PPC budget split: what we would do at three spend levels
Nobody can tell you the perfect percentage. What we can do is show a cautious one at three monthly ad budgets, and say plainly what each would buy.
What we would do with three monthly budgets
Examples, not rules. Percentages are hypothetical starting points to be adjusted by your data.
$1,000 a month in ads
Do not buy links yet. Spend the month fixing the landing page, tracking and negative keywords. If the account converts, plan one test of $100 to $150 next quarter.
$5,000 a month in ads
Reserve about 10 percent for a single quarter: $1,500 total across 10 to 15 placements at $100 to $150, aimed at the page that sells best. Ads stay at $5,000.
$20,000 a month in ads
A first test of $4,000 to $6,000 across several evergreen pages is proportionate. Split into three monthly orders so you can watch each batch before the next.
Why we cap the test at one quarter
Links reveal themselves slowly. A placement that goes live in about 24 hours can still need weeks before Google recrawls the page and months before the picture in Search Console is readable.
So the cap is a time box as much as a money box. Set the amount, pick a review date 90 days out, and write down in advance what result would make you spend more. That sentence prevents the usual drift.
Put a stop rule in writing too. If after 90 days the linked pages show no new impressions in Search Console and no referral visits worth the name, you spend nothing more and you have lost a fixed, known amount. If they do, raise the cap by half, not double. Small steps keep the test honest.
Stop at the cap. Buying more because week three looks quiet is the most common way tests turn into guesswork.
A sample quarter at the $5,000 level
| Month | Ads | Links | What to look at |
|---|---|---|---|
| 1 | $5,000 | $600: 5 placements near $120 | Referral visits, indexation of the linking pages |
| 2 | $5,000 | $600: 5 more, different domains | Impressions on the target page in Search Console |
| 3 | $5,000 | $300: top up only the best niche | Clicks and conversions from organic landing sessions |
| Review | Hold | Decide with data | Cost per visit and cost per lead, ads against links |
Two ways a split goes wrong
The first is borrowing from a winner. A manager shaves 10 percent off a hot campaign to fund links, and then cannot tell whether a sales dip came from the cut or the season. Fund the test from a separate line.
The second is spending all of it on one famous domain. A single $500 placement tells you about one page. Five placements near $100 on relevant sites tell you about a pattern. Variety is how you learn.
Remember the comparison from the cost-per-visit page: you are buying data about a channel before you buy more of it.
Never fund links by cutting the ad line that is currently your best converter. Take the test money from growth budget, not from what pays the bills.
Settings to confirm before you spend the link slice
- The target page is indexable and returns a 200 status.
- Anchor text will be the brand, the URL or a natural phrase, not an exact-match ad keyword.
- Each order is tagged in your sheet with date, domain and price.
- Google Analytics has a referral filter ready for the new domains.
- You know which Search Console queries you rank for today.