We move the budget when the numbers say to.
Three quarters of one account. The mix at the start is not the mix at the end, and the reason it changed is a sentence anyone can check.
Paid social bought the cheapest clicks in the account and the fewest meetings, so over two quarters most of it moved to search, where the intent already existed, and to lifecycle email, where the audience had already paid us once.
Quarter one
Paid social, 46 per cent.Paid search, 22 per cent.SEO and content, 12 per cent.Lifecycle email, 8 per cent.Creative, 12 per cent.Quarter two
Paid social, 34 per cent.Paid search, 28 per cent.SEO and content, 16 per cent.Lifecycle email, 12 per cent.Creative, 10 per cent.Quarter three
Paid social, 18 per cent.Paid search, 32 per cent.SEO and content, 22 per cent.Lifecycle email, 18 per cent.Creative, 10 per cent.
| Channel | Q1 | Q2 | Q3 |
|---|---|---|---|
| Paid social | 46% | 34% | 18% |
| Paid search | 22% | 28% | 32% |
| SEO and content | 12% | 16% | 22% |
| Lifecycle email | 8% | 12% | 18% |
| Creative | 12% | 10% | 10% |
What we run
Table stakes. The list matters far less than the decisions made about it, which is the next section.
Paid search
Demand that already exists, captured before a competitor does.
Paid social
Demand that does not exist yet, built cheaply and read honestly.
SEO and content
The pages that answer what your buyers type at the point they start looking.
Lifecycle email
The people who already gave you an address, worked properly rather than blasted.
Creative
New angles at the rate the platform burns them, which is faster than anyone budgets for.
Measurement
The tracking, the naming and the reporting that make the other five arguable.
How a spend decision gets made
Four steps that close on themselves. The last one is the first one, which is why nothing sits in the account for a quarter because it was somebody’s idea in January.
- 01
Brief
One page: the audience, the objection we are answering, the claim we are testing and the number that would make it a win.
- 02
Test
It runs at a budget large enough to reach a readable result and small enough that being wrong is cheap.
- 03
Read
We read it against the brief's number, not against the number it happened to be good at.
- 04
Reallocate
Winners take budget from losers inside the same cycle. Nothing is left running because it was somebody's idea.
- How long a test runs
- Fourteen days, or until the test has enough conversions to be read, whichever is later.
- What kills it early
- A test is stopped early if it is spending at more than twice the account's cost per qualified meeting with no trend toward it by day seven.
- How often budget moves
- Budget moves once a fortnight. Moving it more often means reading noise; moving it less means paying for a losing quarter.
What we killed, what it cost, and what took its place
Every account has these. An agency that cannot name three is either new or not reading its own reports.
- 01
A short-form video series on a platform the buyer does not use for work. It reached a large audience for very little money and produced nothing downstream.
- What it cost
- About a tenth of the quarter's budget
- Stopped after
- Nineteen days
- What replaced it
- The same creative budget went to search terms that already converted, at roughly four times the cost per click and a fraction of the cost per meeting.
- 02
A gated benchmark report. It collected a large number of addresses, and sales could not get any of them onto a call.
- What it cost
- Six weeks of one writer and a design pass
- Stopped after
- Six weeks, at the first pipeline review
- What replaced it
- The same research went out ungated as three pages that now rank, and the email capture moved to a live tool people use more than once.
- 03
Broad brand keywords in a market where the client had no brand yet. Impressions looked excellent in the platform and did not exist in the CRM.
- What it cost
- About a fifth of one month's search budget
- Stopped after
- Eleven days
- What replaced it
- Competitor comparison terms, which cost more per click and produced meetings in the first week.
The one number
Cost per qualified meeting
Not clicks, not leads, not marketing qualified anything. A meeting counts when it passes the same four gates lead generation applies, and it is attributed to a campaign only if the campaign touched the account before the first reply.
The model we use
We report first-touch and last-touch side by side rather than picking the one that flatters the quarter, and we say which one moved.
Where it is blind
Both models are blind to the same things: someone who saw an ad on a phone and searched your name on a laptop a week later, anything that happened in a group chat, and any deal your founder closed at a conference. We name those gaps in the report instead of quietly assigning them to paid search.
What you see, and when
- Every week
- One page: spend by channel, qualified meetings, cost per qualified meeting, what changed and what we are changing next.
- Every month
- An hour with the people running it. Tests read, budget moved, and the next brief agreed in the meeting rather than sent afterwards.
- Any time
- The dashboard is yours, in your own analytics account, and it does not stop working when we do.
Where this sits in the loop
Campaigns are aimed with sentences we did not write. The objections that go into a brief come from the queue that omnichannel customer support answers all day, which is why the ads sound like your customers rather than like an agency. And the work is judged on meetings rather than clicks, using the same four gates lead generation applies before a lead reaches your reps. A campaign that wins on traffic and loses on meetings is a campaign we stop.
Book a strategy call
Forty minutes with the people who would run your account. We look at where conversations are dropping today and tell you what we would do first.
No deck. If we are not the right shape for the problem, we say so on the call and point you somewhere better.
Prefer email? admin@uniquemarketing.net