Whether you need a manager's nod or a security review, the shape of a successful proposal is the same: a specific outcome, a small cost, a named risk, and a fallback. The instinct to bury the risk is what turns a two-day approval into a two-month one.
Lead with one outcome
Not "improve customer communication". Something countable: reduce no-shows on Tuesday clinics, recover unscheduled treatment, respond to web enquiries within five minutes. A proposal with one measurable outcome is a pilot. A proposal with four is a project, and projects need budget cycles.
State the cost honestly, including the hidden one
| Line | Include |
|---|---|
| Provider line fee | The real monthly number, per line |
| Setup or onboarding | If any — some providers charge four figures |
| Implementation | Developer days, or a provider's setup fee |
| Staff time on replies | The cost everyone omits, and the largest one |
| Fallback channel | You are keeping SMS or email running |
The fourth row is what makes the proposal credible. Someone will ask who answers the replies; having costed it already is the difference between a considered plan and an enthusiasm. The staffing reality.
The risk paragraph
Write it yourself, early, in plain language. Something close to:
Apple publishes no iMessage API, so every vendor in this category reaches the channel by means Apple has not sanctioned. We accept that risk because the channel is additive rather than load-bearing: customer records stay in our own systems, SMS remains a working fallback, and the failure mode is degraded outreach rather than lost contact. The vendor provides a SOC 2 Type II report covering the sending infrastructure, we have a signed DPA, and we have a documented exit path for our numbers.
That paragraph gets approved far more often than a proposal that omits it, because it demonstrates you understood the risk rather than missed it. The full security-review guide.
Do not let someone else discover the Apple problem
If a reviewer finds out from a search that the category is unofficial, and you did not mention it, they will reasonably question everything else in your assessment. Raising it yourself costs you a paragraph. Having it raised for you costs you the proposal.
The objections you will actually get
| Objection | The answer |
|---|---|
| "Isn't this spam?" | Consent, quiet hours, one message per person per week, and here is the opt-out path |
| "Is it legal?" | TCPA applies exactly as it does to our SMS. Same rules, same process |
| "What if Apple shuts it down?" | We lose a channel, not a customer relationship. SMS keeps running |
| "Why not just use SMS?" | Show the pilot number. If there is no pilot yet, ask for the pilot |
| "Who answers it?" | Named person, named hours. Costed in the proposal |
| "What does it cost?" | Per line, per month, plus the reply time. Cancellable monthly |
Ask for a pilot, not a programme
One play, one line, one month, one number to beat. A pilot that size is usually below the threshold that needs real approval, and a pilot with a result is a much easier second conversation than a proposal with a projection. How to define the number you are beating.