Leasing Follow-Up Cadences

research · current state · recommendations · 2026-10-07 · Proposed, pending Fede

Status: proposal only. Nothing has been changed. Data: production, last 60 days.

Answer Current state Recommendations Research Production Camellia All series Sources

The answer

1. Keep the 4-touch series (day 1, 2, 7, 14) for new leads. Replies come from touches 1 and 2. Touches 3 and 4 add little but cost nothing.
2. One schedule for both series at every client. Delete Camellia's two overrides (pending Fede).
3. Do not add a day-3 or day-30 touch.
24 automatic series in total; 6 sent nothing in 30 days, 2 ignore quiet hours.

Current state

Series	Who	Default	Camellia	Western Slope	Last 60 days
New lead, never answered	New leads	1/2/7/14	default	default	Camellia 64 nudges, 7 replies, 6 tours. Western Slope 49 nudges, 14 replies, 6 tours
Replied, then quiet	Leads who went silent	1/2/7/14 (changed today; was day 3, then day 4)	own override, day 3 / day 6	default	Camellia 4 nudges, Western Slope 7. 0 replies, 0 tours
Tour reminders	Booked tours	Day before and shortly before	no override	no override	n/a
After a tour or no-show	Toured leads	Nothing, by design (Sept decision)			n/a
Renewals	Residents	Separate program

Recommendations

Delete Camellia's two saved overrides so every client runs the default. Yes.
Keep 4 touches for new leads; revisit at 200 leads with nudges. Keep.
Replied-then-quiet: 11 nudges, 0 replies, 0 tours in 60 days. Keep on the standard schedule, re-check in 60 days. Keep, re-check.
Per-client schedule setting: it has only produced one accidental difference. Remove it so there is one path, or keep it and show the value on the customer page. Recommend remove (Fede decides; the settings screen is Gera's lane).
No day-3, no day-30. No.

Research: what published studies and our data say

The answer

Do not add day 3. Do not add day 30. Our replies fall off fast after the second text, and no published study we could find supports a day-3 or day-30 touch for apartment leads. The series we have is already as long as the evidence will carry.

Our numbers by text number

Standard series: +24h, +48h, +7d, +14d. Every lead in the window ran this exact series (no client overrides). Reply = the lead texted or emailed back within 48h of that text.

Text	Sent	Replied in 48h	Reply rate	Led to a tour	What published research says
1st (+24h)	49	12	24%	3	Early touches carry most conversions
2nd (+48h)	31	10	32%	1	Same
3rd (+7d)	17	1	6%	1	Diminishing returns
4th (+14d)	12	0	0% (could be up to ~24%)	0	Past the 6-attempt knee

+24h24% (12 of 49)

+48h32% (10 of 31)

+7d6% (1 of 17)

+14d0% (0 of 12)

52 leads got at least one text. 23 of them (44%) replied to at least one. Out of 109 texts sent, no reply was a STOP request. The samples are small, so read the shape, not the decimals.

The two questions: add day 3? add day 30?

	Add day 3?	Add day 30?
Our data	We have never run it, so no direct number. The slot sits right after the +48h text, which got 32%, then a drop to 6% by +7d.	The +14d text got 0 of 12 replies and +7d got 1 of 17. A 30-day text would sit further out on a line that already points to zero.
Benchmark	No day-2-vs-day-3 spacing study found. Best contact strategy found: about 6 calls plus 5 emails, mixed (secondary source).	Nothing found past about 6 to 7 attempts. Leads that need 7 or more attempts convert 45% less (secondary).
Opt-out and complaint risk	A third text in three days is the first spot where a few states cap contact at 3 per 24 hours. We would still be under it. Risk rises with each added daily text. No published opt-out rate by frequency found.	Low legal risk, but an old lead who forgot they asked is the person most likely to complain or hit STOP.
Recommendation	No Not supported. If Fede wants to know for sure, run it on a share of new leads for 60 days and compare.	No Expected to get zero or near zero replies.

Other changes the evidence points to

Idea	What we know	Verdict
Drop the +14d text	0 replies in 12 sends. But it closes the series and the lead is then marked inactive in the property system.	Optional Keep it as the closer, or send that one by email.
Faster first reply	MIT/InsideSales: waiting 30 minutes instead of 5 cuts the odds of reaching a lead about 100 times. We did not measure our own first-reply time in this pass.	Measure first This is probably the biggest lever.
Text first	Texts already go first. Email response rates quoted by vendors are lower, but the figures are unsourced.	Already doing it
Re-engagement after a reply	17 leads, 11 texts sent (day 3 then day 4 or 6 after the last reply), 0 replies. Too small to judge, and none of them reached a tour.	Watch After tonight's change, check again in 30 days.

