Wilshire Law Firm is losing an estimated $7,298 per month because cases are being dropped or accepted from the wrong-fit clients after intake. This gap surfaced three separate times across both pre- and post-purchase complaints, making it the firm's most consistent operational leak. The pattern points to a breakdown in how prospects are screened and matched before a case moves forward.
| Gate | Pass | Detail |
|---|---|---|
| A_volume | PASS | 402 reviews (need 40-5000) |
| B_rating | FAIL | 4.8 stars outside the leak zone 3.5-4.5; abs-negatives branch: 10 1-2-star reviews in the harvested set, need >= 30.0 (flat floor since review_count <= 600) |
| C_recurrence | PASS | top issue mentioned 5x (need 3); 402 reviews >= 300, so also needs >= 2% of the 10-review negative pool (>= 0.2) |
| D_niche | PASS | law-personal-injury: can_pay=True reachable=True |
owner replies to 15% of negatives (signal only as of 1.1, does not gate qualification; <= 30% = reputation-management angle, else pure ops-automation angle). Low reply rate. Pair the automation below with a review-response workflow (reputation-management angle) - these complaints are also sitting unanswered in public, which compounds the rating pressure noted above.
1. Case dropped / wrong fit after intake (intake-qualification, pre-purchase, lost deal): mentioned 3x, severity 5/5, automatability 0.60, est. leak $1,990/mo
Proposed automation: pre-qualification / intake-screening workflow
> "they will not further pursue the case"
> "A month later I got a text ... we've decided not to take your case"
> "they decided to shift me over to Camden law"
2. No status updates / had to chase for updates (comms-responsiveness, post-purchase, repeat/referral loss): mentioned 5x, severity 5/5, automatability 0.85, est. leak $1,161/mo
Proposed automation: proactive status-update automation + AI callback
> "no updates, bad service"
> "after trying to contact someone for an update saying sorry we've decided not to take your case"
> "No communication, hidden fees"
3. Surprise fees / unclear pricing at settlement (pricing-transparency, post-purchase, repeat/referral loss): mentioned 3x, severity 5/5, automatability 0.75, est. leak $697/mo
Proposed automation: automated upfront estimate + written confirmation
> "I was charged extra for this"
> "argue with this firm about their outrageous tax fees they tried to include at the very end"
> "al final inflan las facturas medicas"
4. Slow callback / long wait to reach someone (phone-access, post-purchase, repeat/referral loss): mentioned 1x, severity 4/5, automatability 0.90, est. leak $232/mo
Proposed automation: AI receptionist / missed-call text-back
> "It takes almost 24 hours to get a call back"
These are real recurring complaints too. Automation will not close them, so we are not pitching one; listing them here instead of pretending otherwise.
1. Untrained, rude staff who hung up (staff-attitude): mentioned 1x, severity 4/5, est. leak $663/mo (not automation-addressable)
> "she ultimately hung up on me"
> "the person who calls you back clearly has had no training"
Based on 20 unique complaint reviews. Assumptions (tune per client): each
public complaint stands in for ~8 silent unhappy customers; measured over ~63.3 months of reviews; a
lost law-personal-injury customer is worth $15,000 at
35% margin; a paying-but-burned customer (post-purchase) is
conservatively valued at 35% of a lost deal (lost repeat + referral, not
the historical sale, which is already banked). The per-issue figures above can overlap (one
review often names two issues) so they do NOT sum to this total; the total is derived from the
unique complaint pool, split pre/post by each cluster's classified purchase stage. These are
estimates, shown so the owner can challenge them.
Independent research (Luca, Harvard Business School working paper 12-016) found that a 1-star increase in a business's average rating causes a 5-9% increase in revenue for independent businesses. Every unresolved complaint below is pressure on that same rating, in the other direction.