Why search engine optimization is important: the case you take to a decision maker, with the arithmetic filled in
Not what SEO is, but why it earns a budget line. The channel comparison, a worked payback calculation using your own four numbers, the order the metrics move in, and the four situations where a different spend comes first.
Published by AI Knows Us · Updated 22 October 2026
Search engine optimization matters for one commercial reason that no other acquisition channel shares: the asset you build keeps answering the same question after you stop paying per visit. Advertising buys attention for as long as the card is charged. A page that answers a buying question is a fixed cost that keeps earning, and it accumulates rather than resets. That is the case, and it is an arithmetic case rather than a belief, so this page sets out the comparison, the calculation to run with your own figures, and the honest list of situations where the money belongs somewhere else first.
If what you need first is the definition and the mechanism, that is a different page: what SEO is, in plain English covers crawling, indexing and the four parts of the work. This one is for the conversation where somebody has to approve the spend.
Three reasons it earns a line in the budget
- The demand already exists and is written down. People type what they want. Every other channel starts by guessing who might be interested; search starts with a list of the exact words buyers used, with volumes attached. You are deciding which existing questions to answer, which is a far easier decision than inventing demand.
- The intent is sorted for you. Somebody searching a product category is at a different point from somebody searching a how to question or a supplier in their city. One channel gives you all three, separated by the words themselves, so the page can match the moment.
- The cost per visit falls over time. The page is written once. In month two it earns what it earns; in month twenty, if the subject still matters, it is earning against a cost that was paid long ago. Paid search works the opposite way round, because the auction price per click goes up with competition rather than down with time.
Against the alternatives, honestly
| What you pay for | When it starts working | What happens when you stop | |
|---|---|---|---|
| Paid search | Each click, at auction | The same day | Traffic stops that day |
| Search engine optimization | The work, once per page | Weeks to months | The pages keep answering, and decay slowly |
| Paid social | Impressions and clicks | Days | Traffic stops, and the creative ages quickly |
| Email to your own list | Almost nothing per send | Immediately, to people who already know you | Keeps working, and the list needs feeding |
| Outbound and cold contact | People's time | Immediately | Stops with the activity |
Read that table as a portfolio rather than a contest. Paid search can be read from the day it runs and SEO over a longer window, which is why most businesses that can afford both run both, and why a business that needs revenue in thirty days starts with the row that works the same day.
The calculation to put in front of the decision maker
Four numbers you already have: organic sessions a month, the share of them that convert, the value of a conversion, and what the work costs a month. The example below is an illustration with its assumptions named, and it is worth ten minutes to replace them with yours.
- Now. 400 organic sessions a month, 2 per cent of them become an enquiry that closes, an average first order of 1,200 dollars. That is 8 orders and 9,600 dollars a month of revenue from organic search.
- The change. Suppose the work takes sessions from 400 to 520, a 30 per cent rise, with the conversion rate unchanged. That is 2.4 more orders and 2,880 dollars more revenue a month.
- Against the cost. At 2,000 dollars a month of work, the month that lift lands pays for itself and the gap widens from then on, because the pages stay up while the fee stops when you stop.
- The number to write down. Revenue per organic session, here 24 dollars. It makes every later decision arithmetic: a page you expect to bring 200 sessions a month is worth about 4,800 dollars a month at the same rate, and that is how you choose which page to write next.
Two honest notes to include when you present it. The lift is an assumption until measured, so present it as the hurdle the work has to clear rather than as a forecast. And the first order value understates the case for any business with repeat purchase, so if you know your repeat rate, use lifetime value and say which you used. The full formula and where each input comes from is in how to calculate marketing ROI.
The order the numbers move in
This is the part that causes most of the arguments in month two, because the figure everybody watches is the last one to move.
- Pages indexed moves first. It is a binary and it is visible within days in Search Console.
- Impressions move next. The page is being shown for queries, often ones you did not plan for, and that list is the most useful early output of the whole programme.
- Average position moves after that, usually from somewhere beyond the first few pages towards it, in steps rather than in one jump.
- Clicks move last, because position has to reach the range where people click before the click arrives.
- Enquiries and revenue follow clicks, with your sales cycle added on top.
Agree that sequence before the work starts and the month two conversation is about impressions and query lists instead of about disappointment. Which reports to read and what each number counts is covered in the Search Console guide, and the metrics worth putting in front of a sceptical reader are in SEO KPIs.
