When not to hire a PPC agency

The arithmetic, before anything else

An agency fee comes out of the same budget as the media. So the bar is not “will they improve the account”. It is “will they improve it by more than the fee would have bought as extra spend”. That is a number, and you can work it out before you talk to anyone.

Our management starts at $2,500 a month and moves to 10–15% of spend above $25,000. Divide the fee by the spend and you get the improvement we have to deliver just to break even.

What we would have to improve, at our own published rate

Your monthly ad spend	Our fee	Improvement needed to break even
$3,000 / mo	$2,500	83%
$5,000 / mo	$2,500	50%
$10,000 / mo	$2,500	25%
$15,000 / mo	$2,500	17%
$25,000 / mo	$2,500	10%
$50,000 / mo	$5,000 – $7,500	10–15%
$100,000 / mo	$10,000 – $15,000	10–15%

Source: Computed from the rate published at /pricing. Break-even is fee ÷ spend: the efficiency gain required before the engagement returns more than the same money spent as media.

At $3,000 a month we would have to make the budget work 83% harder before you were any better off. Nobody does that reliably. At $25,000 the bar is 10%, which is an ordinary result from fixing tracking and cutting waste. Somewhere around $15,000 the purchase stops being speculative.

This is not a quirk of our pricing. Run the same division on any agency that charges a flat fee, and on the percentage models the answer is fixed by definition: a 15% fee needs a 15% improvement, forever.

01

Five situations where the answer is no

Each of these is a threshold rather than an opinion, so you can check which side of it you are on today.

Your spend is below the break-even line

Covered above. The honest version of this is not “we are too expensive for you” but “at this budget the money does more as media than as management”. Come back when the spend has grown, or buy a one-off piece of work instead of a retainer.

Your conversion volume is below what the bidding needs

Automated bidding is the whole game on every major platform now, and it learns from conversion events. Meta’s guidance is around fifty conversions a week per ad set; Google’s bid strategies want a comparable monthly figure. Below that, the system is not optimising, it is guessing, and there is no expertise that fixes a data shortage.

What helps: counting an earlier event that actually happens often enough, feeding values back so the system can tell a good conversion from a cheap one, or consolidating campaigns so the volume lands in one place. All three are a week of work, not a retainer.

Your problem is downstream of the ad

If the cohort never pays back, cheaper installs will not save it. If the paywall converts at a third of the category norm, more traffic multiplies the leak. If onboarding loses half the people you paid for, acquisition is the wrong budget line to be arguing about.

Test it before you hire anyone: put your real numbers into the payback calculator and see whether the month you turn profitable moves more when you change install cost or when you change retention. If it is the second one, an ads agency is not your purchase. Paywall optimisation and onboarding are free to read.

The work is smaller than a retainer

A single search account at modest spend, with tracking already right, is a few hours a month. Any retainer priced for a team is a bad fit for that, and a senior freelancer will do it for a fraction of the cost. Market rates for both are on what PPC management costs, and the fee calculator converts whatever anyone quotes you into a comparable percentage of spend.

Nobody can yet measure what happens

If conversions are not tracked, or are tracked wrongly, then every judgement made in the first month is made against a distorted scoreboard, including the judgement about whether the agency is working. Spend the month on the implementation and start the engagement afterwards. It is cheaper and the argument about results never happens.

02

Four where it genuinely pays

The mirror image, and the same rule applies: if none of these describes you, the purchase is probably premature.

You are adding channels, not maintaining one

Running two platforms well is a job someone in-house can hold. Running six, each with its own auction, creative format and measurement quirks, is where a team stops being a luxury. The cost of learning a new platform badly is usually larger than a year of fees.

Creative volume is the constraint

On paid social the targeting levers are mostly gone; the concepts you feed the system are the variable you still control. If your bottleneck is that you can ship four new concepts a quarter and the account needs forty, that is a staffing problem an agency actually solves.

The measurement needs rebuilding

Server-side tracking, a conversion value schema, offline conversions back from a CRM, an attribution model anyone believes. This is project work with a clear end, it pays for itself in better bidding, and it is the one thing we would argue is worth buying even at spend levels where the retainer is not.

There is a date

A launch, a new market, a funding round, a season. Deadlines are where outside capacity is worth paying a premium for, because the alternative is hiring, and hiring takes longer than the deadline.

If you are on the wrong side of this page

Here is the sixty-day version, in the order we would do it. None of it requires us.

Audit your conversion actions. For each one, write down what it represents, whether it is primary, and what your own database says the count was last month. Any disagreement over about ten percent is a problem, and any primary action that does not correspond to money is a bigger one.
Consolidate. Fewer campaigns, more volume in each, so the bidding has something to learn from.
Put values on conversions. A system told that a $20 order and a $2,000 order are the same thing will go and find you more of whichever is cheaper.
Fix the page the ad points at before you change the ad. It is the cheapest percentage in the funnel and nobody needs permission to do it.
Only then look at bid strategies, and give any change three weeks before you judge it.

Questions people actually ask

Is this page just reverse psychology?: Judge it by whether the numbers are checkable. The table is computed from our published rate, which is on the pricing page, and it says plainly that below roughly $15,000 a month of spend we are a bad purchase. An agency doing reverse psychology would not put a threshold on it, because a threshold is falsifiable.
What if I am below the threshold but the account is a mess?: Then buy the fix, not the retainer. A one-off audit, a tracking rebuild or a restructure is a fixed-price project, and you keep running the account afterwards. Paying a monthly fee for someone to maintain a small account they have already fixed is where the money leaks.
Everyone says they beat the algorithm. Why does spend matter?: Because automated bidding needs conversion volume before it can learn anything, and no agency supplies volume. Meta asks for roughly fifty conversions a week per ad set; Google wants a comparable monthly figure for its bid strategies. Under that, the system is guessing, and the skill you hired cannot be applied.
Can I hire an agency for one month to set things up?: Yes, and it is usually a better first purchase than a retainer. Ask for it as a fixed-scope project with a defined deliverable and a date. Any agency that will only sell you a rolling monthly fee is telling you something about how it makes money.
How do I tell a real improvement from a reported one?: Ask what the number was before, over what period, and under which attribution setting. A platform interface will happily show a ROAS that rose because branded search grew or because a conversion action started double-counting. The comparison that matters is revenue against total cost, media and fees together.
You are an agency. Why would you publish this?: Because the alternative is a first month spent discovering it together, and those engagements end badly for both sides. We would rather lose the enquiry now than refund it in ninety days.

Still want a second opinion?

The audit is free and it is written by hand, which means it will also tell you if the answer is that you do not need us. Most people take the document and act on it themselves. That is a fine outcome.

Request the audit See what it costs