The intake-to-matter handoff is where case provenance disappears
Direct answer. Two systems own different halves of the same client. The intake CRM holds the enquiry, the qualification, the disposition and the consent artefact; the case management system holds the matter, the costs, the lien file and the disbursement. Where the transfer between them is a person retyping fields, the values with no immediate operational use — source, campaign, consent timestamp, decline reason — have no reason to survive the retype. Provenance is lost at the moment the record becomes valuable. Matters referred out to co-counsel never reach the second system at all, so they leave no object to age or reconcile.
Two systems, one client, no shared identifier
An intake CRM and a case management system are built for different jobs. The intake CRM is an acquisition object: it holds an enquiry, a qualification state, a disposition, and whatever the capture form or the call-tracking layer wrote at the moment of contact. The case management system is a delivery object: it holds a matter, its deadlines, its documents, its advanced costs, its lien file and its resolution. Where the two come from different vendors, and where the connector between them is a person with one window open beside another, there is no shared identifier and therefore no join.
That is a structural statement rather than a comment about staff. A re-key preserves the values a human needs in order to do the next job. Client name, date of loss, insurer, adjuster contact and injury description all have immediate operational use, so they cross. Click identifier, campaign, consent timestamp and decline reason have no use in the next job, so nothing forces them across. The boundary is the one sitting between qualification and matter opening on the personal injury law hub.
What is actually at risk at the boundary
| Value held in the intake CRM | Why it exists there | What is lost if it does not cross |
|---|---|---|
| Click or lead identifier | Written by the capture form or the call-tracking layer | The matter cannot be joined to the acquisition that produced it, in either direction |
| Campaign and channel | Set at capture from the ad platform or a referral picker | Spend cannot be compared against realised fee by source |
| Consent artefact and timestamp | Written when the enquirer submitted the form or answered the intake script | Later follow-up has no stored record of what was agreed, so a contact decision has to be taken without one |
| Disposition and decline reason | Set during qualification and conflict checking | Declines collapse into one bucket, and conflict cannot be separated from lack of merit |
| Referral counterparty | Set when a matter is sent to co-counsel | There is no object to age, chase or reconcile against an incoming fee |
| Original enquiry timestamp | Set when the lead object was created | Time to qualification cannot be measured, because the matter open date is not the enquiry date |
The last row is worth pausing on. Where the matter record's open date is treated as the start of the relationship, the interval between the enquiry arriving and the qualification completing is not stored anywhere. The point is not that the interval is long or short. The point is that it does not exist as a value, so no operational decision can be taken about it and no change to it can be observed.
Dispositions, conflict checks and the referral line
Conflict checking sits between the enquiry and the matter, and it produces one of the more valuable pieces of data in the whole path: a structured reason the firm did not proceed. Where the result is recorded as free text in a note field, the intake CRM can report that an enquiry closed without signing but cannot report why. Conflict, wrong venue, wrong practice area, no available coverage and no merit are operationally different outcomes with different downstream actions — one of them routes to co-counsel and carries a fee expectation, and the others do not.
A matter the firm cannot take does not stop being economically relevant. Fee division with co-counsel is governed by professional-conduct rules that are adopted and enforced state by state, so the arrangement itself carries requirements this page does not interpret. The systems consequence is independent of that jurisdictional detail: the executed agreement exists as a document in a folder, and the expected fee exists as an expectation held by a person. Neither is a record with a status, an owner and an aging clock, and the case management system is not involved at all, because no matter was opened. The reconciliation that would make this visible is narrow: list intake CRM records closed with a referred-out disposition, compare them against executed fee-division agreements in the document system, and compare both against fee receipts posted to the operating account. Where a referred-out disposition value does not exist in the schema, that reconciliation cannot even be started.
Consent is a field, not a policy
The consent artefact deserves separate treatment, because it is the one value at this boundary that constrains what the firm may do later rather than only what it can report. Federal telemarketing rules at 47 CFR 64.1200 require prior express written consent for autodialed or prerecorded telemarketing calls and texts, and require that a revocation — made by using any reasonable method to clearly express a desire not to receive further calls or text messages — be honoured within a reasonable time not to exceed ten business days [SRC-14]. Separately, for a business that meets the applicability thresholds the statute itself sets — a per-entity legal determination this page does not make — the California Privacy Protection Agency states that businesses must honour user-enabled opt-out preference signals such as Global Privacy Control, carry a conspicuous Do Not Sell or Share My Personal Information or Your Privacy Choices link in the footer or header, and comply as soon as feasibly possible, up to a maximum of 15 business days [SRC-15].
Read structurally, both regimes say the same thing about the data model: permission and its withdrawal are per-record states carrying timestamps, and they have to live where the system that sends the follow-up can read them. A consent value stranded in the intake CRM after the client's record moved to case management is a state the follow-up layer cannot see. A separate dependency sits alongside the record rather than inside it: personal injury lawyer is an enumerated Local Services Ads category in Google's category list for the United States [SRC-08], and Google's screening documentation enumerates per-category screening, including licence checks for professional services categories [SRC-09]. Channel eligibility therefore rests on the licensing documentation the firm maintains, not only on bid configuration.
Questions you can answer from your own systems
- Open a matter at random in your case management system: can you name the campaign, directory or referral source that produced it without asking a person?
- Can your intake CRM list enquiries that closed without signing, grouped by a structured decline reason rather than by free text in a note field?
- Does a referred-out enquiry exist as a record with a counterparty, an expected fee and a status, or only as an email thread?
- Is the consent value captured at intake readable by the system that would send a follow-up message, or does it stop at the intake CRM?
- Is the interval between enquiry arrival and qualification stored as a value anywhere, or is it inferred from the matter open date?
What fixing the boundary does and does not do
Mapping the fields across the boundary makes a set of questions answerable that were previously unanswerable. It does not make the answers favourable, and it does not establish that any particular campaign, referral relationship or intake script produces value. It changes the class of the problem from unmeasurable to measured, which is a precondition for a decision rather than a decision. The work is split across capability families — integration and orchestration on one side of the boundary, CRM and lifecycle on the other; the capability architecture sets out how those are separated. If you want an outside read on which parts of this surface are visible before committing to an internal audit, a Revenue Scan documents what public signals show and names where system access would be required to verify the rest.
Source and evidence notes
-
SRC-14Cornell Legal Information Institute — 47 CFR 64.1200, FCC delivery restrictions Limitation: Primary regulatory text on consent, revocation and timing for autodialed and prerecorded telemarketing. It is reproduced here as a systems constraint, not as legal advice, and it does not address state-level or professional-conduct requirements. -
SRC-15California Privacy Protection Agency — Frequently Asked Questions Limitation: States consumer opt-out rights, the opt-out preference signal obligation, the required link labels and the response window. It does not determine whether a given business falls in scope, which is a legal question outside this page. -
SRC-08Google Local Services Ads Help — Getting started with Local Services Ads, United States Limitation: Confirms the enumerated category list including lawyer categories. It is a signal of which categories Google operates a verified-lead product in; it is not evidence of demand volume, cost or performance. -
SRC-09Google Local Services Ads Help — Business screening and verification requirements, United States Limitation: Confirms that screening is enumerated per category, including licence checks for professional services categories. It describes Google's own eligibility process only, and does not state any professional-conduct or bar advertising requirement.