MedConnect Mobile went live at the Release 2 cutover on 8 May 2026 and the program closed one week later. It is now 161 days into a benefit case worth $1,130,000 a year. This is the first time any of it has been measured — which is the most useful thing this document has to say about itself.
1. The problem with how this program handled benefits
MedConnect Mobile delivered in eight two-week sprints across two releases. It inspected and adapted on scope, on quality, on team health, and on its own process, every fortnight for five months. It measured its business benefits exactly once — never.
That is not an agile failure in delivery. It is a failure to extend the same feedback discipline one step further out. A program that ships increments but evaluates value in a single event at the end has adopted agile delivery with waterfall benefits management, and gets the worst property of each: the frequent releases produce no learning about whether the product is working, and the single evaluation arrives too late to change anything.
The corrective is not a bigger report. It is that benefit measures are treated the way the Definition of Done treats quality: a property of each increment, checked at the release boundary, not a phase at the end.
2. What 161 days of live data actually shows
Values are the CBA's, unchanged. Realization is computed from the operating measures the CBA itself names as its basis — not estimated.
| ID | Benefit | Annual | Readable? | Realized | Run-rate evidenced | Accountable owner | Confidence |
|---|---|---|---|---|---|---|---|
| BEN-01 | Legacy Vendor License Elimination CareLink Classic (Vantix) annual license retired at cutover |
$340,000 | Yes | 100% | $340,000 | M. Delacroix (Sponsor, VP Digital Health) Measured by C. Tyrrell |
High |
| BEN-02 | Provider Throughput / Visit Capacity Gains Faster visit setup (~6 min → under 2 min) frees provider capacity |
$250,000 | Yes | 85% | $212,500 | Dr. L. Nguyen (CMIO) Measured by F. Delgado |
Medium |
| BEN-03 | Reduced Patient No-Show Rate Mobile reminders reduce missed appointments |
$220,000 | Partial | 0% | $0 | M. Delacroix (Sponsor, VP Digital Health) Measured by F. Delgado |
Low |
| BEN-04 | Patient Retention / Satisfaction Mobile-first experience reduces churn to competitor telehealth apps |
$180,000 | No | 0% | $0 | M. Delacroix (Sponsor, VP Digital Health) Measured by L. Bianchi |
Low |
| BEN-05 | Reduced IT/Help Desk Support Burden Support tickets: ~410/mo → under 150/mo (see Vision & Roadmap OKRs) |
$140,000 | Yes | 86% | $120,400 | T. Brannigan (Director, Information Security & Compliance) Measured by R. Okafor |
High |
| TOTAL | Five benefit categories | $1,130,000 | — | 60% | $672,900 | Against a Year 1 target of $678,000 (60% realization). Tracking at 99% of target with 35% of the benefit base unmeasured. | |
3. Reading each line
Legacy Vendor License Elimination
$340,000/yr · 100%Contract terminated at Release 2 cutover; procurement confirmed. Full run-rate secured.
Provider Throughput / Visit Capacity Gains
$250,000/yr · 85%Visit-setup median moved 6.0 min to 2.6 min against a sub-2-minute target — 85% of the intended improvement, not yet all of it.
Reduced Patient No-Show Rate
$220,000/yr · 0%Not claimed. No-show rates are strongly seasonal and 161 days does not contain a comparable prior-year period. A reading exists; it is not yet evidence.
Patient Retention / Satisfaction
$180,000/yr · 0%Not claimed. Churn is measured over rolling 12 months; the first defensible read is May 2027.
Reduced IT/Help Desk Support Burden
$140,000/yr · 86%Ticket volume moved 410/month to 186/month against a sub-150 target — 86% of the intended reduction.
