“I have 50,000 leads in my CRM and I’m not making any money from them”

Fifty thousand dormant records are not dead — they are unworked. Across our own Colliers-era database reactivation campaigns, dormant CRM leads converted to booked qualified appointments at 4.4% on average, with an 8.9% peak. On 50,000 records, 4.4% is roughly 2,200 booked calls. The blocker is almost never interest. It is workload.

At a glance — leads sitting in your database not making money:

• It is not an emergency. The list has been quiet for years. Nothing about it gets materially worse this fortnight.
• Next 24 hours, free: export the database, count five numbers, and ring 20 records yourself.
• Do not mail all 50,000 tomorrow. You get one first send. A single blast burns the sending domain, the number and the list.
• The measured anchor: 4.4% average / 8.9% peak, dormant record to booked appointment, our own campaigns.
• The arithmetic that decides it: five touches across 50,000 records is 250,000 attempts. That is not a staffing problem you can solve by pushing harder.

What is actually happening when I have 50,000 leads in my CRM and I’m not making money from them

Leads sitting in a database not making money is a symptom with one dominant cause, and it is not lead quality. The records stopped producing revenue for one structural reason: your follow-up finished before their buying decision did. Most of those people never said no. They said nothing, and nothing is not a decision. If your sales cycle runs 90 days and your follow-up sequence runs seven, then 83 of every 90 days of live buying intent happen after you stopped talking — and every one of those decisions was made by someone else.

A dormant database is not a list of people who rejected you. It is a list of people who outlived your follow-up.

Is it as urgent as it feels?

Mostly, no — and saying so is more useful than pretending otherwise. A database that has been quiet for two years does not degrade meaningfully in the next two weeks. There is no deadline here that we or anyone else can honestly put on you. What usually makes this feel urgent is a cashflow problem somewhere else in the business, and dormant records look like the cheapest available answer. They often are. They are still not a same-week fix: a properly segmented reactivation sends in waves, so first booked calls typically land weeks after the first send, not days.

There is exactly one real deadline in this situation, and it is yours, not ours: if you are migrating CRMs, cancelling a subscription or winding up a platform contract, export everything today. Databases are lost to expired logins far more often than they are lost to decay.

What to do in the next 24 hours (none of this costs anything)

1. Export the whole database to CSV, including created date, last activity date, source and any opt-out flag. Twenty minutes in any CRM.
2. Count five numbers: total records; records with a valid mobile; records with a deliverable email; records with a last-activity date inside 18 months; records already flagged unsubscribed. Until you have those five, “50,000 leads” is a story, not an asset. It is common for the workable number to be well under half the headline count.
3. Find the consent evidence for one segment. Where did those records come from, and what did the person actually agree to? If you cannot answer that for a segment, that segment does not get contacted this week.
4. Ring 20 records yourself from the newest bucket, at a reasonable hour. Twenty calls is the cheapest test in the business: it tells you whether the list is alive, whether the numbers still connect and what people say when they hear your name, before you spend a cent.
5. Do nothing else. Specifically, do not send to the whole list tomorrow.

What to do in the next 7 days

Suppress first, segment second, send smallest-first. In order: remove opt-outs, hard bounces and disconnected numbers; remove current customers and anyone sitting in an open deal, because double-contacting a live opportunity is how a reactivation campaign loses money instead of making it; then dedupe, which is where most databases shed their first 10–20% of records. Our CRM data hygiene framework covers the suppression and dedupe sequence in detail.

Then split by record age using the ladder below, write one message per segment rather than one message for the list, and send to your smallest warm segment first — 500 to 1,000 records. Measure four things on that wave: delivery rate, reply rate, opt-out rate and booked rate. Those four numbers set the plan for the other 49,000.

The 4.4% ladder: record age, realistic response, and which channel goes first

This is the planning table we use. The 4.4% blended rate is measured — dormant CRM record to booked qualified appointment, averaged across the reactivation campaigns we ran in the Colliers era for Australian buyer’s agents, brokers, planners and coaches, with 8.9% as the best single campaign on record. The multiples are how we plan segments off that anchor, not a second measurement. Full method and context are on our 4.4% average and 8.9% peak reactivation write-up.

The top row is why our 8.9% peak campaign is not a fluke: it was a young, high-intent segment, which is exactly what the ladder predicts. The bottom row is the one operators get wrong — a three-year-old record is a re-permission exercise, not a sales campaign, and treating it as a sales campaign is what generates complaints.

