- Your affordable cost per enquiry comes from average sale value, margin and close rate, and every later argument resolves against it.
- Two campaigns beat eight at small budgets, because fragmentation starves every slice of data.
- Negative keywords are where the savings are: one account went from $96 to $58 per enquiry in six weeks on negatives alone.
- Sending paid traffic to a home page costs roughly half the conversions in accounts where we have measured both.
- Call tracking typically doubles the recorded conversion count for service businesses, which changes what looks profitable.
Almost all published PPC advice assumes a budget large enough for the platform's automation to learn from. Under about 30 conversions a month that assumption breaks, and following the advice actively wastes money.
This is what we do instead for clients spending in the hundreds rather than the tens of thousands.
Work out your ceiling first
Take your average sale value, your gross margin, and the proportion of enquiries that become customers. Multiply them. That is the most you can pay for an enquiry and still make money.
A business with a $1,800 average job, 45% margin and a 30% close rate can afford roughly $240 per enquiry. One with a $90 product and a 25% margin can afford about $22 per order. Those two businesses should not be reading the same advice, and the second probably should not be in paid search at all until the average order value rises.
Write the number down before the account is built. Every argument later, about match types, about bidding, about whether the month was good, resolves against it.
Structure: two campaigns, not eight
At small budgets the enemy is fragmentation. Every campaign, ad group and keyword divides the same money into thinner slices, and thin slices never gather enough data to be optimised.
Our default starting structure is one search campaign for the highest-intent terms, the ones naming the service and the intent to buy, plus one brand campaign if competitors are bidding on your name. That is it. Add a second service campaign when the first is consistently spending its budget and converting.
Inside the campaign, keep ad groups tight: three to eight closely related keywords, three responsive search ads, and a landing page that matches.
Match types: start narrow, widen deliberately
Start on exact and phrase. Broad match works well at scale because the system has enough conversion data to steer it. At $800 a month it will spend the budget on tangential searches while it learns.
Once you have a couple of months of conversions, test broad match in its own campaign with a separate budget, so you can turn it off cleanly if it underperforms.
Negative keywords are the real work
For the first eight weeks, read the search terms report every week and add negatives. It is unglamorous and it is where most of the savings are.
Build three lists you reuse across accounts: informational modifiers (how to, diy, tutorial, free, salary, jobs), competitor names you do not want to appear against, and irrelevant industries that share your vocabulary. On a recent account, negatives alone cut cost per enquiry from $96 to $58 in six weeks with nothing else changed.
The pattern to watch for is a single expensive term quietly taking a third of the budget. It is usually a plural, a job-seeker query, or the name of a product you stopped selling.
Landing pages decide the outcome
The ad's job is to get the click. The page's job is everything after that, and it is the part small accounts neglect.
Rules we hold to. The headline repeats the promise in the ad, almost word for word. The proof sits above the fold: a real client name, a number, or a photograph of actual work. The form asks for the fewest fields that let you follow up, which for most service businesses is name, contact and one sentence about the job. And the phone number is a link, prominently, because a large share of small-business enquiries still arrive as calls.
Sending paid traffic to a home page is the most common mistake we correct. In the accounts where we have measured both, it costs roughly half the conversions. The page-side work is covered in CRO at low traffic.
We build and run small paid search accounts on a flat monthly fee, no percentage of spend, so there is no incentive to push your budget up.
Bidding at low volume
Smart bidding needs conversion data. Below roughly 30 conversions a month the strategies oscillate. Start with maximise clicks and a manual CPC cap, gather conversions for six to eight weeks, then move to maximise conversions once the volume supports it.
Set the cap from your ceiling. If you can pay $240 per enquiry and you convert 8% of clicks, your break-even click is about $19. Bid below that with room for error.
Track calls, or under-report by half
Most service businesses receive more calls than form submissions from paid search. If you count only forms, every decision you make will be wrong in the same direction.
Use call tracking on the ad destination pages, count calls over a sensible duration as conversions, and reconcile monthly against the actual enquiry log. In the accounts we run, adding call tracking typically doubles the recorded conversion count, which changes what looks profitable.
A realistic first ninety days
Weeks one and two: launch, watch daily, add negatives. Weeks three to six: pause keywords with spend and no conversions, tighten ad copy, fix the landing page's worst step. Weeks seven to twelve: adjust the bid strategy, test one new ad group, and produce the first report comparing cost per enquiry against your ceiling.
Do not judge the account before week six. Do not restructure it before week twelve unless something is obviously broken. Restructuring at week three, which is when the pressure usually arrives, resets the learning and costs you the month.
What we do not run at this budget
Display, video and discovery prospecting. Performance Max as a first campaign: it works, but it hides where the money went, which is exactly what a small account cannot afford. Dynamic search ads without a tight page feed. And anything depending on remarketing lists you do not have the traffic to fill.
When to hand it over
Run it yourself if you can give it an hour a week for the first two months and you enjoy spreadsheets. Hand it over when the search terms report stops surprising you, when you are ready to add a second service, or when the account has been sitting on automated bidding with nobody reading it for a quarter. We manage small accounts on a flat monthly fee rather than a percentage of spend, which is set out on the PPC management page, and paid search works better when it is planned alongside the other channels: full-funnel sequencing covers that.
Questions
Is $800 a month enough for Google Ads?
It is enough for a narrow, high-intent campaign in most service categories. It is not enough in categories where clicks cost $15 and up, such as legal, insurance and some B2B software. Check the click cost against your ceiling before committing.
Should I use Performance Max?
Not as your first campaign. It performs well once you have conversion history and a good feed, but it obscures where spend went, which small accounts cannot afford. Start with search, add PMax later if the volume justifies it.
How many keywords should a small account have?
Fewer than you think, often between 15 and 40 across the whole account. Every additional keyword divides the same budget, and a keyword that gets four clicks a month tells you nothing.
How long before Google Ads becomes profitable?
Six to twelve weeks in most accounts we run: the first two spent buying data, the next four cutting waste, the rest improving conversion rate. If it is not close by week twelve, the problem is usually the offer or the page, not the account.
Should I bid on my own brand name?
Only if competitors are bidding on it, or if your organic result does not occupy the top of the page. Otherwise you are paying for clicks you would have received for nothing.