Translated from the Japanese original on mureo.jp.
Look for outsourced ad management at a monthly ad spend of ¥100,000 to ¥200,000 and you can be turned away at the enquiry stage. Even when you are not, the fee in the quote comes back at close to half the ad spend, which is a surprise. This article sets out what is happening at that size, and which options realistically remain.
The overall picture of costs and pricing models, and what each type of provider suits, is in Outsourcing search ad management: costs, and how to pick a provider. This article covers only what is specific to small budgets.
What counts as a small budget here
From around the point where monthly ad spend drops below ¥200,000, the response from providers changes. The dividing line sits there because under the common terms of a 20% rate and a ¥50,000 minimum fee, the effective rate starts to jump at about ¥250,000 of ad spend. Below that, what you pay stops being proportional to what you spend on ads.
Three bands are used as reference points below: ¥100,000, ¥200,000 and ¥300,000 a month. At ¥50,000 a month or less, outsourced ad management barely holds up as an arrangement at all, so running it yourself, or paying once for the initial build alone, is the realistic shape.
Being turned away is not aggressive pricing
Break down the provider’s own margin and the reason is plain. At a 20% rate, a ¥100,000 monthly ad budget brings in a ¥20,000 fee. The work on their side, though — reviewing keywords, adding negatives, rotating ad copy, the monthly report — does not scale much with the size of the spend. A few hours a month and the ¥20,000 has gone on payroll.
The minimum fee is a floor placed to close that gap. The same reason is why many providers set a minimum engagement at ¥200,000 to ¥300,000 a month: below that, the more of the work they take on the closer they get to a loss. The effective rate looking like it leaps on a small quote is not someone taking advantage of you; it is that the amount of work does not rise and fall with the budget.
That also changes how you should look at providers. Comparing quoted rates alone means nothing. What to compare is the amount in yen, and what that amount buys you.
Fix the total budget, and see how much reaches delivery
On a small budget, rather than thinking of ad spend and fee separately, it is easier to decide the total you can put in first and then look at how much of it actually reaches delivery. Here are a rate-based model at 20% with a ¥50,000 minimum, a flat model at ¥40,000 a month, and running it yourself, side by side.
| Total you can put in per month | Rate-based (20%, ¥50k min) | Flat (¥40k/mo) | Run it yourself |
|---|---|---|---|
| ¥150,000 | ¥100,000 to delivery | ¥110,000 to delivery | ¥150,000 to delivery |
| ¥200,000 | ¥150,000 to delivery | ¥160,000 to delivery | ¥200,000 to delivery |
| ¥300,000 | ¥250,000 to delivery | ¥260,000 to delivery | ¥300,000 to delivery |
At a total of ¥150,000, going rate-based leaves ¥100,000 reaching delivery. Run it yourself and it is ¥150,000, a difference of 1.5 times. That difference shows up directly in results. Producing a 1.5x efficiency gap through skill in running the account is not easy.
The table alone does not settle it in favour of running it yourself, though, because it leaves out the time the work takes. The material for that judgment is below, in Where the line falls if you run it yourself.
The constraint that bites before cost is conversion volume
The real reason small budgets stumble sits ahead of the fee. When conversions are few, the platform’s automated bidding cannot finish learning.
Search ads today assume automated bidding, and if you use a target cost per conversion or a target return on ad spend, the machine needs a volume of conversions before it can leave the learning phase. The guideline varies by platform, but around 30 conversions in the last 30 days is one line. Fall short of that and delivery keeps running with bidding accuracy never improving.
Working backwards from your product’s cost per conversion, the ad spend you need comes out like this.
| Assumed cost per conversion | Ad spend needed for 30 a month |
|---|---|
| ¥3,000 | ¥90,000 |
| ¥5,000 | ¥150,000 |
| ¥10,000 | ¥300,000 |
| ¥20,000 | ¥600,000 |
| ¥30,000 | ¥900,000 |
If your product lands between ¥3,000 and ¥5,000 per conversion, learning holds up even at ¥100,000 to ¥150,000 of monthly ad spend. Above ¥10,000 per conversion, the arithmetic says you need ¥300,000 a month. Whether outsourced ad management works on a small budget is settled less by the size of the budget than by the product’s cost per conversion.
When the cost per conversion is high and the budget will not stretch, there are two moves. One is to set an earlier action as the conversion so the volume rises. Switching the measurement point to a brochure request or a quote request does raise volume, with the side effect that low-quality volume also becomes something the optimisation chases. The other is to skip automated bidding and run it manually for a while. While volume is short, not handing it to the machine behaves more straightforwardly.
