The follow-up you keep meaning to send. Sent.

1. 900 inbound calls/month × 28% unanswered = 252 missed calls. The 28% figure is a published average for inbound call handling in high-volume phone verticals; a well-staffed shop will be lower, a two-person office much higher.
2. 252 × 85% = 214 callers who never call a second time. CloudTalk's research puts non-return after a missed call at 85%, with 80% declining to leave a voicemail.
3. 214 × 25% engagement on text-back = 53 live conversations. We use 25% deliberately. Vendor case studies claim 30–40%; we do not model on vendor case studies.
4. 53 × 18–25% close = 9–13 booked jobs. Phone leads convert to appointments far better than web leads, but these are recovered calls, so we discount the rate.
5. 9–13 jobs × $650 average ticket = $6,240–$8,450/month. Swap in your own average ticket; on a $2,400 install average the same model returns north of $23,000.

What this model excludes: after-hours calls, which typically add another 15–30% of missed volume; the referral value of a customer served instead of lost; and the lifetime value of a maintenance agreement sold on a job that would never have been booked. It also excludes gross margin — this is revenue, not profit.

1. Baseline: $180,000/month total revenue, 8% email-attributed = $14,400. That is a typical no-flows starting point; well-optimized programs are reported around 33% of total revenue, so 8% is a long way from the ceiling.
2. Target: 18% email-attributed = $32,400, a $18,000/month lift. We model to 18%, not to the 30–40% seen in mature programs, because the first 90 days are about building the machine, not tuning it.
3. Cart recovery, modeled separately: 1,150 abandoned carts × $95 AOV = $109,250 abandoned. At a conservative 6.7% recovery — the published range is 5–15% — that is $7,320/month.
4. SMS layered on the two highest-intent flows only. Automated SMS averages roughly $0.75 revenue per message against $0.15 for broadcasts; we send far fewer texts than the platform will let us, because opt-out rate is the metric that kills a program.
5. Ramp: roughly 40% of modeled lift in month 1, 75% by month 2, full run-rate by month 3. Flows need traffic to pass through them before the numbers mean anything.

What this model excludes: list growth from the on-site capture we install, which compounds the whole model; margin, which varies enormously by category; and the platform's own reported attribution, which is generous — we recommend measuring against a holdout group instead.

1. 14,000 dormant contacts × 1.2% reactivation = 168 accounts re-engaged. Reactivation campaigns commonly report 1–3%; we model at the bottom of the band because a 24-month gap is a long one.
2. 168 × 25% place an order = 42 orders. Re-engagement is not the same as revenue. Most of the 168 will need a rep involved.
3. 42 × $1,400 average order = $58,800. On a B2B account with reorder behavior, first-order value understates the real number substantially.
4. Segmented sends throughout. Segmented campaigns are reported to generate materially more revenue than unsegmented ones — this campaign is split by former category, order size and lapse window, and the copy differs in each.
5. Suppression is doing real work here. Anyone who has opted out, hard-bounced or complained never enters the campaign. On a 24-month-old list, list hygiene before the first send is not optional — sending to a stale list is the fastest way to burn a sending domain.

What this model excludes: the recurring order value of a reactivated account, which is the actual prize; and the deliverability risk of doing this badly, which is the reason most distributors never try. B2B email ROI runs lower than ecommerce — roughly $28–$32 per $1 — and we would rather say so up front.