Specialist versus full-service agencies

“Full-service” is a phrase from a capabilities page. It is not a description of a building. In the Australian market it usually means: we will not send you away if you mention a channel we do not staff. Sometimes that is a genuine bench. Sometimes it is a media pair, a contracted designer, and a quiet arrangement with a specialist SEO shop whose name never appears on the invoice.

The useful question is not which label you prefer. It is whether you have the time and the skill to integrate the work yourself. One shop is a management choice. A roster is a management job. Both can be right. Pretending they are the same thing is how mid-market teams end up with four vendors, no owner, and a Tuesday that is only status meetings.

This sits inside the wider problem of choosing an agency. Pick the structure before you pick the brand.

What specialist means here

A specialist, in this country, is a shop that will decline work. That is the tell. An SEO practice that will not “just handle the Google Ads” is a specialist. A paid-media shop that refuses to write your brand platform is a specialist. A social studio that will not pretend it can do technical search is a specialist.

The point of that refusal is depth. Specialists live in the ugly parts of one discipline: feed hygiene, crawl budget, query mapping, incrementality tests, creative fatigue, community moderation, the particular way Meta’s attribution lies to a finance manager. They have seen the same failure often enough to recognise it in the first week.

The cost is the edge of their world. A search specialist will not save a product page that cannot explain the offer. A paid-media shop will not invent a brand you have not funded. If the constraint sits outside their lane, they will either go quiet or sell you a neighbouring service they should have declined.

Specialists also concentrate in the same two cities as everyone else, with good independents in Brisbane, Perth and Adelaide. The specialist premium is not a postcode. It is the roster: fewer generalists, more people who still do the work themselves.

What the org chart usually shows

Ask a self-described full-service shop to draw the team that would sit on your account, then ask which of those people are employees. The honest drawings are smaller than the website.

A common Australian pattern is a core of account, strategy and one or two channel leads, with production, development, incrementality, or “content at scale” sitting in partner studios. Another pattern is a network office that can theoretically reach into Singapore or London and, in practice, will not do so for a $15,000 retainer. A third is a founder-led independent that was excellent at one thing and hired a generalist so the capability page would look even.

None of those is a moral failure. The failure is the pretence. If the work will leave the building, you need the name of the building. You also need to know whether you are paying a margin on a freelancer you could have hired yourself.

Read how to read an agency proposal with this in mind. The team slide is where full-service either becomes real or collapses into a supply chain.

The coordination tax

Every extra vendor adds a meeting, a brief, a disagreement about attribution, and a person who thinks the landing page is someone else’s job. That tax is paid in your calendar, not on their invoice.

Two shops can be cheaper in fees and more expensive in outcome if nobody owns the join. SEO wants the page thin and fast; paid wants the page long enough to qualify a lead; the CRM owner wants a field that neither page has. A full-service shop with a real integrator — an account lead who can make those three people sit down — is earning the extra fee. A full-service shop that just forwards Slack messages is charging you to be a receptionist.

Be honest about your own bench. If you have a marketing manager and a founder who approves captions on Sundays, you do not have an integration function. You have a bottleneck. In that shop, one agency is often the less romantic and more accurate answer.

If you have a capable in-house lead, a clean brief, and the stomach for two sets of numbers that will not match, a roster can outperform a generalist. The in-house lead is the product. The agencies are components.

When one shop is the cheaper answer

One throat to choke is worth paying for when the work is interdependent and the budget is not large enough to fund proper specialists in each lane.

A mid-market company spending $8,000 a month on labour and $20,000 on media does not need an SEO agency, a media agency and a social studio. It needs a plan that will not trip over itself, and a single reporting line finance can read. The quality ceiling will be lower than a specialist roster. The waste will be lower too.

One shop is also the cheaper answer when the problem is political. If sales, product and marketing do not agree on the offer, three agencies will not agree for them. They will amplify the argument and invoice it. Hire one grown-up team, or fix the offer first.

Seasonality points the same way. If you go dark in January and spike in May for EOFY, a single retainer that can flex hours across channels is less clumsy than three notice periods.

White-label, partners and the work you thought you bought

Ask the ugly question in the first week, not the last: who actually does the specialised work, and do we ever meet them?

White-label is widespread. A full-service shop wins the relationship, then a specialist does the technical audit, the feed, or the creative under another letterhead. You may never hear the specialist’s name. That can still be fine if the integrator is strong and the specialist is stable. It is not fine if you are paying for senior judgement and receiving a PDF from a person who cannot email you.

Put partner names in the scope. Put a rule in the agreement that subcontractors who touch personal information are disclosed, because that is an OAIC problem when it goes wrong. Put a rule that you can speak to the specialist if the work is technical. If the shop refuses, they are protecting a margin, not a method.

The inverse exists too. Specialists quietly add “full-funnel” pages and then subcontract the parts they used to decline. The label moved. The org chart did not.

Running a roster without becoming the project manager

If you choose specialists, write the joins before you write the retainers.

Name the owner of the website, the owner of the offer, the owner of the accounts, and the owner of the number you will argue about. Give every shop the same brief excerpt so they cannot invent a private version of the strategy. Set one measurement source. Decide in advance who loses when paid search and organic fight over a brand term.

Run a single monthly meeting with all of them, not a daisy chain of recaps. The daisy chain is how bad news ages. Keep the group small enough that a mid-market manager can still do their actual job. Two specialists and one production partner is a roster. Five is a second career.

Write the exit while you are still polite. Specialists who hold a login can hold a category. The same ownership rules apply as with a full-service shop: your accounts, your data, their tools. The brief that makes this possible is the one in writing a brief agencies will actually read.

A choice you can defend to a CFO

Finance does not care about full-funnel. It cares about labour, media, duplication and delay.

Put both options on one page. Option A: one shop, one fee, named gaps, a lower ceiling. Option B: two or three specialists, a higher ceiling, an internal owner, and the hours that owner will spend. Include GST. Include the likely media treatment. Include the cost of a missed handover — a week of paused ads, a month of conflicting landing pages.

If option B only works because you assume the internal owner has spare time, it does not work. Australian marketing teams are already thin. A structure that needs a hidden project manager is a structure you will resent in quarter two.

Revisit the choice when the work changes. A specialist SEO retainer that was right at $6,000 can become the wrong object once you add international, a replatform and a marketplace. A full-service shop that was right for a launch can become a generalist tax once paid search is the whole job. The structure is not a personality. It is a response to a brief.

If a shop will not decline a channel, you are not looking at a specialist. You are looking at a brochure that has not met a roster.

Choose the smallest structure that can do the job without you inventing a department to hold it together. That sentence will keep you out of more bad retainers than any reel of integrated work.