4. What each release was supposed to move
This is the table the program should have had at Sprint 0, and the reason section 1 calls the omission expensive rather than merely untidy. Mapping the six epics onto the five benefits takes about an hour and makes every release boundary a measurement opportunity.
| Epic | Released | Benefit it was meant to move | Readable at that release? |
|---|---|---|---|
| 1 — Patient Identity & Onboarding | R1, 27 Mar | Enabler. Moves nothing on its own | No — correctly |
| 2 — Scheduling & Reminders | R1, 27 Mar | BEN-03 no-show rate | Partial — needs a seasonal cycle |
| 3 — Video Visit Experience | R1, 27 Mar | BEN-04 retention / satisfaction | No — 12-month measure |
| 4 — Provider Console | R2, 08 May | BEN-02 visit setup time | Yes, from day one |
| 5 — Data Migration & Decommission | R2, 08 May | BEN-01 Vantix licence elimination | Yes — binary, on the termination date |
| 6 — Messaging & Refills | R2, 08 May | BEN-05 help desk volume | Yes, within weeks |
Three of the six epics carried a benefit that was readable almost immediately, and all three sat in Release 2. The other three carried benefits that genuinely cannot be read quickly — and saying so up front is as valuable as measuring the fast ones, because it stops a steering committee expecting a retention number in June that will not exist until May.
5. Why velocity is not on this page
The program tracked velocity, burndown and a release burnup, and all three were useful for forecasting delivery. None of them appears here, because none of them is a benefit. Story points completed measures how much the team produced; every measure in section 2 concerns whether producing it changed anything for a patient, a provider or the help desk.
The Test Strategy already makes the equivalent move by excluding velocity from quality metrics. The same logic applies with more force to value: a team can raise velocity for a full quarter while delivering nothing anyone uses, and a benefit register that accepts output as evidence will report that as success.
6. Who owns this now that the program is gone
The squads dissolved at program closeout on 15 May 2026. Unlike a program that hands a system to an operations function, there is no handover event here — MedConnect Mobile is a live product, and the product is the operation. What actually stopped at closeout was not ownership of the product but ownership of measurement: it sat with the delivery squads, and it left with them.
| What is needed | Owner | Cadence |
|---|---|---|
| Benefit measures reported alongside product metrics, in the same review | C. Tyrrell (Product Owner) | Monthly |
| Clinical measures (visit setup, no-show) validated as clinically meaningful, not just statistically moved | Dr. L. Nguyen (CMIO) | Quarterly |
| Help desk volume tracked against the 410/month pre-cutover baseline | T. Brannigan | Monthly |
| Full-year reads on no-show and retention, which need a seasonal cycle | M. Delacroix (Sponsor) | Once, May 2027 |
Decision DEC-06 corrected this program's staffing from 13 people to 28 once it became clear the original estimate could not deliver the scope. The same class of error applies here in miniature: nobody was assigned to measure benefits, so the work did not get smaller — it simply did not happen.
7. Open gaps
| ID | Gap | Consequence | Owner |
|---|---|---|---|
| BG-01 | No benefit measure was read between cutover and this document — 161 days. | Two releases' worth of evidence about whether the value hypothesis held was available and unused. The readings recovered here were reconstructed after the fact, which is why two carry Medium rather than High confidence. | C. Tyrrell (Product Owner) |
| BG-02 | $400,000 of annual benefit (35%) depends on measures that need a full seasonal cycle — no-show rates and churn. | Neither can be honestly read before May 2027. Any Year 1 claim before then covers at most 65% of the base. | M. Delacroix (Sponsor) |
| BG-03 | Assumption A-01 and risk R-01 were closed with the program's registers at closeout, but the benefit case runs for seven years. | The risks that threaten benefits — adoption plateau, competitor response, PulseConnect SDK dependency — now have no register. They are product risks, not program risks, and need a product-level home. | C. Tyrrell (Product Owner) |
| BG-04 | The pre-cutover baselines survived only because they were written into the CBA (410 tickets/month, ~6 minute visit setup). Nothing else was captured. | Fortunate rather than planned. Benefits that were not given a numeric baseline in the business case — retention in particular — have nothing to be compared against. | F. Delgado |
8. What good looks like from here
Three changes, none of which requires a new ceremony.
Benefit measures join the monthly product review. They are four numbers. They belong next to adoption and crash rate, read by the same people at the same time, not in a separate annual exercise.
Every future release states which benefit it is meant to move, before it ships, and the release is reviewed against that claim afterwards. A release that moves nothing is worth knowing about while the next one is still being shaped.
The May 2027 full-year read is scheduled now, with an owner, because it is the only point at which no-show and retention become defensible and it falls a year after everyone involved has moved on.
None of this recovers the 161 days already spent. It is written down because the next program will make the same omission by default, and the correction is cheap only if it is designed in at the start rather than reconstructed at 161 days like this one.