How to grow a recruitment agency without renting your candidate flow.

This is written for a principal running an operating agency, somewhere between $2M and $8M in billings, with consultants on the desk and a growth line that has gone flat.

If you are starting an agency, this is the wrong article. You need clients, cash flow and a licence check, in that order, and none of that is here. The setup list for a new agency is over here instead.

For everyone else: the plateau usually has a structural cause, and it is rarely the one people name.

Why agencies plateau

Ask a principal why growth stalled and you will hear a version of three answers. The market softened. We lost a big account. We cannot find good consultants.

All three can be true and none of them explains why revenue tracks headcount so tightly. The structural answer is that most agencies have no demand asset. They have two rented channels.

Candidate flow is rented from job boards. You post, you pay, you get applications. Stop paying and it stops within a week. Worse, the board owns the relationship and sells the same candidates to the agency down the road.

Client flow is rented from the founder’s network. It worked brilliantly for the first five years, because the founder knew everyone. It does not scale, because the founder has a finite address book and a finite calendar, and no new consultant inherits either.

Both channels have the same defect: nothing accumulates. Year six starts from roughly where year five started, minus the relationships that moved on. The only lever left is more consultants, which is why revenue and headcount move together and why margins do not improve with scale.

The rented-channel test

There is a quick way to find out how exposed you are. Take the three channels that produce most of your candidates and most of your clients, and for each one ask two questions.

If the price doubled tomorrow, what would we do? For a job board the honest answer is usually “pay it”. There is no substitute channel to switch to, so the board can set the price. Boards understand this, which is why posting prices move the way they do and why the negotiating position is poor in every renewal conversation.

If it disappeared tomorrow, how long until we replaced it? A founder’s network is difficult to transfer. An owned audience, email list or site with useful search visibility remains with the agency.

Most agencies run this test and find that every channel they depend on fails both questions. That is not a reason to stop using job boards, which remain the fastest way to fill a live role. It is a reason to stop treating them as a strategy, because a channel you cannot leave controls your margin instead of you.

Compare annual job-board spend with the cost of building an owned audience and search presence. Boards remain useful for live roles, while owned content and contact data can keep producing discovery after the initial work is paid for.

The two-sided problem nobody accounts for

Here is where most marketing advice fails recruiters. It assumes one audience.

A recruitment agency has two, and they want opposite things from the same website.

An employer wants evidence you can fill a role that has been open for four months. They want proof of specialisation, a sense of your market knowledge, and enough signal that the introduction call will not be a waste of an hour.

A candidate wants roles, salary information, and a reason to believe you will not treat them as inventory.

Point one website at both and you usually get a homepage that says “we connect great people with great companies”. It attracts no search traffic, because nobody searches that exact phrase.

The workable arrangement is one content operation producing two distinct streams that meet at the brand.

The traffic asymmetry, stated plainly

Candidate-side content is cheap to produce and pulls high volume. Salary guides, role explainers, career-path pieces, and interview content all attract searches, and there are a lot of candidates.

Employer-side content is scarce, hard to write, and attracts very little traffic. There are perhaps a few hundred people in your city with authority to appoint a recruitment partner in your specialisation, and they search rarely. One employer enquiry is worth more than a thousand candidate visits.

Agencies that measure content by sessions can overproduce candidate content because it attracts more traffic. The report can improve while employer enquiries remain flat, so the two audiences need separate targets.

Employer-side demand: how buyers actually search

Employers do not search “recruitment agency”. They search the problem.

They search the role and the city. They search a salary benchmark because they need to know whether their budget is realistic before they brief anyone. They search whether a role can be filled at all in the current market. They search a specific credential or a visa condition. Sometimes they search a competitor’s name.

What ranks for those searches is almost never an agency, because agencies write about themselves. “We place great people” ranks for nothing, because nobody types it.

The pages that work are the ones that answer the employer’s actual question before the enquiry:

Salary and market-rate content for your specialisation. The most-used employer research asset in recruitment, and the one that most obviously demonstrates market knowledge. Update it annually with a stated methodology and a date.

Role scarcity and time-to-fill reality. What a role genuinely takes to fill in this market, and why. Uncomfortable to publish and disproportionately persuasive, because it is the conversation the employer is already having internally.

Hiring process content. Structured interviewing, assessment, offer negotiation, counteroffer risk. This is you being useful before you are engaged.

