We run a second website. It is not a client project and it is not a demo. It is a technology repair and review site we built ourselves, and we use it as the place where we find out whether what we tell clients is actually true. It has been indexed for about eleven months. In that time it has gone from effectively invisible to attracting a volume of search impressions that is three orders of magnitude larger than where it started. This is an honest account of what that took, because the numbers are the least interesting part of the story.
We are deliberately not publishing the absolute figures. It is our own commercial data and there is no reason to hand competitors a map. What we can share is the shape of the curve, the timeline, and the work behind it, which is the part that would actually help you judge whether anyone is doing this properly for your business.
Key takeaways
- The first four months produced almost nothing, and that is normal rather than a sign of failure.
- The inflection came after months of consistent publishing, not after a single clever change.
- Technical foundations did not create the growth, but they were what allowed it to happen at all.
- Internal linking turned out to matter far more than we expected, and it is the thing most sites neglect.
- The work is unglamorous, repetitive and mostly invisible, which is precisely why it is easy to under-deliver on.
- Search volume is not the same as business value, and for a local trade the two can point in very different directions.
Who this is for
This is written for a business owner who has been pitched SEO and cannot tell the difference between a good proposal and an expensive one. It is also for anyone who has paid for search work before, seen a monthly report full of charts, and never quite been able to connect it to enquiries.
You do not need to care about technology repair sites to get something from it. The point is not our subject matter, it is the timeline and the labour, both of which apply to any site trying to earn search traffic honestly.
One thing to be clear about before we go further: this is our own site, in a different sector, with different economics to a local trade. We will come back to what does and does not transfer, because pretending everything transfers would be exactly the kind of thing this article is arguing against.
Why we built a second site at all
Every web design studio claims to understand search. Very few of them are willing to demonstrate it on something where the results are theirs to own and theirs to lose. Client sites are not a fair test, because you inherit their history, their constraints and their patience.
So we built our own. It let us test decisions at a scale and speed that no client engagement allows, and it meant the cost of being wrong landed on us rather than on somebody paying us. When we tell a client that a particular approach works, it is usually because we have already watched it work or fail somewhere we could afford to find out.
It also keeps us honest in a way that is slightly uncomfortable. When you have a site of your own with a public search footprint, you cannot tell yourself comforting stories about why the numbers are flat. The data does not care how good the strategy document was.
The first four months: almost nothing
The site went live and then, for four months, essentially nothing happened. Impressions were in the low thousands a month, which sounds like a number until you realise it means a handful of people a day glancing past you. Clicks were in the tens. Some months were worse than the month before.
This is the phase where most projects die. Not because the approach was wrong, but because four months is long enough for a business owner to lose faith, and long enough for whoever is doing the work to quietly slow down. If you have ever cancelled an SEO engagement at month three, this is almost certainly the phase you cancelled in.
Nothing about that period looked like progress from the outside. Internally it was the busiest stretch of the whole project, because that is when the foundations, the structure and the initial body of content were being built. The reward for all of it arrived months later.
The most dangerous point in any search project is month three, when a large amount of real work has been done and none of it is visible yet.
The shape of the curve
From roughly month five, impressions began to climb, and then kept climbing every single month without exception. Not doubling once and settling, but compounding: each month meaningfully ahead of the last, for seven consecutive months.
By month eleven, monthly impressions were around three orders of magnitude above where they started, and clicks had grown at a similar rate. The curve has not flattened. The most recent month is the strongest so far, and the trajectory suggests it is still early rather than mature.
The important detail is not the size of the multiple. It is that the growth was gradual, continuous and boring. There is no single month where something dramatic happened. There is no clever tactic we can point at. It is the accumulated effect of a lot of small things landing on top of each other.
The timeline, month by month
Put in a table, the pattern is easier to see than in prose. The middle column is what an owner would have seen on a report. The right-hand column is what was actually going on underneath it.
| Period | What was visible | What was actually happening |
|---|---|---|
| Months 1 to 2 | Almost no impressions, no clicks | Foundations, structure and the first body of content |
| Months 3 to 4 | Still flat, one month worse than the last | Publishing cadence established, indexing issues resolved |
| Month 5 | First visible movement | Earlier pages beginning to be indexed and ranked |
| Months 6 to 8 | Clear month-on-month growth | Compounding, plus internal linking work connecting pages up |
| Months 9 to 11 | Growth accelerating, still climbing | Same routine, maintained without interruption |
Notice that the right-hand column barely changes after month three. The work did not get cleverer as the results improved. It got more repetitive, and the results improved anyway.