When our texts land (small samples)

Most texts go at 8am local, because the quiet-hours hold pushes them to the first allowed slot: 32 of 109 sends, 22% replied. By weekday, Wednesday did best (7 of 16) and Thursday and Sunday worst (1 of 12 each). With samples this small, none of this is a finding. Times are Mountain, approximate.

How we counted

Source: follow-up records and conversation messages in production, two clients, leads whose first reply was sent Aug 7 or later (145 leads; 52 had at least one text actually sent).
Reply = any inbound message after the text, before the next text, up to 48h. Some replies may be to Clara's earlier message, so reply rates are an upper bound.
Tour = a confirmed tour booking within 3 days of that text, from the conversation log. Approximate.
Only one stage-one series appears in the records. Earlier schedule versions exist in the code history (the 5-step series was added in June, the editable follow-up settings in July) but no lead in this window ran a different one.

Risk rules to stay inside

Federal: no marketing-style calls or texts before 8am or after 9pm in the recipient's time zone. 47 CFR 64.1200 (primary). Whether it covers a reply to an inquiry is unverified.
A STOP, or any reasonable way of saying it, must be honored within 10 business days (rule in effect since April 2025). Same source. A broader "stop applies to everything" rule is delayed to Jan 31, 2027 (Hunton, secondary).
Oklahoma and Maryland: 8pm cutoff and a cap of 3 contacts per 24 hours on the same subject (Kelley Drye, secondary). Our series never comes close.
CTIA: a business replying to a consumer who asked to be contacted is conversational or informational; adding promotional calls to action may change that. It says nothing about how many follow-ups are allowed. CTIA principles (primary).
Carriers filter on opt-out and complaint rates, but Twilio publishes no number (Twilio). A "stay under 1% opt-out" rule of thumb exists only on reseller pages (anecdotal).

Outside evidence and how much to trust it

Claim	Number	Source	Trust
Reach rate, 5 vs 30 minute response	About 100x odds of contact, 21x of qualifying	MIT/InsideSales study	Primary, B2B leads, not apartments
Firms responding to a web lead	37% within 1 hour, 23% never	HBR	Primary, figures via summary (paywalled)
Converted leads reached by attempt 6	93%	Velocify study via Cirrus Insight	Secondary
Best cadence: 6 calls plus 5 emails	+128% conversions	Same	Secondary
Renters expect a reply within a day	72%	Zillow 2017	Unverified, from a search snippet
SMS vs email response	45% vs 6%	Repeated by vendors, attributed to Gartner	Unverified, treat as anecdotal
AI leasing: more tours	+30% lead-to-tour	Zuma	Vendor marketing

Not found: published multifamily follow-up-count or tour-conversion data from Knock, Funnel, RentCafe, Entrata, RealPage, NMHC or NAA; opt-out rates by texts per week; any day-2 vs day-3 spacing study.

Production, Aug 8 – Oct 7

Aggregates only, no names. Situs Group: 149 leads, 0 series (not live yet).

Who entered a series and how many touches went out

Series	Entered	Got 1 or more	Touch 1	Touch 2	Touch 3	Touch 4	Median day of each
Camellia new	66	22	22	19	12	11	1 / 2 / 7 / 14
Camellia replied-then-quiet	36	2	2	2	0	0	3.2 / 6.2
Western Slope new	85	31	28	15	5	1	1 / 2 / 7 / 14
Western Slope replied-then-quiet	48	5	5	2	0	0	3.0 / 4.1

Replies within 72 hours of a touch

Series	Touch 1	Touch 2	Touch 3	Touch 4	Tours after a touch
Camellia new	3 of 22	4 of 16	0 of 11	0 of 10	6
Western Slope new	8 of 24	5 of 9	1 of 5	n/a	6
Replied-then-quiet, both clients	0 of 6				0

No application followed any touch.

How each series ended

Outcome	Camellia new	Camellia re-engage	Western Slope new	Western Slope re-engage
Replied before any touch	39	5	34	6
Replied after a touch	7	0	16	0
Tour booked or applied	2	15	5	11
Lead's last message said not to nudge	–	10	–	12
Closed, home gone, or merged	0	1	2	3
Send failed or no contact	1	0	4	0
Ran the full series	11	2	9	2
Still running or no outcome	6	3	15	14

Findings

Most leads never get a nudge because they answer before day 1.
Replies come from touches 1 and 2.
Camellia's override affected exactly 2 leads. Neither replied.
No nudge went to a vendor or a resident in the window.
3 replied-then-quiet series look stuck past day 17 (being checked).

How Camellia ended up on 3 and 6

Saved 2026-07-23 at 23:21 UTC by Gera in the Follow-ups settings card, one day after the card shipped.
No note. No decision in the 07-20 Camellia review (which set the then-default "3 days, then +1 day", Sean's words), the standups, or Slack.
A second Camellia override, saved 07-22, equals the default (no effect).
Western Slope, Situs and Yale have none.
Last night's default change to 1/2/7/14 left Camellia stuck on two touches.

