“My referrals have dried up” — what changed, and what replaces them

“My referrals have dried up” almost always means one measurable thing changed: your active-referrer count — the number of people you actually delivered for in the last 18 months. Referred work is proportional to that number, not to how many people know your name. The ABS puts annual job mobility at 7.2%, so a network ages even when you do nothing wrong.

• The test: count people who have referred you at least once ever, then count how many of those you delivered for or genuinely spoke with in the last 18 months. The gap between the two numbers is your decay.
• Noise or decay: enquiries down 4–8 weeks with a stable active count is seasonality. Two quarters down with the active count halved is structural.
• Next 24 hours: list the active referrers. Nothing else.
• Next 7 days: re-contact them, and work your own dormant enquiry list before you buy a single cold lead.
• The honest order: warm first, cold last. Cold outbound is the slowest and most expensive way to replace referred work.

Is it seasonal noise, or have my referrals actually dried up?

Half of what feels like collapse is a normal trough. The distinguishing test is not enquiry volume — that is the symptom both causes produce — but whether the supply side moved: the number of people currently in a position to refer you.

The rule worth writing on the wall: two quiet months is noise, two quiet quarters with a falling active-referrer count is decay. Only the second one justifies rebuilding anything.

Why do consulting referrals stop when nothing went wrong?

Referral networks are not assets that sit still. They decay unless something replenishes them, and the replenishment is delivery — being in the room, being remembered accurately enough to be described to a third party. Three forces run against you at once, and none is about the quality of your work:

• People move. Just over 1.0 million Australians changed employer in the year to February 2026, a job mobility rate of 7.2% (ABS, Job mobility, February 2026) — down from 7.7% the year before, so this force is easing slightly rather than worsening. Compound 7.2% over three years and roughly one in five of a three-year-old contact list has changed employer at least once; fewer in practice, because a minority of people move repeatedly. That is the share of your referrers who now sit against a new budget, a new procurement process and an incumbent supplier.
• Businesses close. The ABS counted 2,814,778 actively trading Australian businesses at 30 June 2026, against an exit rate of 13.8% for 2025–26. That is the whole economy including sole traders, not your list specifically — but it is the right order of magnitude for how fast any contact list ages.
• Relationships fade without contact. The nearest real evidence is not from marketing at all: Roberts and Dunbar followed 25 English school-leavers and the 1,291 people in their personal networks across an 18-month life transition. Emotional closeness to friends fell significantly over that period (b = −0.62, p < 0.001) while closeness to family rose (b = +0.27), and the decline in friendships was mitigated — not prevented — by effort put into the relationship (Human Nature, 2015). Read this as a mechanism only. Twenty-five teenagers leaving school is not a sample of anybody’s referral network, the study measures felt closeness rather than work sent, and none of its coefficients are a rate at which referrals decay. What it establishes is the general shape: unlike family ties, chosen relationships weaken when contact stops, and effort slows that rather than stopping it. Do not read a number out of it.

Put together: a referral business does not fail, it expires, on a schedule set by how long ago you last delivered for someone.

The active-referrer rule — the arithmetic behind referral decay

The active-referrer rule: referred work is proportional to the number of people you delivered for in the last 18 months, not to the number of people who know your name. Five columns below, four of them in the formula — and all five are already in your invoices and inbox:

Referred revenue per year = B × C × D × E. In the example: 7 × 0.9 × 0.5 × $32,000 = about $101,000. Run the same arithmetic on the active count you had two years ago — say 19 — and it produces 19 × 0.9 × 0.5 × $32,000 = about $274,000. Nothing changed about the work, the pricing or the market. Only B moved.

The decision that falls out of it: if B × C × D × E is below the revenue the business has to cover, the gap is structural, and waiting is not a plan. It also tells you exactly how big the replacement channel has to be, which is the number most rebuild plans never establish.

What to do in the next 24 hours

Nothing here costs money, and none of it involves hiring anyone. It exists so that tomorrow you are working from a list rather than from a feeling.

