SEO challenges are the constraints that make it hard to get traffic and enquiries from Google even when the site is technically sound: portals that take up page one, fewer clicks per search, long timelines, constant algorithm updates, incomplete data, limited resources and demand that is often small. On 16 commercial searches we measured on Google Italy on 8 October 2026, 37% of the organic results on page one belonged to directories, marketplaces and portals, and 45% to the sites of individual businesses.
An SEO error can be fixed. A challenge can't: you work around a challenge, by choosing where to compete.
It sounds like a textbook distinction. It isn't, because it changes where you put the budget.
Most of the articles on Google UK's first page for "seo challenges" talk about something else: AI, algorithm updates, mobile-first indexing, the technical issues found in audits. Several of them date from 2023 or 2024. We've already listed the errors, with data from 418,125 audits, in our guide to SEO mistakes. Here we're talking about what's left when the errors are down to zero and the traffic still doesn't come.
Two projects we work on show the scale. Macropix, an LED screen manufacturer based in Milan, took five years to go from position 88 to position 2 for "monitor pubblicitario" (advertising monitor). Today it has 25.55% share of voice on the ledwall cluster, ahead of Amazon at 18.08%.
Sunscape, which develops solar farms in Italy, went from 1 to 38 keywords on page one out of 143 in six months, with monthly impressions rising from 329 to 5,809.
Same method, very different timelines. Why?
Because the two companies were up against different challenges. Here are the seven we come across most often, each with a number next to it:
- who really occupies your page one, with our test on 16 Italian results pages and three repeated on Google UK;
- how many clicks are left after maps, product listings and AI Overviews;
- how long it takes, and why it depends on the search more than on the site;
- how to work through 10 algorithm updates in 19 months;
- what you can't see in Search Console;
- where to put the few resources you have, and why small demand is an advantage.
Let's start.
Challenge 1: your competitor is a portal (and it takes 37% of page one)
On a commercial search a business's first opponent often isn't another business: it's Yell, Checkatrade, Booking.com, Rightmove or a comparison site. To measure it, on 8 October 2026 we used DataForSEO to run 16 searches on Google Italy from a smartphone, across eight different sectors, and classified every organic result on page one by hand.
| Search (Google Italy) | Monthly searches | Element above the first organic result | Organic results | Portals, directories, social | Big brands | Sites of individual businesses |
|---|---|---|---|---|---|---|
| assicurazione auto (car insurance) | 201,000 | none | 10 | 7 | 2 | 0 |
| divano letto (sofa bed) | 90,500 | none | 9 | 2 | 3 | 4 |
| hotel firenze (hotels Florence) | 40,500 | hotel box | 7 | 4 | 1 | 1 |
| caldaia a condensazione (condensing boiler) | 40,500 | none | 8 | 5 | 0 | 1 |
| noleggio auto lungo termine (long-term car leasing) | 40,500 | none | 10 | 2 | 0 | 7 |
| idraulico milano (plumber Milan) | 27,100 | local business carousel | 10 | 7 | 0 | 3 |
| scarpe running uomo (men's running shoes) | 18,100 | product listings | 10 | 2 | 7 | 1 |
| corso inglese online (online English course) | 12,100 | AI Overview | 9 | 1 | 2 | 6 |
| traslochi roma (removals Rome) | 3,600 | local business carousel | 10 | 3 | 0 | 7 |
| dentista torino (dentist Turin) | 2,400 | local business carousel | 10 | 2 | 1 | 7 |
| commercialista roma (accountant Rome) | 1,600 | none | 10 | 6 | 0 | 4 |
| agenzia immobiliare napoli (estate agent Naples) | 1,000 | local business carousel | 10 | 5 | 5 | 0 |
| avvocato divorzista milano (divorce lawyer Milan) | 590 | none | 10 | 1 | 0 | 9 |
| impresa ristrutturazioni bologna (renovation company Bologna) | 110 | local business carousel | 10 | 6 | 0 | 4 |
| software gestionale pmi (SME management software) | 90 | AI Overview | 8 | 0 | 0 | 7 |
| lavorazioni cnc conto terzi (subcontract CNC machining) | 10 | none | 10 | 3 | 0 | 7 |
| Total | 9 results pages out of 16 | 151 | 56 (37%) | 21 (14%) | 68 (45%) |
The average says little. The figure that matters is the split by search volume.
On the 8 searches above 10,000 searches a month, the sites of individual businesses hold 32% of the places on page one. On the 8 below that threshold, 58%.
On "assicurazione auto" (car insurance), 201,000 searches a month in Italy, there isn't a single independent broker's site: seven results out of ten are comparison sites and portals such as Segugio, 6sicuro and Supermoney. We repeated the search on Google UK on 10 October 2026, and "car insurance" (450,000 searches a month) looks the same: no independent broker, four comparison sites (Compare the Market, TotallyMoney, MotorEasy, Uswitch), five big brands including Churchill, the RAC and Santander, and an AI Overview on top pointing to Confused.com and Compare the Market. On "avvocato divorzista milano" (divorce lawyer Milan), 590 searches, nine out of ten are law firms; on "divorce solicitor manchester" (320 searches in the UK), nine out of ten too.
