How SEO Actually Grows a Business (With Real 90-Day Data)
Every business that has ever paid for ads knows the feeling: the day you stop paying, the traffic stops. SEO behaves differently, and the difference is not philosophical — it shows up in the numbers. This post explains the mechanism by which search actually grows a business, and uses real figures from the sites we run to show what each stage looks like when it is working, and when it is not.
SEO is an asset, advertising is an expense
A paid campaign buys attention for as long as the budget lasts. A ranking page is a small piece of property that keeps producing while you sleep, and keeps producing next quarter. The economics only work if you understand that the payback is delayed: a page published today typically does nothing for four to twelve weeks, and then starts compounding.
That delay is why most businesses quit. They measure SEO on a 30-day window, see nothing, and go back to ads. The correct unit of measurement is a 90-day window compared against the previous 90 days.
The four-stage funnel that search growth actually follows
Search traffic is a chain, and it always breaks at one specific link. Diagnosing which one is most of the job:
- Coverage — does Google have your pages at all? If a URL is not indexed, nothing downstream matters.
- Impressions — are you appearing for queries people search? This measures topical reach.
- Position — are you appearing high enough to be seen? Below position 10, clicks are close to a rounding error.
- Click-through rate — given that you appear, do people choose you? This is a copywriting problem, not a ranking problem.
Multiply the four together and you get clicks. Improve the weakest one and clicks move. Improving a strong one has almost no effect, which is why generic "do more SEO" advice fails.
What a broken link in the chain looks like in real data
Here is one of our own properties, a construction tools site in Nepal, over the 90 days from 9 May to 6 August 2026. Its plot area conversion calculator earned 53,945 impressions at an average position of 7.0 — and 127 clicks. That is a click-through rate of 0.24%.
Nothing is wrong with that page's coverage, reach or ranking. All three are strong. The chain breaks entirely at stage four: the title and description are not winning the click. No amount of new content fixes that. Rewriting one title does.
Contrast it with another page on the same site: the property valuation tool, 229 impressions at position 5.8, 10.04% click-through rate. Same site, same authority, forty times the CTR. The difference is the promise the search result makes.
Growth, when it works: three real examples
These are Search Console numbers from properties we operate, comparing the 90 days ending 6 August 2026 against the previous 90 days.
| Property | Clicks | Previous | Change | Impressions | Previous |
|---|---|---|---|---|---|
| An online grocery store | 1,781 | 562 | +217% | 98,868 | 17,761 |
| A construction tools site | 1,425 | 594 | +140% | 115,168 | 18,895 |
| A new architecture site | 199 | 0 | from zero | 9,536 | 35 |
| An established design blog | 1,524 | 2,773 | −45% | 148,589 | 248,502 |
Notice the fourth row. It is there on purpose.
The part nobody puts in a case study: decay is real
That established site has more impressions than any other property in the portfolio and it lost 45% of its clicks. Nothing dramatic happened — no penalty, no migration. Its pages simply aged while competitors refreshed theirs, and its average position drifted to 10.2, right at the edge of page one where clicks fall off a cliff.
This is the single most under-appreciated fact about SEO as a business investment: the asset depreciates if you do not maintain it. A page about cement prices in 2025 stops being the best answer in 2026. Traffic does not fall in one visible crash; it leaks a few percent a month until someone notices a year later.
The practical consequence is that a content plan without a refresh plan is a treadmill. Roughly half of the work on a mature site should be updating what already ranks.
Branded versus discovery traffic: two different businesses
On the grocery store, the single query mountemart produced 173 clicks from 296 impressions — a 58.45% click-through rate at position 1. Meanwhile bounty chocolate price in nepal produced 12 clicks from 875 impressions at position 6.
These are two entirely different growth engines, and conflating them is how businesses fool themselves:
- Branded search converts extraordinarily well, because the person already decided to buy from you. But it does not grow the business — it captures demand that your brand, ads or word of mouth already created. If branded traffic is 80% of your search traffic, SEO is not acquiring customers for you; it is just routing them.
- Discovery search — product, price and question queries — converts worse per visit but is the only kind that brings you people who had never heard of you. This is the growth number to watch.
Split them before you judge whether SEO is working. A traffic chart that only goes up because your brand got popular is measuring your marketing, not your search strategy.
Impressions are the leading indicator
Both growing properties above show the same pattern: impressions grew far faster than clicks — 5.6x and 6.1x respectively, against 3.2x and 2.4x on clicks. That gap is not a failure. It is the normal shape of early growth.
What it means is that Google started showing those sites for far more queries, mostly at mid-page positions where clicks are thin. The impressions arrived first; the clicks follow as positions consolidate and titles get sharpened. If your impressions are flat, you have a content and authority problem. If impressions are climbing and clicks are not, you have a ranking or copy problem — a much better problem, and a much cheaper one to fix.
Traffic is not the point — engagement is the bridge to revenue
The construction tools site recorded 1,929 sessions and 1,574 users in GA4 over the same window, with a 68.3% engagement rate and an average session of 152 seconds. Two and a half minutes on a calculator page is a person doing real work with real intent — someone actually costing out a house.
That is the number a business owner should care about, because it is the closest proxy to revenue that search data offers. Ten thousand visitors who bounce in eight seconds are worth less than five hundred who spend two minutes. When you evaluate an SEO investment, ask for engagement rate and session duration alongside clicks, or you are only seeing the top of the funnel.
How to actually run this for your own business
- Fix coverage first. Check how many of your pages Google has indexed. Publishing more pages while half your site is uncrawled is wasted money.
- Find the broken stage. Pull every page with high impressions and low CTR. Those are titles to rewrite this week, not articles to write next quarter.
- Rescue before you build. List every page that has lost traffic year over year and refresh the top five.
- Separate branded from discovery in every report you look at.
- Publish for jobs, not for keywords. A calculator that answers "how many square feet is 4 aana" beats a 2,000-word article about land measurement, every time.
- Write down what you shipped, on the chart. Attribution is impossible three months later without notes.
- Measure in 90-day blocks. Anything shorter is noise.
The honest summary
SEO grows a business when it is treated as an operations discipline rather than a content campaign: index everything, find the one broken stage in the funnel, fix it, maintain what already works, and keep score over quarters. Our fastest-growing property tripled its clicks in 90 days. Our largest one lost nearly half of its traffic in the same window, doing nothing wrong except standing still.
Both outcomes came from the same discipline, applied or not applied.
Want the detail behind these numbers? Read the full 90-day portfolio report, or the walkthrough of how we run this loop inside GharNaksa SEO.