Specialisation depth. Ten pieces on one niche beat 50 across everything. Depth is what both search engines and AI assistants use to work out what you are authoritative about.

Proof from your own placements. Time to fill, offer-accept rate, retention at 12 months. Most agencies will not publish these figures, which is what makes the ones that do stand out to a buyer comparing several firms.

Do this well and the employer-side content also becomes your consultants’ business-development material, which is a second return on the same spend.

Where employer branding fits is a longer conversation: your clients’ inability to attract candidates is often why your roles stay open, and helping fix it is a service you can charge for. We cover that separately in employer branding for recruitment agencies.

Candidate-side demand: cheap, useful, and easy to over-invest in

Candidate content earns its place for three reasons, none of which is traffic volume.

It builds name recognition that makes outbound calls answerable. A candidate who has read your salary guide takes the call. That is the whole return, and it does not show up in any analytics report.

It builds a talent pool the agency owns rather than rents. An email list of qualified candidates remains available between live roles; a job-board subscription is a recurring cost.

It generates the raw material for employer-side content. Every conversation about why candidates decline offers is market intelligence, and market intelligence is what employers pay attention to.

The discipline is to keep it proportionate. Set a ratio in advance, spend the majority of your effort on the employer side despite the traffic numbers arguing otherwise, and review the split quarterly against enquiries rather than sessions.

What this looked like in one account

AI Talent on Demand is a founder-led AI and technology recruitment specialist in Australia. The business had strong networks and needed search to introduce it to candidates and buyers beyond them. We built the positioning, the website, and a compact content system from launch.

In absolute terms, monthly organic clicks grew from 22 in the baseline month to 557. Monthly search impressions grew 22×, from roughly 3,400 to 76,000. As at August 2026 the site held page-one positions for more than 500 keywords.

The important caveats, stated in full. These are Google Search Console search-visibility metrics, not placement or revenue results. The site launched in November 2025 and the SEO programme began on 24 March 2026, with February 2026 as the baseline month, so the multiples run from a small base. Page-one count means queries averaging positions 1 to 10 with at least five impressions, over the 28 days to 9 August 2026.

What generalises from it is the sequence rather than the multiples. Positioning first, so the site says what it is specialist in. Then the pages that answer what buyers and candidates in that specialisation actually search. Then volume, once the shape is proved. A founder-led firm with genuine niche depth and no search presence has more headroom here than a generalist agency with a large existing footprint, and the numbers reflect that.

The specialisation decision, worked through

Step one of the sequence below is the one principals stall on for years, so it is worth walking through properly.

The objection is always the same: narrowing loses revenue. A generalist agency that names one market appears to be turning away everything else.

In practice it turns away almost nothing, because existing clients do not read your homepage. They already know you and they call you. The homepage is read by strangers, and a stranger can only hire you for something you have actually claimed on the page.

Consider two versions of the same firm.

Generalist. “We are a leading Australian recruitment agency delivering talent solutions across a range of industries.” This ranks for nothing, differentiates from nothing, and gives an employer no reason to choose it over the four other firms on the list. It also gives your consultants nothing to lead a business development call with.

Specialised. “We place allied health professionals in regional Victoria and New South Wales. Physiotherapy, occupational therapy and speech pathology, permanent and locum.” Every noun in that sentence is something a buyer searches, every claim is checkable, and a physiotherapy service manager in Bendigo now knows within two seconds that this is the right call to make.

The second firm can still fill a finance role for an existing client; it just doesn’t put that on the homepage.

How to choose, if it is not obvious:

• Where is your fill rate highest? Look at the last two years. Specialise where you already win, not where you wish you won.
• Where do clients return? Repeat business is the market that values what you actually do.
• Where do your consultants have real language? Depth is not fakeable. Buyers in specialised markets detect a generalist inside one conversation.
• Is the market big enough? A specialisation supporting only three potential clients is a job, not a strategy. Somewhere between a few dozen and a few hundred realistic buyers is the workable range.

You can hold more than one specialisation. Each needs its own section of the site, its own salary content and its own pages. Holding none of them means a stranger has no way to find you.

Who does this work

The plan above fails most often on resourcing rather than on strategy, so be honest about the options before starting.

A consultant writing content between placements. It appears inexpensive, but billing work correctly takes priority. Few agencies sustain this model without protected time and editorial ownership.