So are the big searches lost from the start?
For a single site, almost always yes, at least in the short term. There are two exceptions in the test: "noleggio auto lungo termine" (long-term car leasing), where 7 results out of 10 are dealers and leasing companies, and cases where the company is already a national brand (Cisalfa and Nike for "scarpe running uomo", men's running shoes).
The practical consequence: where portals hold 6 or 7 places out of 10, it pays to be inside the portal (a complete profile on Yell, Checkatrade or Booking.com) and to move the website onto more specific searches. It's the strategy we describe in our guide to real estate SEO, where an agency beats the portals in its own area, and in the one on SEO for hotels, which doesn't take on Booking.com for "hotels in Milan".
Challenge 2: fewer clicks for every search (9 results pages out of 16 don't start with a website)
Every element Google places above the first organic result holds back some of the clicks. In our test the first organic result wasn't the first thing on the page in 9 searches out of 16: five times there was the local business carousel, once the hotel box, once the product listings, twice an AI Overview. The "People also ask" block was there in all 16.
On commercial searches in Italy the AI Overview is still rare. The UK is further ahead: on "car insurance" and "seo challenges" the AI Overview was the first element on the page when we checked on 10 October 2026, and "divorce solicitor manchester" opened with the map of three firms. On informational questions the AI answer is the norm: in our test on 24 desktop searches in Italy on 6 October 2026 it appeared on 6 informational questions out of 6, as we describe in our article on SEO trends.
And when it's there, it weighs. The two most solid figures are American, and should be read as such:
- Pew Research Center tracked the browsing of 900 US adults in March 2025: with an AI summary, people clicked a result in 8% of visits, without one in 15% (Pew Research, July 2025);
- Ahrefs, comparing 300,000 keywords between December 2023 and December 2025 on desktop, calculates that an AI Overview goes with a 58% lower CTR for the first result (Ahrefs, February 2026).
For the UK there's no equivalent click study, but there is a share: according to Ofcom's Online Nation 2025, about 30% of searches now show an AI Overview. What you can measure is your own figure: on our page about zero-click searches we explain how to estimate it with Search Console.
There's one detail in the test, though, that's worth more than the percentages...
On "software gestionale pmi" (SME management software) the AI Overview cites TeamSystem, Zucchetti, SAP, Factorial and Odoo. None of the five is among the 8 organic results on the same page. The AI answer doesn't draw only from page one, and whoever owns it plays a different game: how to measure it is in our guide to AI visibility.
How many of your clicks end up with a portal or an AI answer?
On 16 Italian searches, 37% of page one belonged to portals and directories, and "car insurance" in the UK looks the same. Let's look together at the queries that bring enquiries to your site and who really occupies them, to decide where it makes sense to push SEO and where it doesn't.
Book a strategy consultationChallenge 3: timelines (fewer than two new pages in a hundred reach the top 10 within a year)
How long it takes to climb on Google depends on the search more than on the site. Ahrefs followed 1 million URLs first seen in September 2023: within a year only 1.74% had reached the top 10 for at least one keyword. On a filtered sample of English-language pages with real content, the share rises to 6.11% (study by Patrick Stox for Ahrefs).
On category keywords it's even slower. In the same study, across 1.3 million US keywords, the page in first position is 5 years old on average (it was 2 in 2017) and 72.9% of the pages in the top 10 are more than 3 years old.
That's Macropix's curve: position 88 in 2020, position 2 in 2025. Sunscape, on the other hand, added 37 keywords to page one in six months because it was working on regional searches, with one page per region.
So SEO is no use if you're in a hurry?
It is, if the goals have two deadlines: three to six months for long, specific searches, years for category keywords. The five stages, with timings measured in Search Console, are in our guide to how long SEO takes to show results.
Challenge 4: Google changes while you work (119 days of updates in 19 months)
From 13 March 2025 to 8 October 2026 Google released 10 official ranking updates: 5 core updates and 5 spam updates. Adding up the durations listed on the Search Status Dashboard, an update was rolling out on about 119 days out of 574: one day in five.
The longest was the August 2025 spam update, 26 days and 15 hours. The latest started on 24 September 2026 and lasted 13 days and 16 hours.
The problem isn't the update itself. It's that if you change the site during a rollout, you no longer know what to put the effect down to.
Three rules we apply on our projects:
- record in the same calendar every change to the site and every Google rollout;
- don't touch the main pages while a core update is rolling out, unless there are errors blocking indexing;
- compare two windows of equal length before and after the rollout, starting on the same day of the week.
What to do when the drop is real is explained in our guide to Google algorithm updates and in the one on finding the cause of a traffic drop.
