Mapping a Conveyancing Workflow Before Buying Software
Harborline Legal is a 22-person firm in a regional city. Residential conveyancing generates 60% of revenue and runs through a 38-step process documented only in a two-page memo written in 2019. Files pass between reception, two paralegals, a fee earner and the accounts team, creating five handoffs. Average cycle time from instruction to completion is 47 working days against a client-facing promise of 30. Last quarter, 18% of files were reworked because the initial client pack was incomplete, and 31 clients complained that nobody told them what was happening. Both paralegals report spending about six hours a week chasing information by phone and email, and overtime is up 22% year on year. The managing partner is convinced the answer is a new case management module and wants to buy one this quarter. She has asked you to run a two-week process review first, using existing staff, with no budget for external consultants and no authority to change headcount.
1. How would you scope and build a current-state process map that the managing partner will actually trust? Set out your steps, the data you would collect and how you would validate the result.
Read the worked answer
Scope the boundaries first. Define start (instruction received) and end (completion and registration, plus file closure and billing). Without a stated boundary the map sprawls and nobody owns it.
Map by role, not by department. Use swimlanes for reception, paralegals, fee earner, accounts, and the client and lender as external lanes. The handoffs between lanes are where the 47 days are hiding.
Collect two kinds of evidence:
- Quantitative: pull timestamp data from the practice management system for 30-40 files closed in the last 90 days. For each step record arrival time, start time, completion time, and who touched it. This lets you separate touch time (actual work, likely 6-9 hours per file) from wait time (likely the bulk of the 47 days). Count rework events, chase emails and calls per file, and complaints per 100 files.
- Qualitative: interview one person per role and walk three live files end to end. Ask where they wait, what they re-key, and what they check twice.
Facilitate, do not dictate. Run a half-day session with the people who do the work, build the map on a wall, then mark each step with frequency, average duration, owner and percentage of files where the step is incomplete first time. Version the map and record assumptions.
Validate before publishing. Walk the draft map with the two paralegals and a fee earner and ask them to correct it. Then test it against five closed files: if the map cannot explain a real file's timeline, it is wrong. Confirm the documented process differs from the actual process and say so explicitly in the report, because that gap is itself a finding.
2. Which steps are likely to be non-value-add, how would you classify and size the waste, and what future-state changes would you sequence? Assume you cannot buy software yet.
Read the worked answer
Classify every step into three buckets.
- Client value-adding: advice, drafting, negotiation, and communication the client would willingly pay for.
- Business-value-enabling: identity and anti-money-laundering checks, conflict checks, file opening, quality review. These are necessary even though the client would not pay for them directly. Optimise them, do not remove them.
- Non-value-add: waiting in queues, chasing missing information, re-keying the same data into the case system, the accounts package and a lender portal, duplicate checks, batching work until Friday, and internal handoffs that exist only because of seating arrangements.
Size the waste before proposing fixes. Using the file data, calculate total wait time versus touch time, count chases per file, and multiply rework hours by a loaded hourly cost. If 18% of files rework for two hours each across roughly 400 files a year, that is a quantifiable annual figure the managing partner can compare with the software quote.
Sequence changes by effort and evidence:
- Quick wins (weeks 1-4): a single standard client pack with a checklist, issued at instruction, that closes the incomplete-information cause. One named contact per file. A status-update trigger at four fixed milestones, so clients hear from the firm before they complain. Daily rather than weekly handoff of files between reception and the paralegals.
- Standard work (weeks 4-10): a one-page standard work instruction per role, a visual file board showing where each file is waiting, and a rule for what counts as a complete file at each stage.
- Then reassess software, now knowing which steps you want the system to support.
Key argument to make: automating a process before simplifying it locks the waste in and makes it harder to see. The rework and chasing costs will not disappear because a module was purchased.
3. What measures would you put in place to prove the changes worked, and how would you stop the team optimising one metric at the expense of another?
Read the worked answer
Measure the flow, not just the average. Use median and 90th-percentile cycle time, because the average hides the files that go badly wrong. Add touch time per file, number of handoffs, first-time-right percentage at file opening, chases per file, complaints per 100 files, and overtime hours.
Define each metric operationally. State exactly which clock starts and stops, which system is the source, and who owns the number. A cycle-time metric that starts at the first fee earner entry rather than at instruction will look excellent while the client still waits three weeks for contact.
Guard against perverse incentives:
- If you only measure cycle time, staff will push work downstream or skip checks. Pair it with first-time-right and a quality review of a sample of files.
- If you only measure touch time, staff will record less time rather than work faster. Pair it with realisation and write-off data.
- If you only measure client complaints, the team will suppress complaints. Pair it with a short post-completion client survey and a sample callback.
Establish a baseline first. Measure four weeks before any change, publish the baseline, and re-measure at 4 and 12 weeks. Use a small run chart rather than a single before-and-after comparison, so you can see whether a shift is real or noise.
Report at the right altitude. The managing partner needs six numbers on one page and a short narrative of what moved and why. The operational team needs a weekly file board and a rework log. Holding a monthly review with both groups keeps the story consistent and lets you test whether the improvement is being felt by clients as well as by the spreadsheet.