SEO versus paid search when the budget is small

A founder with fifteen hundred a month to spend on getting found asks the obvious question: agency retainer for search rankings, or ads. The advice they get back is almost always the same shape. Ads are fast but you stop and it stops. SEO is slow but it compounds. Both sentences are true, and together they are close to useless, because they describe the channels as if the only variable were time.

The variable that matters at this budget is not time. It is what you know. SEO at fifteen hundred a month is a bet placed on a demand curve nobody in the building has measured. Paid search at fifteen hundred a month is a purchase — and the thing you are buying is not the clicks. It is a list, arriving in about two weeks, of which words people type before they spend money with a company like yours. That list is worth more than the traffic it comes with, and it is the input SEO needs before it is worth starting.

Fifteen hundred a month buys information, not customers

Run the arithmetic before the strategy. If clicks in your category cost around four dollars, fifteen hundred dollars buys roughly 375 of them in a month. Spread across thirty keywords in a tidy campaign structure, that is twelve clicks per term — a number from which nothing can be concluded about anything. Concentrated on five terms, it is seventy-five clicks each, and seventy-five clicks is enough to see whether a query produces enquiries at a rate anywhere near what your landing page manages elsewhere.

So the first discipline is subtraction. A small test budget is only a test if it is pointed at few enough things to produce a readable result on each.

What two months at this level will genuinely tell you: whether the query exists in volume at all, what an impression costs in your market, which of your assumed search terms are actually commercial and which are people doing homework, and — the underrated one — what your prospects call the thing you sell. That last output regularly contradicts the internal vocabulary. Companies bid on the term they use in the deck and discover the money is on the term customers use on the phone.

What it will not tell you: whether a 3.1% conversion rate is meaningfully better than 2.4%. At a few dozen conversions, those are the same number wearing different clothes. Treat the early data as a filter that separates zero from not-zero, not as an optimisation surface. Small budgets earn the right to make big decisions, never fine ones.

Some terms are permanently cheaper to buy than to rank for

The compounding story assumes that every keyword eventually pays for its own page. Several categories never do.

Long-tail queries with high intent and low volume are the clearest case. A term searched twenty times a month, where fifteen of those searches are people ready to buy, is a good term — and it will never justify the page, the internal links and the sustained patience needed to take and hold the top organic slot for it. Buy those. Buy them indefinitely. The arithmetic does not change with scale.

The second category is any result page already owned by aggregators, directories and review sites. If the first five organic results are all lists of vendors, you are not competing for a ranking, you are competing to be included in someone else's ranking, which is a partnerships problem rather than a content problem. Paid is the only route to the top of that page that you control.

The third is anything seasonal or occasional. A term that matters for six weeks a year cannot be defended organically at a sane cost, because the effort profile is continuous and the demand profile is a spike.

None of this means SEO is dead. It means the terms worth building for are the ones with enough repeat volume to amortise the work — which, again, is something paid search can tell you before you commission a single article. The related question of whether the content that ranks actually returns the money it costs is worth answering separately before committing to a content plan.

Branded search is a tax, and it is usually worth paying

Almost every small advertiser eventually notices that a big share of their spend goes on their own company name — clicks from people who already knew who they were. It feels like paying a toll to reach your own front door, and roughly speaking, it is.

Pay it anyway, under two conditions. If a competitor is bidding on your name, the ad slot above your organic listing belongs to whoever wants it most, and letting them have it means handing over people who typed your name specifically. And if your own result has been pushed below a stack of ads and other features, the organic listing you are relying on is not where you think it is.

The honest way to settle this internally is not an argument. Pause the branded campaign for two weeks and watch total branded sessions and enquiries, not just the paid line. If the total barely moves, the ads were buying traffic you already had, and the money should move to non-branded terms. If it drops noticeably, you have found out what the tax is actually for. It is a cheap test and an uncomfortable one, which is why the argument usually gets had instead of the fortnight getting spent.

The break-even sum, with the numbers you are missing

The decision is arithmetic, and the arithmetic needs four inputs most companies can produce in an hour: cost per click, the share of clicks that become enquiries, the share of enquiries that close, and the gross margin on the resulting contract. Chain them.

Input	Consultancy	Low-ticket service
Cost per click	$6.20	$6.20
Landing page enquiry rate	4%	2%
Clicks per enquiry	25	50
Cost per enquiry	$155	$310
Enquiry-to-customer rate	20%	10%
Cost per customer	$775	$3,100
Gross margin per contract	$1,800	$900
Verdict	Works, with room	Never works

The figures above are placeholders for your own; the shape is the point. Two multiplications separate a channel that funds itself from one that cannot be fixed by better ad copy. When cost per customer lands within a factor of two of gross margin, paid search is a real channel and worth building on. When it lands at three times margin, no amount of bid management closes that gap — the problem is the price of the product or the length of the sales cycle, and both are outside the ad account.

This same sum is what makes SEO defensible or not, incidentally. Ranking is not free traffic; it is traffic paid for up front in work. If the customer economics fail at $6.20 a click, they fail at the equivalent cost of producing and maintaining the pages that would have replaced those clicks. The channel was never the problem.

Then, and only then, the SEO plan writes itself

Assume the test comes back positive on four or five terms. Now SEO becomes a specific, scoped job rather than a subscription: build the best page on the internet for each proven query, in descending order of proven value. Not a blog calendar. Not thirty posts about industry trends. Four or five pages that deserve to win, aimed at words you have watched convert with your own money.

Keep the paid campaigns running on those same terms while the organic work matures, because it takes longer than anybody's patience. Plan on the honest timeline rather than the pitch deck one — the lag between publishing and seeing anything in the numbers is the single most common reason a programme gets cancelled one quarter before it would have worked.

Sometimes the correct answer is that nobody is searching

The scenario nobody plans for: two weeks in, the campaigns are approved, the bids are competitive, and the impressions are in the low hundreds. Not expensive clicks. Barely any clicks, because barely anyone is typing the words.

This is the most valuable outcome the test can produce, and it costs a fraction of the budget to reach. It means search is a harvesting channel for demand that, in your category, does not yet exist in that form — common for genuinely new products, for categories bought through referral, and for anything sold to a market small enough that the buyers all know each other. Spending a year on SEO for those queries produces a beautifully optimised page that ranks first for nothing. The right move is to stop and go and look at which acquisition channels hold up when you actually measure them, because the answer is somewhere else entirely — outbound, partnerships, a community, a stage.

What we would do with the fifteen hundred

Two months of tightly focused paid search on no more than five terms, treated as a research line in the budget rather than a growth line, with the break-even sum run before the first ad goes live so you know in advance which result would count as good. Then, if the numbers clear, keep the winning terms running and put the remaining budget into a small number of pages built for the queries that proved themselves. If the numbers do not clear, you have spent two months of budget to avoid spending a year and a half of it on a ranking that would not have paid for itself, which is the cheapest useful thing search marketing does all year.