Why compounding matters more than any single win
Search growth of this kind is not one page succeeding. It is hundreds of pages each earning a small amount of visibility, and then those pages reinforcing each other. The effect is closer to interest accruing than to a switch being flipped.
That has an uncomfortable implication for anyone buying SEO. If the value comes from accumulation, then a three-month trial cannot tell you whether it is working. You will be judging the project at precisely the point where the graph is flattest and the work is heaviest.
What "doing the work" actually meant
Here is where the honest part starts. The reason the curve looks like that is not insight. It is volume, cadence and consistency, sustained for the better part of a year without interruption.
Content was published continuously rather than in bursts. Not a flurry at the start and then silence, which is the pattern almost every neglected blog follows, but a steady output maintained month after month, including the months where nothing was happening and it felt pointless.
Everything was automated that could be automated, because the alternative was not doing it. Publishing, scheduling, health checks, link maintenance and reporting all run on a schedule rather than on somebody remembering. That is the only reason the cadence survived contact with the rest of the business.
The bit that surprised us
Internal linking mattered far more than we expected. Not a related-posts widget at the bottom of a page, but deliberate, contextual links between articles, including links pointing backwards from older pages to newer ones so that new work inherits some of the standing of the old.
We ended up building tooling specifically to manage it: to map which pages link to which, to identify pages with nothing pointing at them, and to repair links whose targets had changed. Pages with no inbound links are effectively invisible no matter how good they are, and you cannot spot that by eye once you pass a few dozen pages.
What the search data actually taught us
The most useful lesson had nothing to do with volume. It was about the shape of the queries that win. Almost everything that earns real traffic on that site is problem-shaped: somebody has hit a specific error, or something has stopped working, and they are typing the exact symptom into Google at the moment it is annoying them.
Very little of it is brand-shaped or category-shaped. Nobody arrives by searching the name of the site, and comparatively little arrives from broad category terms. The traffic comes from precise, urgent, narrow questions, which is also where the intent is strongest, because a person with a specific problem is a person who wants a specific answer right now.
That reframed how we think about content generally. The instinct is to write about what you sell. The data consistently says you should write about the problem the customer is experiencing just before they realise they need what you sell, in the words they would actually use to describe it.
How that translates to a local business
A trade does not have error codes, but the principle holds exactly. The searches with real intent are not "plumber Chelmsford", they are the specific worry: a boiler making a noise, a leak under a sink, a bill that has jumped, whether a job needs building regulations approval.
Pages that answer those questions honestly tend to attract fewer visitors than a broad page would, and convert far more of them. That is a better trade for a business that only needs a handful of good jobs a month, and it is a much cheaper page to write well than a generic services page nobody reads.
The parts nobody sees
None of the technical work created a single visitor by itself. What it did was remove the ceilings. A site that loads slowly, or that Google cannot crawl properly, or that quietly serves duplicate versions of the same page, will underperform no matter how much content you pour into it.
So the foundations were dealt with first and then kept in good order: how the site is structured, how pages are linked, how quickly they render, how they are described to search engines, and whether anything was accidentally blocking indexing. These are one-off decisions that keep needing to be re-checked, because a rebuild or a plugin can undo them silently.
The monitoring side matters as much as the setup. We check indexing coverage, link health and page performance on a schedule, because problems in this area do not announce themselves. A page can drop out of the index and nothing will tell you unless you are looking.
Why results like this are rarer than they should be
Nothing in the previous two sections is a secret. There is no proprietary method here. Everything we did is documented publicly, discussed openly, and available to anyone willing to read for a weekend.
The difficulty is not knowledge, it is stamina. The work is repetitive, it is largely invisible, and for the first several months it produces no evidence that it is working. That combination makes it unusually easy to let slide, and unusually hard for a client to detect when it has been let slide.
That is the honest tension in this industry. A strategy document can be written in a week. The work it describes takes a year, and the gap between the two is very difficult to see from the outside until enough time has passed that you have already spent the money. We are not suggesting that is anyone's intent. We are saying the structure of the thing makes under-delivery easy and detection slow, which is worth knowing before you sign anything.
Questions worth asking anybody, including us
- What will actually be published, and how often? A specific cadence you can check beats a promise of "ongoing content".
- Can I see the work as it happens? Access to the site, the analytics and the search data should be yours, not held on your behalf.
- What happens in months one to four? Anyone who implies you will see results in six weeks is either inexperienced or telling you what you want to hear.