The constraint does not disappear by handing the work outside. How far the person in charge can move narrows in the same way, so when you take a quote it is worth asking whether automated bidding’s learning will run at this budget. A provider who cannot answer is safer avoided.
Five kinds of provider available at this size
Flat-rate agencies. The monthly amount is fixed regardless of how much you spend, so the effective rate does not jump. Because the cost does not rise in proportion when you raise ad spend, it works in your favour if you intend to scale. Scope tends to be narrower, so verify individually whether ad copy production and reporting are included.
Freelancers. Some take the work at ¥20,000 to ¥30,000 a month, which keeps the cost lowest. Going directly to someone with long hands-on experience is another advantage. In exchange, you place the risk of operations stopping, through illness or the state of their other accounts, with a single person. Settle before signing what remains if you can no longer reach them.
Paying once for the initial build alone. Have the account structure, keywords, ad copy and conversion tracking built in one go for somewhere around ¥100,000 to ¥200,000, then run it yourself from there. No monthly fee arises, so it holds up even at ¥50,000 of monthly ad spend. The person who built it leaves, so be ready for not being able to ask later why the design is what it is.
Spot consulting. You run the account yourself and buy only one or two consultations a month. Around ¥10,000 to ¥30,000 an hour. For people who can do the hands-on work themselves, this can come out the best value of the lot.
AI-run ad management. Because there is no payroll structure underneath, it is often offered flat, and priority does not shift with the size of the ad spend. The smaller the budget, the wider the gap against a rate-based model. There are cautions on how to delegate, covered below.
Where the line falls if you run it yourself
As the table above shows, at a total of ¥150,000 running it yourself puts 1.5 times as much into delivery. Whether outsourcing is still worth it turns on how you value the working time.
Running search ads yourself takes ten to twenty hours a month on reviewing keywords, adding negatives, rotating ad copy and checking the numbers. The first month or two goes beyond that on design and setup. A ¥50,000 fee divided by 15 hours a month is ¥3,300 an hour. If your own time is worth more than that, outsourcing is the rational move. If it is worth less, running it yourself and adding what you save to the ad budget grows faster.
There is a second question beside the time, which is whether you can keep it up. Ad ops is not a job you sit down to once a month; it is a job of touching things a little at a time. Estimate honestly whether it will survive three months alongside your main work. The material for that judgment, and what happens to the account and to learning when you switch from a provider to in-house, is in Bringing ad ops in-house.
Four ways to narrow, to make a small budget work
When the budget is small, the wider you spread it the thinner it gets. Narrowing these four makes results easier to reach on the same amount.
- Narrow to one platform. Rather than splitting evenly between Google and Yahoo!, leaning on one gets you to the volume learning needs sooner. For search ads, starting with Google alone is the straightforward move.
- Narrow the keywords. Putting in every word that might be related scatters the budget thin. Starting with only the words close to a purchase or an enquiry, ten to thirty of them, and adding as you watch the search terms, is faster.
- Narrow the geography and the hours. If your trading area is fixed, narrow the delivery locations and cut the hours when you cannot take an enquiry. The smaller the budget, the more directly the share of wasted spend shows up in results.
- Settle on a single definition of a conversion. Measuring and optimising for several points at once splits the volume and learning does not progress. Decide on one first, and add more once volume has built up.
When to raise the budget
Once you have narrowed things down and the campaign is running, the next thing you hesitate over is when to move the number. The smaller the budget you started with, the more likely you are to move the figure without holding a reason to raise it, and at that point the increase becomes a matter of feel. There are three things you can judge from.
The first is whether you are failing to show up because the budget is short. Google Ads reports lost impression share (budget), which shows the proportion of times delivery stopped at your daily cap while the demand was there. If a figure appears there, raising the budget recovers what you are leaving behind. If the value stays near zero and you raise anyway, there is nowhere for the spend to go, so it goes unused.
The second is whether the cost per conversion you are getting is in line with your target. If it sits inside the target, an increase gives you reason to expect volume to grow in proportion. Raising it while you are above target only increases the volume of transactions that do not add up. Moving the number before the cost per conversion is inside your target scales the loss along with it.
The third is a check on the low end. If the daily budget is only a few times the cost per click, you get a handful of clicks a day and never accumulate anything to judge from. At ¥800 a day against a ¥300 cost per click, that is two clicks a day and sixty in a month. At that size you cannot tell good ad copy from bad, so narrowing the placement to bring the unit cost down comes before any increase. To check whether volume is reaching the automated bidding learning phase, use the figures given earlier under the conversion volume constraint.