A side effect of testing, not a decision.

Every automatic series Clara runs

24 series, from code on main and the production database (read-only), 2026-09-07 to 2026-10-07. n/m = not measured.

Series	Receives	Default schedule	Channel	Override in prod	Live at	Sends, 30 days (Camellia / Western Slope / Situs / test)
New lead never replied	prospect	+24h, +48h, +7d, +14d; ≥20h apart; 8am–9pm; 48h grace	text, email fallback	Camellia row identical to default (no-op)	Camellia, WS	33 / 49 / 0 / 83
Replied, then went quiet	prospect	same 4 steps from last reply; no 20h gap, no grace	text, email fallback	Camellia: 2 steps, +3d, +6d	Camellia, WS	2 / 7 / 0 / 1
Tour offered, not confirmed	prospect	+24h, +3d; dropped at 7-day hold	text	none	Camellia, WS	0
Tour reminders	prospect	24h and 1h before; NO quiet hours	text	none	Camellia, WS	22 / 29 / 0 / 16
Weekend tour check-in	prospect, staff	Fri 10:30; start −2.5h; staff −1h	text	none	Camellia, WS	0
Application link after tour	prospect	tour start +60 min, once	email + text	link mode only	Camellia, WS	12 / 0 / 0 / 0
Toured, never applied	prospect	+1d, +3d, +7d from link	text	none set	Camellia, WS	33 / 0 / 0 / 3
Staff check-in after tour	tour host	+30 min after tour end	text	company switch: WS on, Camellia off	WS	≤15 WS
Staff tour emails	property inbox	immediate	email	none	all	18 / 26 / 0 / 113
Renewals	resident	Day 0 letter+text+call; +2d, +9d; −60/−30/−15d	text, MMS, voice	renewal policy	Camellia (1 of 2)	32 / 0 / 0 / 61
Past-due rent	resident	+21d, +28d, +31d; ≤3/mo; human approves	text	none	Camellia	0 sent, 17 awaiting approval
Work-order question unanswered	resident → tech/manager	+12h nudge, +24h escalate	text	none	Camellia	0
Visit heads-up / check-in	handyman	window −10 / +10 min	text	none	Camellia	0
Rating after repair	resident	once	text	on unless off	Camellia	11
Vendor dispatch / quote chase	vendor	+1d, +3d	email	none	all w/ maintenance	0
Vendor redial	vendor	+4h, +24h, +72h, cap 6; NO quiet hours	voice	none	approved only	0
Manager to-do reminders	PM	first notice only (24h×3 retired)	email	Camellia in; WS out (70 props)	Camellia	n/m
Lease/transfer signed notice	accounting	once PDF fetched; retries forever	email	recipient lists	Camellia	5
Unit claim to confirm	PM	+0, every 24h ×3	text	none	all	n/m
Turnover charges nudge	PM	once +48h	text	1 Camellia property	Camellia	n/m
Weekly stale-lead list	company	Wed 16:00Z	email	Camellia set	Camellia	n/m
Weekly owner report	owner	Mon 15:30Z	email	1 Camellia property	Camellia	n/m
Daily leasing activity post	Slack	daily, edited in place	Slack	2 Camellia, 1 WS	Camellia, WS	n/m
No-reply-15-min watchdog	our #alerts (+1 resend)	once, then page hourly	Slack, email	none	all	n/m

Findings

Only real follow-up override in prod: Camellia quiet-lead, day 3 / day 6 (saved 07-23).
Six series sent nothing in 30 days: tour not confirmed, weekend check-in, vendor chases, vendor redial, work-order nudge, visit heads-up.
Past-due rent: 17 messages waiting for human approval at Camellia.
Situs: nothing runs yet (not live; 115 inquiries, 0 conversations).
The settings screen disagrees with the code in 7 places:
- Manager reminders are listed as every 24h x3, but only the first notice sends.
- Visit heads-up goes to the handyman, not the resident.
- Unit-claim reminders go to the manager, not the resident.
- The 60-day renewal nudge and the Day-0 letter are missing from the list.
- Post-tour timing counts from tour start +60 min, not tour end.
- Lease and transfer signed notices are not listed.
- Tour reminders are shown as 8am-9pm but have no quiet-hours hold.
Two series ignore quiet hours: tour reminders (1h before an early tour) and vendor redial.
Two never stop on their own: signed-lease PDF fetch and the 15-minute no-reply alert resend.
New-lead and quiet-lead series share one timetable but follow different rules (20h gap and grace only on new-lead).
23% of last month's outbound (1,762 of 7,724) belongs to deleted conversations, mostly test.

Sources

Outside sources: listed in the research section above.
#17387: one schedule for both series (2026-10-07)
#17371: skip vendor threads
#17419: retry when no mailbox
#17428 and #17450: run-next-touch tool