- Who owns the accounts? If the analytics property, the search console and the tag container are in an agency's name, leaving becomes expensive.
- What would make you tell me to stop? The most useful answer any supplier can give is the circumstances under which they would advise against their own service.
What transfers to a small business site, and what does not
We should be careful here, because this is exactly the point where a case study usually overclaims. Our second site is in a sector with enormous search volume and a global audience. A plumber covering three towns is playing a completely different game.
What transfers is the shape of the thing: the slow start, the compounding, the primacy of consistency over cleverness, and the fact that technical foundations set a ceiling on everything else. Those hold regardless of sector.
What does not transfer is the scale. A local trade does not need hundreds of thousands of impressions and should not chase them. A few dozen of the right people, in the right postcodes, at the moment they need you, is worth more than a large number of visitors who will never call. For most local businesses the highest-return work is not a content programme at all: it is the Google Business Profile, a handful of genuinely useful service pages, and a site that loads fast and makes it obvious how to get in touch.
Common mistakes
These are the patterns we see most often when someone brings us a site that has had search work done on it. None of them are exotic, and all of them cost months.
They tend to share a root cause: a focus on the visible, reportable parts of SEO over the unglamorous parts that actually accumulate. Charts are easy to produce. Consistency is not.
Read this list against your own site, or against the last report you were sent, rather than against what you were promised at the outset. The gap between the two is usually where the problem is hiding, and it tends to be a gap of consistency rather than one of competence.
If more than two of these describe your situation, the issue is probably not that the wrong tactics were chosen. It is more likely that the right ones were started and then quietly not sustained, which is a much more common failure and a much harder one to see from a monthly report.
- Judging the project at month three. That is the flattest part of the curve and the heaviest part of the work.
- Publishing in bursts. Twelve articles in one month then nothing for six months performs far worse than two a month for a year.
- Ignoring internal linking entirely. New pages with nothing pointing at them start from zero every time.
- Treating technical work as a one-off. A rebuild or a plugin update can silently undo it.
- Chasing volume over relevance. Ranking for something with no commercial intent produces traffic and no enquiries.
- Not owning the accounts. If you cannot see the search data yourself, you are taking someone's word for it.
- Measuring rankings instead of enquiries. Position is a proxy. Phone calls are the thing.
A realistic way to approach this
If you take one thing from our own eleven months, let it be the timeline rather than the multiple. Plan for the shape of the curve and you will make far better decisions about what to spend and when to judge it.
The order below is roughly how we would sequence it for a small business, and it deliberately front-loads the cheap, fast wins before anything that takes a year to pay back. That ordering is not an accident: it means you get some return early, which makes it far easier to stay patient through the slow part.
It also means that if you decide to stop after step four, you have still gained something permanent rather than having paid for an unfinished programme. Structuring the work so that abandoning it is not a total loss is, in our view, simply what an honest sequence looks like.
- Fix the foundations first. Speed, structure, indexing and mobile experience. This does not grow traffic, it removes the ceiling on it.
- Claim and complete the Google Business Profile properly. For a local business this is usually the highest-return hour available.
- Get the core service pages genuinely useful. One page per service, written for a customer rather than for a search engine.
- Set up measurement before you start. Search Console and analytics, owned by you, so you can judge the work honestly later.
- Then, and only then, consider a content programme. And commit to twelve months or do not start, because six months of it is the worst of both worlds.
- Review at month six and month twelve. Not month three.
That last point is the one that costs people the most money. A programme abandoned at month four has absorbed all of the cost and none of the return, which is a worse outcome than never having started.
Frequently asked questions
How long before I see results from SEO?
Why will you not publish the actual numbers?
Does this mean I need hundreds of blog posts?
Was this done with automation?
How do I tell whether an agency is actually doing the work?
What is the single biggest technical mistake you see?
Is rankings or traffic the right thing to measure?
Where this fits
We are not publishing this to argue that every business needs a content programme. Most do not. We are publishing it because the honest shape of search work is rarely shown, and because it is difficult to judge a proposal when you have never seen what the year behind it actually looks like.
The reason we are comfortable putting our own timeline in public, flat months and all, is that the flat months are the honest part. Any case study can show you a graph going up. Far fewer will tell you how long you have to hold your nerve first, or admit that most of the interesting work happened before anything moved.
If you want the foundations dealt with properly, that is what we build: websites designed around enquiries and made to load quickly, and SEO that starts with what is actually holding a site back. If you would rather just ask whether any of this is worth it for your business, get in touch and we will give you a straight answer, including when the answer is no.