There is an order to how you raise it as well. With automated bidding, moving the budget a long way at once sends the campaign back into the learning phase and the numbers are unsettled for several days. Raising it 20 to 30 percent at a time, with a few days in between, leaves you something you can compare before and after. Adding a platform or widening keywords at the same time as the increase makes it impossible to tell which one did the work, so change one thing at a time.
What to verify in the quote
When you take a quote on a small budget, asking these five reduces later regret.
- The minimum engagement size and the minimum fee. Have them also put in yen how the fee changes when you raise ad spend.
- What the fee covers. Report production, ad copy creation and advice on the landing page are sometimes billed separately.
- How many accounts the person in charge carries. Small accounts tend to drop in priority. Ask whether the arrangement is one that does not scale with budget.
- Contract lock-in. Three to six months is common. A one-year lock-in is too long at this size.
- Whose name the ad account is in. Create it in your own company’s name and the operational data stays with you when you cancel. This is not one to give ground on.
A further breakdown of what the fee is made of is in Are ad agency management fees too high?, and where to look in the report once you have delegated is in Reading your agency’s ad report.
Cautions when delegating to AI
AI-run ad management is an arrangement that suits small budgets. There is no payroll structure underneath, so priority does not shift with the size of the ad spend, and it is often offered flat. Delegate badly, though, and a different cost appears: hand it over without the criteria for judgment and changes based only on the surface of the numbers get executed quickly. Three lines are worth verifying: that changes touching budgets or bids can require approval, that ceilings it cannot cross can be set as numbers, and that a record of the changes made remains so you can revert. This is explored further in Is it safe to hand ad ops to Claude Code?.
The mureo tool we build is this AI-run ad management as a product. Pricing is flat rather than a rate on ad spend, so the effective rate does not jump even at ¥100,000 of monthly ad spend. AI runs operations daily across platforms from Google Ads to SmartNews, and shows you on screen what was spent where and why each judgment was made. Changes to budgets or bids go through human approval, and the ceilings you set cannot be crossed from the AI side.
Frequently asked questions
Are there agencies that will take a ¥100,000 monthly ad budget?
Providers that advertise flat pricing, and freelancers, will take it. Providers centred on rate-based pricing will often turn you away on the minimum engagement size. Since the reason for the refusal sits in the structure of their margin, negotiating it down is not realistic. Looking for a different type of provider is faster.
Can the minimum fee be negotiated?
It can come down, but what comes down is reflected in the amount of work. Rather than negotiating the amount alone, a conversation about cutting the work included in order to cut the amount lands more easily. Making the monthly report a simpler form, or producing the ad copy yourself, for example.
Which should I raise first, the ad spend or the fee?
While you are short of the volume automated bidding’s learning needs, ad spend comes first. Until volume is there, skill in running the account barely shows in results, so paying more in fees is hard to recover. Once volume has built up there is more room to work with, and the judgment changes.
Is it fine to run it myself while the budget is small and outsource once it grows?
That order is no problem. Having run it yourself once actually leaves you able to judge whether the reports are any good after you delegate. At the switch, mind only two things: create the ad account in your own company’s name, and keep the negative keyword list and the history of changes so far.
Does performance-based pricing remove the risk on a small budget?
It depends on how results are defined. Contracting per enquiry pushes effort toward volume regardless of quality. On a small budget the denominator is small, so low-quality volume mixed in hits harder. Confirm in the contract wording what counts as a result. Few providers offer it in the first place, which is the other reality.
Summary
Below ¥200,000 of monthly ad spend you can be turned away from outsourced ad management, but the reason sits in the provider’s margin, not in their pricing. Compare with the total fixed and the amount reaching delivery falls by exactly the fee a rate-based model takes, so weighing that difference against the price of your working time is the axis of the decision. What bites before cost is conversion volume, and whether you reach automated bidding’s learning is settled almost entirely by the product’s cost per conversion. The providers available are five: flat-rate agencies, freelancers, a one-off initial build, spot consulting, and AI ad ops. Whichever you pick, narrowing four things — platform, keywords, geography and the measurement point — is the precondition for making a small budget work. Once it is running, treat an increase as something to make only when two things line up: the budget is visibly costing you impressions, and the cost per conversion is inside your target. The overall picture of costs and how to pick a provider is in Outsourcing search ad management, and the cautions specific to Google Ads are in Outsourcing Google Ads management.