Mass Tort Firms Are Built to Sign Cases. They Are Not Built to Keep Them.

A mass tort partner forwarded me her firm's pipeline last quarter. 14,200 active plaintiffs across four bellwether dockets. Roughly $8 million spent on lead acquisition over the prior twelve months. A 22-person intake team running 24/7 in two languages. Sign-up conversion at 41%, which she was proud of.

Then she shared the number that ended the conversation.

Between the month a plaintiff signed and the month the case actually resolved or moved to a Master Settlement Agreement, the firm lost an average of 19% of plaintiffs. Withdrawals. Non-responses to required questionnaires. Failure to return signed authorizations. Loss of contact when phone numbers changed and nobody updated the file.

19% of 14,200 plaintiffs is roughly 2,700 cases. At an average per-plaintiff net fee of $11,000 in that specific MDL, the firm was watching $29 million walk out the door over the life of the docket. Not from bad facts. Not from a bad trial. From a workflow nobody owns.

Mass tort firms are built to sign cases. They are not built to keep them. That structural gap is the single biggest opportunity in mass tort operations right now, and AI Fusion is the only thing that closes it cleanly.

The silent phase nobody automates

Walk a mass tort case through its lifecycle and you find three very different operational phases.

Phase one is acquisition and sign-up. Ad spend, lead vendors, intake call center, retainer execution, initial PFS or fact sheet completion. Most firms invest 80% of their operational budget here. The playbook is well-developed and well-staffed.

Phase three is settlement or resolution. Disbursement, lien resolution, client check delivery. Most firms have a team and a process, even if it is slow (we covered this in detail in the post on settlement disbursement gaps).

Phase two is the silent phase. Between sign-up and resolution. For a typical MDL that is 14 to 36 months of discovery, bellwether trials, science day rulings, expert depositions, and waiting. From the firm's perspective there is work happening. From the plaintiff's perspective nothing is happening, because nobody is telling them anything.

This is where the 19% goes. Three reasons, all of them fixable.

The first reason is phone number turnover. Roughly 12 to 15% of US adults change their cell number every year. Over a 24-month silent phase that is a 25 to 30% chance the contact info you have on file is dead. When the firm finally needs the plaintiff to sign a release or respond to a discovery request, the call goes nowhere.

The second reason is questionnaire fatigue. Most MDLs require updated plaintiff fact sheets at multiple intervals. The first one gets returned 75% of the time. The third one gets returned 45% of the time. The plaintiff has not heard from the firm in eight months, the form arrives unannounced, it looks like spam, and it goes in the trash.

The third reason is reasonable abandonment. When a plaintiff hears nothing for nine months, they assume the case died. They sign with a competitor. They tell their cousin not to bother joining. The firm finds out twelve months later when the questionnaire deadline hits.

None of these are sign-up problems. All of them are retention problems. And almost no mass tort firm has a retention team.

Why the standard fixes do not work

Every mass tort firm has tried something to close this gap. The four standard approaches all fail in predictable ways.

Monthly newsletters. The firm sends a generic "MDL update" email or text to the entire plaintiff list. Open rates are 18 to 24% the first month, 8% by month four, and effectively zero by month six. The content does not feel like it is about the plaintiff's case. It is about the docket. Plaintiffs do not care about the docket. They care about their case.

Call center quarterly outreach. The firm hires a vendor to "check in" with plaintiffs every 90 days. The vendor reads a script. The plaintiff cannot get any real answer to their specific question because the script reader has no access to the case file. The plaintiff hangs up annoyed. Phone numbers still go dead between calls.

The "we will call when there is news" model. The most common approach, and the worst. There is rarely news for any individual plaintiff during the silent phase. So the firm calls nobody. So plaintiffs leave.

A dedicated plaintiff liaison hire. Some firms hire a person whose job is to communicate with plaintiffs. At 14,000 plaintiffs and one liaison, the math is roughly 670 plaintiffs per touch per month if the liaison does literally nothing else. The math does not work at mass tort scale. Hiring three liaisons makes it $250,000 a year and still does not work.

The shape of the failure is consistent across all four. They treat retention as a broadcast problem (newsletters, scripted calls) or as a staffing problem (more liaisons). It is neither. It is a per-plaintiff, per-case-event, per-channel coordination problem. That is exactly the problem AI Fusion is built for.

What the AI Fusion build looks like

For mass tort retention we deploy three coordinated agents inside one AI Super Agent. They share state through a unified knowledge layer that sits above your CMS (Litify, Filevine, Smart Advocate, or whatever you run), your phone system, your SMS platform, your email, your e-signature provider, and your fact sheet portal.

The Contact Health Agent

Runs continuously. For every plaintiff in the docket, it watches three signals.

It tracks the last time the plaintiff successfully received and engaged with any outbound message (open, reply, click, picked-up call). It tracks the last time the plaintiff initiated contact in any channel. It tracks the carrier-validation status of the phone number on file using a number-validation API (Twilio Lookup or equivalent).

When any plaintiff hits 75 days of no engagement, the agent triggers a soft re-engagement: a personalized SMS referencing the case, the docket, and a casual check-in line. Open rates on these messages in the firms we have deployed with run 64 to 71%, because the message is specific.

When a plaintiff hits 120 days with no engagement and a soft re-engagement that did not get a response, the agent escalates. It dials the number to verify it still routes to the plaintiff (not just connects, actually verifies it is the plaintiff). It cross-references the address on file with USPS NCOA data. If both fail, it flags the plaintiff for paralegal outreach with a full failure report attached.

This is the contact decay problem solved before it costs you a case.

The Case-Event Update Agent

Runs on triggers from the case management system, the bellwether court docket (via PACER feed), and any internal firm milestones (sample selection, deposition scheduling, expert reports filed, MSA terms posted).

