fictional case, the next actions are to verify CRM capabilities, collect the baseline, and have the sales manager approve option A's test criteria.

No implementation has taken place. No time savings, sales results, or client outcomes are claimed. The report can be used by the client's own team. Any Syndesi implementation would need a separate scope, price, and schedule.