Why SEO isn’t producing a result: how to evaluate SEO results

What counts as an SEO result

The results chain: from indexing to inquiries

Work on a site affects the result in sequence. Each link depends on the one before it: a page outside the index won’t get impressions, and without clicks there won’t be inquiries from search.

Link	What question it answers	What to watch	Where to look
Indexing	Are the pages you need in search	Indexed and excluded pages	The “Page indexing report” in Search Console; “Pages in search” in Yandex Webmaster
Visibility	Does the site show up for the queries you need	Impressions, queries, average position	The Performance report in Search Console; “Search query analytics” in Webmaster
Clicks	Are people clicking through from the SERP to the site	Clicks, CTR	The same Search Console and Webmaster reports
Target actions	Are search visitors doing what the site exists for	Goals and conversion of visits from search	Yandex Metrica or another web analytics system
Contribution to the business	Is search bringing in inquiries and sales	Inquiries and deals that came from search	The company’s CRM and inquiry records

A few terms from the table:

• An impression is a link to the site appearing in search results for a query; a click is a visitor going to the site from search results.
• CTR is the number of clicks divided by the number of impressions.
• A goal in Yandex Metrica is an action by a visitor that the site owner cares about — for example, viewing a specific page or clicking the “Place order” button.
• Conversion is the share of target visits — that is, visits where a goal was reached — among all visits.

The last link isn’t fully visible to a web analytics system. A deal might close later, by phone or in an office. That’s why the contribution of search to sales is judged from company data: how many inquiries and deals came from search, and how they turned out.

Why rankings are an intermediate metric

A ranking is where a link to the site sits in the SERP. In the Search Console report, the average position is the average rank of the topmost result from your site. It’s an averaged value: the site ranks higher for some queries and lower for others, and one number doesn’t show that. There’s a separate article on what makes up a site’s visibility and how to read it.

Rising rankings mean visibility has changed. But if clicks and target actions haven’t grown, there’s no result for the business yet. Rankings are useful as a signal about the state of one link in the chain, not as the final outcome of the work.

How to evaluate SEO results: metrics and data sources

Evaluating SEO, or SEO analytics, means collecting data for every link in the chain and comparing it over time. There are usually three sources. Google Search Console and Yandex Webmaster show what happens in search before someone clicks through to the site. A web analytics system shows what visitors do on the site itself. Google splits its own tools the same way: Search Console provides data on the site’s performance in search results, while Google Analytics covers how visitors interact with the site.

Google Search Console: the Performance report

The “Performance on Search results” report shows clicks, impressions, CTR and average position. You can group the data by query, page, country, device, search appearance and date. Periods in the report can be compared — for example, a recent period against the previous one, or against the same period last year.

Look not at the site as a whole, but at the pages you’ve worked on and the queries tied to them. If impressions are growing but CTR isn’t, people see the site in the SERP but don’t choose it. In that case, check how the page appears in search results. There’s a separate article on how the title and description affect the choice in the SERP.

Yandex Webmaster: search queries and pages in search

For Yandex, the same links are viewed in Webmaster:

• “Search query analytics” — impressions, clicks, CTR, average position and average click position by query. Webmaster only counts impressions and clicks up to position 50, so the statistics cover part of the data.
• “Page statistics in search” — how the site’s pages attract visitors from the SERP: impressions, clicks and CTR broken down by page.
• “Pages in search” — the indexing link: how many pages participate in search, how many were added and excluded, and why a page was excluded.

There’s a separate article on where to find these reports and what else is in Webmaster.

Web analytics: goals and conversions from search

Search Console and Webmaster end at the click. What the visitor did next is shown by a web analytics system, such as Yandex Metrica.

To see the result, you set up goals in Metrica. These can be submitting a form, clicking a phone number, paying for an order, or moving to a messenger. Conversion is calculated as the ratio of the number of target visits to the total number of visits.

Metrica flags visitors from search as a separate source. The “Search engines” report shows which search engines sent visitors to the site; Metrica determines the source from the referrer — the address the visitor came from. You can see goal statistics in reports built by visits: you pick the goal in the report table’s header. That shows how many goal completions search specifically brought, and what its conversion rate is. Other web analytics systems work the same way: you need configured target actions and traffic split by source.

Which metrics to pick for your site

You don’t need to watch everything at once. One or two metrics per link is enough:

• indexing — how many important pages are in search, and whether any are among the excluded ones;
• visibility — impressions of target pages for the queries assigned to them;
• clicks — clicks and CTR for those pages;
• target actions — one or two goals that reflect the site’s purpose: an inquiry, a call, a purchase; reading articles for a media site;
• contribution to the business — inquiries and sales from search, according to the company’s data.

