There is a particular kind of disappointment that arrives at month five of a link building retainer. The links exist. They are on real websites. The invoices have all been paid. And the page you were trying to move has not moved. Nobody has done anything dishonest, and yet the money is gone.
In almost every case we have seen, that outcome was decided before the contract was signed — not by the agency's competence, but by the absence of a brief. The client asked for "link building", received a proposal for link building, and only discovered at month five that the two parties had meant entirely different things.
So write the brief first. It takes about ninety minutes, it costs nothing, and it changes the conversation from "what do you sell" to "here is what I need — can you do it". Eleven points.
1. The exact URL you are trying to move
Not the domain. Not "our blog". One URL, or at most three, with the reason each one matters commercially. "Links to the homepage" is a request with no success criterion attached, and an agency will happily fulfil it forever.
If you cannot name the page, you are not ready to hire. Go and find out which pages actually generate pipeline, then come back. That exercise alone frequently changes the answer.
2. The term you want that URL to own
One primary term, plus the two or three variants you would also accept. This is what makes the engagement measurable. It also lets any competent agency tell you within a day whether your target is realistic, which is worth knowing before you commit a quarter's budget.
3. Who is currently above you, and by how much
List the four URLs outranking you. Pull each one's referring domain count. Put your own next to it. That single line — "we have 19, they average 78" — reframes the entire negotiation, because it converts a vague ambition into a countable gap.
Any agency that quotes you a monthly link volume without doing this subtraction is guessing, and the guess will be shaped by what they want to sell rather than what you need to buy.
4. What has already been tried, and what it produced
Previous agencies, freelancers, marketplace purchases, in-house attempts. Be specific about what was delivered, because it determines whether the first quarter should be acquisition or cleanup. Roughly one engagement in five should begin with a disavow file rather than an outreach campaign, and nobody discovers that from a sales call.
5. Your budget, stated honestly
The instinct to hide the number is understandable and counterproductive. Link building has a hard cost floor: manual prospecting, scoring, pitching, writing and an editor's time cannot be delivered below roughly $150 per placement by anyone, anywhere. Below that price the inventory is coming from somewhere else.
Publishing your budget lets an agency tell you what it genuinely buys. Withholding it produces a proposal reverse-engineered from what they think you will accept.
| Monthly budget | Realistic output |
|---|---|
| $1,500–$3,000 | 5–10 quality placements, or one linkable asset per quarter instead |
| $3,000–$6,000 | 8–16 placements plus basic monitoring |
| $6,000–$15,000 | 16–40 placements plus digital PR |
6. Your approval capacity
This is the single most underestimated line in the brief. Somebody on your side must be able to read and sign off a draft article within three working days, repeatedly, for a year. If that person does not exist, the campaign will run at half speed regardless of what you pay.
Name them in the brief. If you cannot, say so and let the agency price the delay in.
7. What you will not accept
Publications you refuse to appear on. Competitors you will not be listed beside. Claims your legal team will not allow. Whether paid inclusions are acceptable at all, and whether you need them disclosed.
Getting this on paper early saves a genuinely awkward conversation at month three.
8. The reporting you need, and who reads it
A report for a growth lead and a report for a CFO are different documents. Specify which columns you need, how often, and in what format. Our own bias is a live spreadsheet over a monthly PDF — you can check a spreadsheet on the day something goes live rather than four weeks later — but the important thing is that you state a preference rather than accepting whatever arrives.
9. Your timeline and what happens if it slips
Be explicit about the horizon. Three to six months for ranking movement on a mid-competition term; nine to twelve for a contested category head term. If you have a board meeting in ninety days that needs a number, say so — an honest agency will tell you link building is the wrong instrument for that deadline and suggest something else.
10. Category exclusivity
Ask whether the agency currently works with any of your direct competitors, and whether they will contractually agree not to during your engagement. Some will, some will not, and both answers are legitimate. What matters is hearing it on the first call rather than the fourth.
11. Your exit
Minimum term, notice period, and what you keep when you leave. Gap analyses, prospect lists and content you paid for should transfer to you. If they do not, you are renting your own strategy.
The three questions that do most of the filtering
If you send nothing else, send these. In our experience they narrow a ten-agency shortlist to two or three in a single round.
Where does your inventory come from? A real answer names a method — manual outreach, existing relationships, a prospect index built in-house. A weak answer names a price.
Can I see the domains you rejected last month, and why? Almost nobody keeps this. The ones who do are usually worth talking to.
What happens if a link disappears in month eight? The good answer is a written replacement warranty. The revealing answer is "that doesn't really happen".
What a good response looks like
You are not looking for enthusiasm. You are looking for a proposal that engages with your specific numbers — that references your gap, your budget, your timeline — and that contains at least one thing you did not ask for and did not want to hear.
The best proposal we ever lost to a competitor opened by telling the client their target keyword had almost no commercial search volume and they should pick a different one. They were right. We had missed it.
A proposal that agrees with everything in your brief is a proposal written by someone who has not checked your brief. Somewhere in a good response there should be a sentence that begins "we would push back on point four".
One thing the brief cannot do
It cannot tell you whether the work will actually be done well. For that you need to audit live placements from the agency's last quarter, which is a separate exercise and takes about ninety minutes. We have written that one up separately.