Four situations where another spend comes first
- The business needs revenue inside thirty days. Paid search, outbound and your existing list are the channels that answer that timeframe. Start the search work in parallel if there is budget for both, and judge it on its own clock.
- Nobody is searching for the category yet. A genuinely new product has demand expressed as a problem rather than as a product name. The work then is content about the problem, which is slower and is better funded as demand creation than as search optimisation.
- The queries are owned end to end by marketplaces. In some categories the first page is aggregators, marketplaces and comparison sites. The right move is often to be present inside those places first and to own the long tail and the branded searches on your own site.
- The site cannot be changed. Where a head office, a franchisor or a locked platform owns the pages, the realistic levers are the profile, the listings and the pages you can control. Buy the fix to that first, because optimisation work on a site nobody can edit is a retainer with no mechanism.
What it costs, so the case is complete
A budget request with no price is a conversation that gets postponed. The shape in most markets is a monthly fee for ongoing work, a one off fee for a technical fix or a migration, and a per page or per project price for content. What matters in the comparison is what each tier actually produces each month, which is set out tier by tier in what SEO costs and what each price tier buys. Take two quotes for the same written brief and compare the deliverables rather than the headline number, because the same fee buys a technical audit from one supplier and four blog posts from another.
Why it matters more now than it did two years ago
Search results now carry AI generated summaries, and buyers also ask assistants directly rather than typing a query. That changes where the answer is displayed and it keeps the underlying asset the same, because the summaries and the assistants are built by reading pages. The practical consequence is that the work which earns a citation in an assistant's answer is largely the work that earned a position in search: a page that answers one question completely, is reachable by a crawler, states facts plainly and is corroborated elsewhere. One job, two surfaces, and the second surface names a shortlist of two or three companies rather than listing ten.
How to tell within ninety days whether it was worth it
- Record the baseline on day one: impressions, clicks, average position, indexed pages, and revenue per organic session. A programme with no baseline cannot be judged, only argued about.
- Count pages that moved, not the site average. An average hides the three pages that doubled and the forty that did nothing, and the three are the evidence.
- Read the query list monthly. New queries appearing is the earliest sign the work is landing, and it also tells you what to write next.
- Check one commercial number. Enquiries, bookings, trials or orders attributed to organic sessions, with the attribution method written down before the data arrives.
Straight answers on whether this is worth doing
Is SEO still worth it with AI answers in the results?
The asset is the same asset. AI answers are assembled from pages, and the page that earns a citation is the page that answers one question well and can be reached and parsed. What changes is the measurement, because an answer can name you without sending a click, so brand searches and assistant citations belong in the reporting beside clicks.
When can we judge whether it is paying back?
Read the numbers in the order above: indexing and impressions in Search Console, then positions, then enquiries by page. Google's SEO Starter Guide says: "Some changes might take effect in a few hours, others could take several months. In general, you likely want to wait a few weeks to assess whether your work had beneficial effects in Google Search results." When the commercial case is settled depends on competition, your offer and your budget, and an established site has more to build on than a new domain. The order above is the thing to watch in the meantime.
Is it cheaper than paid search?
Per visit, over a long enough period, usually yes, and the comparison only holds if you count the fee against the visits the pages are still earning a year later. In the first two months paid search is better value, which is the honest version of the answer.
What makes it fail?
Three things, in order: pages written for a query nobody searches, a site a crawler struggles to reach, and stopping at month four. The third is the most common and the most expensive, because the spend is incurred and the compounding never arrives.
Does a small business really need it?
A small business with a local market usually needs the listing, the profile and five good pages rather than a programme, and that work is cheap and durable. The ninety day version of it is in SEO for beginners.
Who owns it internally?
One person who can approve a page going live and get a developer's attention. The most common structural reason this work stalls is that nobody holds both of those, so the drafts queue and the technical fixes wait.
The same case, put to the other surface
Everything above is argued in clicks, because clicks are what a search channel has been measured in for twenty years. There is now a second place the same pages get read, where a buyer asks an assistant who to use and gets back two or three names with a sentence each, and where being named is the whole result whether or not a click follows. Any case for this work is stronger with a reading of that answer attached, because it shows the current position in words rather than in positions. AI Knows Us asks those buying questions across the main assistants and records which companies get named and what the answers were built from.