Why the refusal loses and the quote wins
The request arrives in a Monday meeting with a shape you'll recognise. Why don't we come up for "digital marketing"? Or "software". Or whichever two-word term the founder typed into their phone over the weekend and didn't find themselves in.
The instinct is to explain search intent. Don't. You'll be right and you'll lose. A lecture about intent lands as a reason you can't do the thing, and three vendors have already said they can.
There's a rule buried in this worth stating flatly: whoever says no owns the risk and gets none of the credit. Nobody thanks you for money saved on a term nobody pursued. If the company later loses ground to a competitor who did pursue it, the meeting where you said no is the one that gets remembered.
So don't say no. Say what it costs. You're not the person who decides which keywords are worth having — you're the only person in the room who knows the price. Quote it the way a builder quotes an extension: scope, timeline, and what you'd stop doing to make room.
That single move turns a fight about your judgement into a comparison between two numbers. You are considerably better at the second argument.
Work out which request you've actually received
Before pricing anything, find out which of four things has been asked for, because three of them are cheaper to satisfy than the one that landed on your desk.
- I want to see us where I look. They searched something and we weren't there. Often the real fix is what shows up when someone searches the brand, not the category term.
- A competitor appeared and it stung. A competitive-response request wearing a keyword costume. Usually satisfied faster by taking the three comparison queries where that competitor is already beating you.
- Somebody asked in a board meeting. The requirement is an answer to give, not a ranking to hold. A one-page note on what the term costs and what you're doing instead is the deliverable.
- We genuinely need the category term. Rare, real, and dealt with at the end of this page.
Pricing a head term, with the method
Price it like work, not like a probability. Five steps, no forecasting software, and every input is something you can show on a screen.
- Read the SERP, not the difficulty score. Write down what the top five results actually are — a category leader's hub page, a marketplace, Wikipedia, a directory. That list is Google telling you which format it accepts here, and it's frequently one you don't have.
- Count the gap in pages, not in links. For each of the top three, count the supporting pages they hold on that topic. Then count yours. The difference is the cluster you'd have to build before you're even considered.
- Convert the gap into months at your real publishing rate. Not the aspirational rate from the plan. Four pages a month against a thirty-page gap is seven to eight months of the whole content line before you're eligible — not before you win.
- Price the off-page half at market. The incumbents didn't arrive on content alone. Use what links actually cost in India rather than a placeholder, and count the outreach months as well as the rupees.
- Add what stops. The line people leave off, and the one that decides the meeting. Name the pages and fixes that won't happen this quarter because the production line is pointed here instead.
What the quote says out loud, and what it never contains
The output is a range with a start date and a checkpoint — nine to fourteen months, this much production, this much outreach, first review at month four — and never a position. Say that part out loud, because it protects you. Google's own guidance for buyers states that no one can guarantee a #1 ranking on Google. If a vendor promises the position next week, you've already told your boss what that promise is worth.
One caution about the inputs everyone quotes. Average monthly search volume is rounded and averaged over twelve months, and the competition column in Keyword Planner counts advertisers bidding, not organic difficulty. Quote the volume as a band.
The twelve-query alternative, and how to make the difference visible
A counter-proposal has to be concrete or it reads as deflection. Build the list in the meeting, on the screen, not afterwards in a document nobody opens.
Twelve queries that name a buying moment rather than a category. The shapes are predictable enough to draft live: X price, X cost in India, X vs a named competitor, best X for a segment, how to choose an X, alternatives to the incumbent. Every one is somebody with a wallet already out.
| Line | The head term | Twelve buying-intent queries |
|---|---|---|
| Pages required | A hub page plus a cluster big enough to close the gap | Twelve to fifteen pages, each finishable in days |
| Time to first movement | Months, and the early ones look like nothing at all | Weeks, because the competition on each is thinner |
| Off-page work | Sustained outreach; incumbents have years of it banked | Little to none for most of the twelve |
| What a win looks like | One position, watched daily by the whole company | Twelve small streams of people already comparing |
| What failure costs | A quarter of the content line, plus your next request | One or two pages that didn't land, replaced next month |
| Who the traffic is | Students, jobseekers, competitors, and some buyers | Buyers, mostly, at the point of choosing |
Buy the answer instead of arguing for it
Run two weeks of paid search: one ad group on the head term, one on three of the twelve. Cap the spend at a number nobody has to approve twice. Report cost per qualified lead, not cost per click.