1. Build column A and column B above. One spreadsheet, two columns, an hour or two. Most consultants are surprised by B, and the surprise is the diagnosis.
2. Rank referrers by work sent, last 24 months. If one name is over 40% of your revenue, you have a concentration problem to solve before a network problem.
3. Pull every unconverted enquiry and unsigned proposal from the last 24 months into the same sheet. This is your dormant list, and it is the cheapest pipeline you own.
4. Send three messages tonight — to the three people on the active list you would most want to work with again. Not a pitch: a specific, useful sentence about their business. That is the entire message.

If your bank position is the actual emergency rather than the pipeline, that is a conversation with your accountant before it is a marketing problem, and the free cash flow statement template on business.gov.au’s cash flow guidance is the version of that conversation that costs nothing.

The next 7 days: work the warm list before you buy a single cold lead

The fastest replacement for referred work is not new demand. It is demand you already generated and never converted — past clients, dead proposals, enquiries that went quiet. In our own campaigns it outperforms cold outreach by a wide margin, because the trust that referrals were supplying is already in the relationship.

Across the AI reactivation campaigns we have run for Australian brokers, planners and high-ticket consultants — including Colliers — dormant databases have produced 4.4% average and 8.9% peak, measured as dormant leads converted to booked qualified discovery calls. That is our own record in high-trust AU verticals, not an industry benchmark, and the method is set out on our database reactivation results page. The levers that move that rate — channel by record age, reply latency, touch two — are broken down in how to increase your database reactivation rate.

Run it yourself in week one: 40 records a day, personalised, from your own address, with a reply window you can staff. The honest cost is time and latency — replies arrive at 9pm and go cold by morning, and a domain that has never sent volume lands in spam if you push it hard.

Below roughly 150 records, doing it yourself is the right answer and nobody should be selling you anything. Above that, the constraint stops being the list and becomes who answers it.

“Outbound isn’t how my industry works”

Partly true, and worth conceding precisely. Cold outbound into bespoke advisory work has a long ramp and a low reply rate, and done clumsily to people who know you it costs standing. Buying lists in a market of a few hundred qualified buyers is close to useless.

What is not true is that the only alternative to referrals is cold email. The sequence that works for advisory businesses runs warm to cold: your own dormant enquiries first, then lapsed clients, then the second-degree network your active referrers can open, then documented expertise that makes you findable, and only then genuinely cold contact into named accounts. Each step has a lower cost per booked conversation than the one after it. The ranked version of that channel list, with what each one costs, is in how consultants get new clients in Australia, and the benchmark numbers sit in cost per booked meeting for consultants.

The point of a rebuild is not to replace referrals. It is to stop the business depending on a number — your active-referrer count — that you were not measuring and could not control.

Frequently asked questions

My referrals have dried up — how long should I wait before I do something?

Waiting is only sensible while the active-referrer count is stable. If you have been delivering steadily and enquiries are down for eight weeks, that is within normal variation for a referral business. If you have been heads-down on one large engagement for a year and your active count has fallen from 19 to 7, waiting makes it worse: the arithmetic above compounds against you every month you do not replenish.

Is it me, or has the market gone quiet?

Check the supply side before you conclude anything about demand. In the year to February 2026 the ABS recorded a job mobility rate of 7.2% and, separately, a business exit rate of 13.8% for 2025–26. Between the two, a meaningful slice of any three-year-old contact list is in a different seat or gone. That is usually a bigger effect than anything happening to the market you sell into.

Do referrals come back on their own?

Only if delivery resumes. Referrals are a lagging output of delivery volume, so they return when you are back in enough rooms to be described accurately to a third party — which is why the first repair is re-contact, not a new website. The lag is typically a full sales cycle behind the work that generates it, so a network rebuilt this month shows up in enquiries next quarter, not next week.

Isn’t reaching out to old contacts going to look desperate?

Only if the message is about you. A note that references something specific about their business, sent by name from your own address, reads as a professional keeping in touch — which is what it is. The version that reads badly is the mass one: same paragraph, obvious merge fields, sent to 400 people at once.

What actually replaces referred work fastest?

Your own dormant list, on the evidence we have. The 4.4% average and 8.9% peak above are dormant leads converted to booked qualified discovery calls, from our AU campaigns in high-trust verticals; cold channels in the same businesses take a quarter or more to produce comparable volume. For consulting specifically, we have set out the model, including where it does not suit bespoke work, on our lead generation for consultants page.