Pick metrics based on the site’s goal, not on what’s easiest to calculate. Query frequency in keyword-research tools is a demand estimate, not a result of the work: it shows how many people search for a topic, not how many came to you. We don’t give universal “normal” values for CTR or conversion — they depend on the topic, the query type and the kind of SERP. Compare the site against its own past.

How to compare “before and after” without fooling yourself

Comparing “before and after” is the main way to evaluate a site’s promotion. But the difference between two periods isn’t always explained by SEO work. Google names several groups of reasons why search traffic changes: algorithm updates, technical problems, security problems, spam problems, seasonality and shifts in user interest, and a site migration.

Seasonality and demand shifts

User behavior changes because of seasonal factors and new trends, and the popularity of queries can drop. Google says outright: in that case, a traffic decline is almost certainly explained by external circumstances.

So compare comparable periods. In the Search Console performance report you can compare a period not only with the previous one, but with the same period last year. Google suggests looking at broader interest-in-topic trends in Google Trends. If demand for a topic dropped across the board, a drop in impressions isn’t a mistake in your work on the site.

Site changes and search updates

According to Google’s help docs, key algorithm updates can affect a site’s rankings. Technical errors can keep Google from crawling and indexing pages. During a site migration, rankings can be unstable. Small ranking fluctuations happen constantly and can recover on their own, without any action from you.

Keep a log of site changes: release dates, redesigns, template changes, migrations, edits to important pages. Then a jump in the data can be tied to a specific event instead of guessed at. Without a log, it’s easy to mistake the results of your SEO work for the aftermath of someone else’s release.

Compare groups of pages, not the whole site

A site’s overall traffic can hide changes moving in opposite directions: one section grows, another drops, and the sum looks flat. So look at the changes for the pages and queries that were actually worked on. Separately, look at sections the work didn’t touch: if they changed the same way, the cause is most likely a shared one.

Why SEO isn’t producing a result: where the chain breaks

If there’s no result, first use the data from the sections above to determine which link in the chain it’s lost at. Then check the areas responsible for that link. Below are eight diagnostic areas. There’s a separate article on how an audit checks each of these areas.

Area	Symptom in the data	Where to check
Crawling and indexing	The pages you need aren’t in search, there are no impressions	The page indexing report, “Pages in search”
Demand and keyword set	The page is indexed, but there are almost no impressions	Whether there’s demand for the topic, whether queries are assigned to the page
Intent	There are impressions, rankings are low, the SERP shows other page types	The SERP for the target queries
Content	There are clicks, but the page doesn’t hold visitors or lead them further	Search visitors’ behavior in web analytics
Structure and internal links	An important page sits far from the homepage or has no incoming links	The section hierarchy and links to the page
External factors	The data changes without any changes on the site	Seasonality, demand, search updates
Conversion	Visitors from search arrive but don’t become customers	The path after the click: the offer, the form, how inquiries are handled
Measurement	There are inquiries, but analytics doesn’t show them, or shows them as “not from search”	Goal setup, tags, the attribution model

Pages aren’t crawled or indexed

Symptom: the pages you need aren’t in search, and there are no impressions for them. Check the page indexing report in Search Console and “Pages in search” in Webmaster: they show whether a page is excluded and why. Find out which pages are actually in the index and why before you start changing the text.

There’s no demand for the pages

Symptom: the page is indexed, but there are almost no impressions. The topic may barely be searched for, or the page may be written in phrasing that doesn’t match the queries. First check whether people search for the page’s topic using the words it’s written in. If there’s demand, check whether query groups are assigned to the page.

The page doesn’t match the query’s intent

Symptom: there are impressions, rankings are low, and the SERP for the query shows a different kind of page. For example, you’re showing a service page, but search is returning how-to articles. Check whether your page matches what people are searching for with that query.

The content doesn’t answer the question

Symptom: there are clicks, but visitors leave quickly and don’t move further through the site. That means the page attracts people but doesn’t help them. Check whether the text answers the question that brought the person there, and whether the page has a clear next step.

Important pages are weakly tied into the site’s structure

Symptom: an important page sits deep in the structure, or almost nothing links to it from other pages. Check where the page sits in the section hierarchy and whether internal links lead to the important pages.