“They’re cold and they’ll complain” — the consent question, answered honestly

Two different things get confused here. An unsubscribe is not a complaint; it is the mechanism working, and every unsubscribe is a record you were never going to convert removing itself for free. A complaint is a regulatory event with a defined test, and the test is about consent and about how fast you stop.

In Australia, the ACMA’s fact sheet on the Spam Act 2003 requires that an unsubscribe option is “actioned within 5 working days”, that it costs nothing, and that it still works for at least 30 days after the message was sent — and the same fact sheet is explicit that if another business sends on your behalf, “your business is liable, and you cannot outsource your risk”. In the United States, the FCC’s 2024 TCPA Report and Order (FCC 24-24) requires callers to honour do-not-call and revocation-of-consent requests “within a reasonable time… not to exceed 10 business days after receipt”.

The practical read: age is not the legal problem. Consent and your opt-out plumbing are. A five-year-old record with documented consent and a working stop mechanism is safer to contact than a six-month-old purchased record with neither. Our outbound compliance checklist sets out the operational side of that.

Where we are not the answer: if the records came from a bought list, if you cannot evidence consent for a segment, or if the database carries health, credit or other sensitive data, the next call is to your privacy officer or a lawyer, and to the regulator’s own guidance — the ACMA in Australia, the FCC and FTC in the United States. That is a legal question, not a marketing one, and nothing on this page is legal advice.

Why 50,000 records is not a human workload at any staffing level

Run the arithmetic with your own numbers; the shape does not change. Five touches per record across 50,000 records is 250,000 contact attempts. Assume a dialler-equipped person sustains 60 connect attempts a day, which is generous over a full week. One attempt per record is 50,000 dials, or 833 person-days — about 3.3 years for one person at 250 working days a year. Five people get one attempt across the list in roughly eight months, by which point the freshest records are eight months older and have slid a rung down the ladder. Five attempts each is 4,167 person-days: 16.7 person-years. Multiply that by your own fully loaded cost per head and the number speaks for itself.

The part that actually breaks teams is not the outbound volume, it is the inbound. A wave of 5,000 sends produces replies within minutes, concentrated in the first two hours, and a reply that waits until tomorrow is a reply you have already lost. Manual reactivation fails at the reply queue, not at the send button. That inbound load is the specific problem AI SMS and voice agents exist to absorb, and it is what our database reactivation service is built around — you can also work down through the rest of our database reactivation cluster for the campaign mechanics.

The fix that stops the database going dormant again

Reactivation is a clean-up. The recurrence fix is one rule: every record carries a next-action date, or it is dormant by definition. A CRM does not have a dormant-lead problem; it has a next-action-date problem. Any record with no next action and no activity for 90 days should auto-enrol into long-term nurture rather than quietly aging in a list view.

Then measure it. Put the dormant-record count on the weekly dashboard next to pipeline. If that count grows week on week, your follow-up sequence is shorter than your sales cycle, and you are rebuilding the exact problem you just paid to fix.

Frequently asked questions

What do I do with old leads that are three years old?

Treat them as a re-permission exercise, not a sales campaign: a single email that reintroduces you and asks whether they still want to hear from you, then archive the non-responders. On our ladder that segment plans at roughly a quarter of the 4.4% blended rate. Working it harder than that is where complaints come from.

Is it legal to text or call leads who haven’t heard from me in years?

Age itself is not the test; consent and your opt-out handling are. Australia’s Spam Act 2003 requires an unsubscribe that is “actioned within 5 working days” and still works for at least 30 days after sending, per the ACMA’s own fact sheet. In the US, the FCC requires revocation requests to be honoured within no more than 10 business days. Check your consent records before your copy, and take legal advice for anything sensitive.

How many appointments should 50,000 dormant leads actually produce?

Use 4.4% as the planning blend and expect the young segments to carry it: 50,000 records at 4.4% is about 2,200 booked qualified appointments, but only after suppression, and your workable count will be lower than your headline count. Run the ladder segment by segment rather than applying one rate to the whole file.

Will reactivation actually increase revenue, or just activity?

Only if the offer and the sales process behind it already convert. Reactivation puts conversations on the calendar; it does not fix a call that does not close. Our measurement methodology defines our 7x average sales lift as trailing three-month closed-deal revenue at month six over the three months before launch, and discloses on the same page that the median is closer to 4x — the average is not the typical case.

Should I clean the list before I send, or just send?

Clean first, always. Sending to unsuppressed records damages deliverability for every future send, and deliverability is the one asset you cannot buy back quickly. Dedupe, suppress and validate, then send to 500–1,000 records before you touch the remaining 49,000.