When something happens that genuinely affects the plaintiff's case, the agent drafts a personalized update. Not a generic docket bulletin. A message that says: the bellwether trial in your MDL started this week, the firm is monitoring outcomes, and here is what the next 60 days look like for your specific case. It pulls the plaintiff's name, the relevant facts of their claim (from intake notes in Litify or Smart Advocate), and references the case number.

These messages go out via the plaintiff's preferred channel. Email if they engage with email. SMS if they engage with SMS. Voice (recorded by an attorney once, personalized by AI at delivery) if they are voice-preferred.

The Communications Supervisor reviews every outgoing message for compliance, accuracy, and privilege risk before it sends. Anything ambiguous goes to a paralegal as a draft. We covered the supervisor architecture in detail in a separate post.

The metric we watch: percentage of plaintiffs who received at least one substantive, case-specific update in the prior 90 days. Before AI Fusion that number at most mass tort firms is under 25%. After deployment it runs 92 to 97%.

The Inbound Response Agent

This is the one most firms underrate. It runs the inbound side: every call, text, and email that comes from a plaintiff.

When a plaintiff calls to ask "what is happening with my case," the AI Super Agent answers in seconds. It pulls the case file, the docket status, the last firm action, and any pending requests from the plaintiff. It gives a specific, accurate, plain-English answer. If the plaintiff has an open questionnaire, it reminds them. If the plaintiff has changed phone numbers since the last touch, it updates the file.

The number that matters here is inbound resolution rate without human escalation. Generic chatbots run 12 to 20% on this metric in mass tort because they cannot read the CMS. An AI Super Agent built for mass tort runs 78 to 84%. The 16 to 22% that escalates is the genuinely hard stuff: status update on an expert deposition, a plaintiff wanting to withdraw, an ethics-adjacent question. Those go to a paralegal with full context attached.

The math on retention

Take the firm I opened with. 14,200 plaintiffs, 19% silent-phase attrition, $11,000 average net fee.

Industry data we have collected from firms running AI Fusion suggests retention improves 50 to 65% in the silent phase once proactive case-event updates and inbound resolution are deployed. Conservatively, the 19% attrition rate drops to 9.5%.

That recovers 1,349 cases over the docket. At $11,000 net per case, that is roughly $14.8 million in recovered fee revenue.

The AI Fusion deployment cost for a docket of that size runs in the low six figures per year for the platform plus integration, and one or two paralegals stay on retention escalations instead of running a call-center broadcast. Net ROI is north of 70x over the docket life.

The number is not theoretical. The firms with mature deployments are quietly building structural advantages that show up at MSA time. When a defendant's settlement administrator is calculating per-plaintiff values, the firm with documented, engaged, contactable plaintiffs gets paid first and gets paid more. The firm with 30% non-responsive plaintiffs takes haircuts.

What this is not

A few clarifications, because mass tort partners are trained to push back hard on automation claims and they should be.

This is not a chatbot. We covered the difference in the AI chatbot vs AI Super Agent post. A chatbot cannot pull a plaintiff's docket status from Litify and write a case-specific update. It can only echo a script.

This is not legal advice. The AI Super Agent does not opine on case value, settlement strategy, or legal merits. It surfaces facts from the case file and the docket, it routes anything strategic to a human, and the Communications Supervisor enforces that line on every message.

This is not a CMS replacement. You keep Litify or Smart Advocate or whatever you run. AI Fusion sits above your stack with read access to your case data and write access to client communications. We integrate. We do not migrate.

This is not a one-size-fits-all template. Every MDL has its own questionnaire schedule, its own science day timing, its own bellwether calendar. The agent gets configured per docket. That configuration work is what the deployment month is for.

The implementation playbook for a mass tort docket

For a partner thinking about how this rolls out, here is what the first 60 days look like.

Days 1 to 10. Audit the docket. Pull contact health data on every plaintiff. Most firms discover 12 to 18% of their plaintiff phone numbers are already dead. That alone is a wake-up call.

Days 11 to 25. Configure the Case-Event Update Agent against the specific MDL. PACER feed integration. Identify the bellwether milestones, science day calendar, MSA terms posting dates. Build the per-plaintiff personalization template against your CMS schema.

Days 26 to 40. Run in draft mode. Every outbound message goes to a paralegal for approval before it sends. The AI learns your firm's voice and your specific case facts. Accuracy converges fast because the underlying data is consistent.

Days 41 to 55. Selective auto-send. Categories that hit 99.9% accuracy graduate. Case status updates, questionnaire reminders, contact verification messages. Strategy-adjacent messages stay in draft permanently.

Days 56 to 60. Inbound goes live. Plaintiff calls and texts route through the Super Agent first. Anything ambiguous escalates to a paralegal with the full case context.

By day 90 the docket retention number starts moving. By day 180 the math compounds. By the time the next bellwether hits, the firm has a documented, contactable, engaged plaintiff base while competitors are still calling dead numbers.

The takeaway

The mass tort firms that win the next cycle of dockets are not going to win on lead spend. Lead acquisition is already a commoditized, expensive, well-staffed function. The advantage is gone.

They are going to win on retention. Specifically on the silent phase between sign-up and resolution, where everyone else is losing 19% of their plaintiffs to silence.

If your docket retention number is not on your dashboard, it should be. If you do not know what it is, ask your case manager today. If it is over 12%, you have a workflow problem that is costing you eight figures over the life of the docket. AI Fusion is the only thing that closes it without doubling your headcount.

---

Want the retention math run on your specific docket? Book a 20-minute call. We will pull the contact-health audit and project the recovered case value before you sign anything. Book a call or start with Voice AI free.