A junior marketing hire. A full-time salary plus on-costs, for someone who can run social channels and update the website, and who has neither the market knowledge to write a credible salary guide nor the seniority to challenge a principal on positioning. They usually end up doing candidate content, because it is the part they can do alone, and the traffic asymmetry above explains where that leads.

A generalist digital agency. Competent at ads and web build, and typically unable to write persuasively to a hiring manager in your specialisation, so you end up writing the substance yourself while paying them for the layout and the ad spend. The channel itself is worth buying well: see Google Ads and Meta ads management.

A fractional arrangement. Senior marketing capability for a few days a month, which for most agencies at this size matches the actual shape of the work: strategy and editorial judgment need to be senior, production does not need to be full-time. We have written up the arithmetic on fractional marketing teams separately.

Whichever you pick, two roles must exist. Someone senior owns positioning and edits everything for market credibility. Someone else does production. Collapsing those into one junior person is the most common and most expensive version of this mistake, because the output can look fine to anyone outside the market while a buyer inside it spots the gap in the first paragraph.

One more resourcing point that costs nothing. Your consultants are the market-intelligence source, and 20 minutes of recorded conversation about why an offer was declined is worth more than a day of desk research. Build the extraction habit early, so the content keeps sounding like a recruiter talking rather than a marketing department describing recruitment from outside.

Measure booked meetings, not clicks

The scorecard for this work has one number on top: qualified employer conversations booked. Everything else is a supporting indicator.

Track alongside it:

• Enquiries by source, separated into employer and candidate
• Time from first site visit to enquiry, which tells you whether the content is doing the persuading or your consultants are
• Candidate contact-rate on outbound, which is where name recognition shows up
• Roles won without a competitive pitch, the cleanest signal that positioning is working

Demote clicks, sessions, rankings, and impressions to a second page of the report. Treat them as diagnostics rather than outcomes — a recruitment content programme judged on them will optimise itself into producing candidate content forever.

One caution on attribution. Search rarely gets credit for an employer enquiry, because the buyer reads three articles in March, meets you at an event in May, and enquires in July naming the event. Ask on the enquiry form where they first heard of you, separately from what prompted them to make contact. The gap between those two answers is the content’s real contribution.

A 90-day sequence

Five steps, in order. A principal can start these without an agency.

1. Pick the specialisation you want to be known for and say it on the homepage. One sentence naming the market, the roles and the geography. This is the decision most agencies avoid and every subsequent step depends on it.
2. Publish a salary guide for that specialisation. State the methodology, state the sample, date it. This is the single highest-value employer asset in recruitment and most agencies still do not have a current one.
3. Write the five questions your buyers ask in the first meeting, as five pages. You already know them. Your consultants repeat them weekly.
4. Fix the enquiry response time. Measure it first; most agencies find it is far worse than assumed outside business hours. An enquiry answered in five minutes and one answered next morning are different businesses.
5. Set the scorecard before the traffic arrives. Agree with your leadership team now that this is judged on employer conversations, so that at month four nobody points at a session count.

Continue for four quarters and judge progress against employer conversations, candidate response and roles won. Organic search usually builds slowly; stopping after a few months often means judging it before enough useful pages have had time to rank.

Frequently asked questions

How do you get clients for a recruitment agency?

Consistent business development plus a demand asset that works while nobody is calling. In practice: choose one specialisation, publish the salary and market-rate content employers research before briefing, answer the questions they ask in first meetings, and publish your own performance data. Outbound then lands warmer because the name is already familiar.

How do you grow a recruitment agency without adding consultants?

Increase revenue per consultant rather than consultant count. The levers are better client mix, higher fill rates through specialisation, faster response times, and inbound enquiries that arrive without business-development hours attached. Marketing affects three of the four.

How long does content marketing take to work for a recruitment agency?

Expect candidate-side traffic within three to six months and employer-side enquiries between six and 12. Employer demand is scarcer and the buying cycle is longer, so the employer side always lags. Budget for four quarters before judging it.

Should a recruitment agency market to candidates or clients?

Both, with effort weighted toward clients despite the traffic numbers arguing otherwise. Candidate content is cheap and high-volume, and it builds the name recognition that makes outbound work. Client content is scarce and low-volume, and it is where the revenue is. Set the ratio deliberately, because measuring by sessions will drift you to the candidate side automatically.

Do you need a licence to run a recruitment agency in Australia?

Requirements vary by state and by the type of placement, and labour-hire licensing applies in several jurisdictions. Check the current rules with your state regulator, because this is not a marketing question and the answer changes.