External factors: competitors, demand, search updates

Not everything depends on the site. Google lists algorithm updates and seasonal shifts in interest in queries among the causes of traffic change. On top of that, the result doesn’t depend only on you: sites that show up for the same queries are also evolving. That’s our own explanation, not wording from a help doc. If the data changed without any change on your site, first check demand and compare periods as described above.

Visitors arrive but don’t become customers

Symptom: target traffic from search is growing, but inquiries haven’t increased. The problem might not be SEO, but what happens after the click: the offer, the price, the inquiry form, how fast the sales rep responds. SEO is responsible for bringing people with the right query to the site. What they decide on the site also depends on other parts of the business.

There’s a result, but you can’t see it

Symptom: inquiries are coming in, but analytics doesn’t show them, or logs them under a different source. Before concluding “SEO isn’t working,” check the measurement itself:

• goals — are they set up, and do they still fire after the site’s latest update, for example after a form was replaced;
• tags — UTM tags can get lost on a redirect to another page; the source is also lost if the landing page has no counter, or a browser extension blocks it from loading;
• the attribution model — the method for crediting different sources’ contribution to a conversion. In Metrica, visits get distributed across sources differently depending on the model you choose. Google also names different attribution models as one reason Search Console and Google Analytics data don’t match.

What a report on SEO work should include

A good report explains what changed in the chain and why. Questions it should answer:

1. What was done over the period and why — what problem each action solved.
2. Which links in the chain changed, and what data shows it.
3. Which pages and queries were compared, not just the site as a whole.
4. What adjustments were made when comparing periods: seasonality, demand, site changes.
5. What didn’t work and why.
6. What’s next and in what order.

Signs of a weak report: it only has rankings; it’s only a list of work performed; the actions aren’t tied to data; it contains promises about rankings or timelines. That doesn’t necessarily mean the work is bad. But a report like that can’t tell you whether there’s a result.

When every action is tied to “before and after” data, SEO becomes a managed system: decisions are made from numbers, not from gut feeling. How our SEO work with regular “before and after” checks is set up is described on the service page.

When SEO doesn’t fit, or isn’t the main channel

SEO isn’t a universal answer. It’s unlikely to become the main channel if:

• there’s almost no demand for the topic — people don’t search for the product using words that could find it;
• you need a fast, one-off burst of visitors by a date — for an event or a sale, for example;
• there’s no resource for ongoing work — SEO requires constant changes and data checks;
• the site or the offer is still changing — optimizing something that’s about to be redone soon isn’t efficient.

For fast traffic, PPC advertising is usually the tool. It’s a separate channel with its own logic and budget.

Why SEO results can’t be guaranteed, and where to start

Google says nobody can guarantee a number-one spot in its search results, and advises being wary of any specialist who promises guaranteed high rankings. In its beginner’s guide, Google notes that it has no recipe that would automatically put a site in the top positions.

A simple conclusion follows for evaluation. What’s worth evaluating is movement through the chain and the quality of the decisions: what was checked, what was changed, what the data showed. A promised ranking is a poor criterion, and the promise itself is a reason to be cautious.

It’s worth starting with two things: recording the baseline state for every link, and finding the link where the result is getting lost. An SEO audit that records the site’s baseline state is a good fit for this: it gives you something to compare “after” against.

Frequently asked questions

Rankings went up, but inquiries didn’t increase — what should I check?

Walk through the chain after visibility. Did clicks and CTR grow for the same pages? If there are clicks, check whether visitors from search are reaching goals in web analytics. If goals are being reached, but the company isn’t getting more inquiries, look at the “conversion” and “measurement” areas.

Search traffic dropped — does that mean SEO stopped working?

Not necessarily. Traffic changes because of seasonality and shifting interest in the topic, algorithm updates, technical problems, and site migrations. Compare the period against the same period last year, check the dates against your site change log, and only then draw conclusions.

Why don’t the numbers from Search Console, Webmaster and Metrica match?

They’re counting different things. Search Console and Webmaster count clicks in the SERP, while analytics systems count visits and sessions on the site. Clicks and sessions are counted differently, so the numbers will most likely diverge. Search Console also processes the data further — for example, removing duplicates — and doesn’t show rare queries in the report. Webmaster only counts impressions and clicks up to position 50. Metrica won’t see a visit if the counter didn’t load. Compare the trend within each system, not the absolute numbers between them.

What access and data do you need to evaluate SEO results?

Access to Google Search Console and Yandex Webmaster, a web analytics system with goals configured, a log of site changes, and data on inquiries and sales with the source noted. Without the last item, you can evaluate the early links in the chain, but not search’s contribution to the business.