That test ends the argument in a way no deck does, and it can end it against you — which is why it works. If the head term produces qualified leads at a sane cost, the request was right and you can fund it properly. If it produces four hundred clicks and a job application, the conversation closes without anyone being told they were wrong.
Paid search isn't a perfect proxy — the intent mix behind an ad click differs, and the ad sits above results your organic page would sit below. It's an indicator, not a proof, and the longer comparison between the two channels is worth having for the follow-up question.
Convert both columns into the number leadership already argues in
Sessions lose this argument. Leads and gross profit win it.
Run both columns through the model finance already uses: leads, close rate, gross profit per deal, and the month the cumulative lines cross. If you've built that page once to get the budget approved, you're reusing it rather than starting again.
Two numbers always get asked for. What the head term is worth if you win it — volume band, a click-share range for positions you could realistically hold, your own conversion rate rather than a benchmark. A forecast built that way is defensible and deliberately unexciting. Then the same arithmetic for the twelve, so the comparison is like for like.
Know the trap in your own reporting too. Average position in Search Console is the topmost position your site held, averaged across every query it appeared for — so "we're at 14 for it" can be one decent query hiding forty bad ones.
When you present it, lead with the business metric and volunteer the bad news early. The bad news is that the honest forecast is a wide range with a long lead time. Saying it first makes the rest believable.
When they insist anyway
Sometimes you lose the argument. Losing it well is a skill, and it's what protects your standing in the next one.
- Take the work properly. Slow-walking an instruction you lost is the fastest way to stop being asked, and it's more obvious than you think.
- Write down the start date, the checkpoint and the stop rule before anything ships. Month four, and here is what has to be true by then.
- Define progress in a metric that can actually move. Impressions and the count of distinct queries the cluster surfaces for, not position on the head term. Position won't move for months, and everyone reads that silence as failure.
- Build the supporting cluster first. Those pages earn traffic on their own merits whether or not the head term lands, so a cancelled project still leaves something behind.
- Log the opportunity cost once, without commentary. A dated line naming what was deferred reads as evidence at the checkpoint, not as an accusation.
- Say what would change your mind. If the cluster is pulling impressions for the head term by month four, you were wrong and you'll say so. That sentence is what buys the stop rule its credibility.
The one case where the vanity term is the right call
It exists, and dismissing it out of hand is how SEOs get a reputation for being small-minded about brand. Check the obvious thing first: it may not be a vanity term at all.
Plenty of head terms are buying queries. Somebody typing "crm software" or "payroll software" is shopping, and shortness doesn't make a term uncommercial. Open the results. If Google is showing product pages, pricing pages and comparisons, the buyers are there and your boss is right. Intent is read off the results page, not off the word count.
Then the genuine strategic case. When the term names a category you intend to own, the payback window is longer than a quarter and the second-order effects are real: a page ranking for the category name earns links passively, because people cite it when they write about the category. If leadership is playing that game deliberately and can fund it for two years, it's a defensible bet — a brand investment with an SEO component, budgeted as such.
The failure mode is neither of those. It's a head term chosen because it sounds like the business, funded for one quarter, judged in month three and abandoned in month five with the cluster half-built and the twelve buying queries never started. That's worse than never beginning, and preventing it is why you quote the thing instead of refusing it.
On the agency side the move is identical and the record-keeping matters more. Put both options in writing, let the client choose, note the choice with a date. Engagements fail in month six not because the term didn't move, but because nobody can find the email where the trade was made explicit.