Challenge 5: measuring (Search Console hides almost half your clicks)
Search Console doesn't show every query: the rarest ones are anonymised to protect privacy, so they count in the totals but don't appear in the table. Ahrefs measured how much they weigh across 887,534 properties and 22 billion clicks: in April 2025, 46.77% of clicks came from hidden queries, and for many sites the share is between 45% and 80% (Ahrefs, February 2026, global sample).
In plain terms: a report built on keywords tells half the story. And the missing half is precisely the long searches, which for a B2B company are often the closest to a quote request.
The second problem is that position and clicks don't move together. On a site we work on, average position improved by 64% in eight months, clicks by 36%. If you only look at traffic you underestimate the work; if you only look at positions you overestimate it.
The way out is to measure by page and by enquiry, not by keyword: which URLs brought in enquiries, with forms and phone calls tracked. How to build an SEO report like that we explain separately, and how to take it to the board is in our guide on how to prove the value of SEO.
Challenge 6: resources (make fewer pages, not more)
The resource a business most lacks for SEO usually isn't budget: it's the time of the people who know things. The engineer who knows the tolerances, the owner who knows why a customer picks one supplier over another. Without them the copy becomes generic, and generic copy doesn't get traffic: 96.55% of the pages in Ahrefs' index, around 14 billion, get no visits from Google (Ahrefs, 2023).
AI has made the shortcut easier. According to the ONS, in June 2026 around 35% of UK businesses with 10 or more employees used AI, against around 12% in late 2023. Writing a hundred pages costs little. But Google's spam policies explicitly prohibit "scaled content abuse", and in 2026 there have already been four spam updates.
At Macropix we did the opposite: one page for each product line and one for each major installation, from the LED wall at CityLife in Milan to the displays for Milan Design Week. Few pages, each with something only the company could write. What the data says about generated copy is in our guide to AI content and SEO.
Challenge 7: small demand (10 searches a month can be enough)
In B2B and in niches the demand exists, but it's tiny. "Subcontract cnc machining" gets 40 searches a month in the UK according to Google Ads data; in our Italian test, "lavorazioni cnc conto terzi" (subcontract CNC machining) got 10 and "software gestionale pmi" (SME management software) 90. If you choose keywords by looking at volume, you throw them out.
And that's a mistake, for two reasons.
The first is in the test: on exactly those two Italian searches the sites of individual businesses hold 7 places out of 10 and 7 out of 8. On "subcontract cnc machining" on Google UK, on 10 October 2026, machine shops' own sites held 6 places out of 10. Competition from portals, challenge number one, is almost absent there.
The second is value. An industrial contract is worth thousands of pounds, and a few enquiries a year pay for a project. How to choose and weigh these searches is in our guide to zero-volume keywords and in the one on SEO for manufacturing.
Which challenges you can solve and which you work around (the table to keep)
Four challenges out of seven don't depend on you. You can only decide how to move around them, and where to start checking.
| Challenge | Up to you? | What to do | First check |
|---|---|---|---|
| 1. Portals on page one | No | Be present on the portal, move the site onto specific searches | How many places out of 10 belong to portals on your 20 main searches |
| 2. Fewer clicks per search | No | Target the searches where your site is the answer, measure impressions too | Ratio of clicks to impressions by page in Search Console |
| 3. Long timelines | Partly | Separate goals for long searches and category searches | How many queries you have between position 11 and 20 |
| 4. Google updates | No | One calendar for changes and rollouts | Rollout dates on the Search Status Dashboard |
| 5. Measurement | Yes | Measure by page and by enquiry | Forms and phone calls tracked by URL |
| 6. Resources | Yes | Fewer pages, written with the people who know the product | Pages with no clicks in the last 12 months |
| 7. Small demand | No, but it's an advantage | Work on low-volume, high-value searches | Average value of a contract divided by the annual cost of SEO |
Working out which of the seven weighs most on your site is how every SEO consultancy project of ours begins and, for the part about generated answers, our AI SEO service.
In short (and the challenge nobody talks about)
In our Italian test, 37% of page one belongs to portals and directories.
In 9 results pages out of 16 the first result isn't a website.
Fewer than two new pages in a hundred reach the top 10 within a year.
One day in five a Google update is rolling out.
Almost half your clicks come from queries you can't see.
The challenge the industry talks about least is telling a business where SEO isn't worth it. In 5 of the 16 searches in our test, independent sites had one place out of ten at most, and "car insurance" in the UK had none: there, the most honest answer is a profile on the portal or a paid campaign, not six months of content (the comparison with numbers is in our guide to SEO and SEM). Choosing where not to compete is the first decision in a strategy, and it's the one that saves the most.
Frequently asked questions about SEO challenges
It depends: on commercial queries above 10,000 searches a month it's competition from portals and directories, which in our test on Google Italy on 8 October 2026 held 41% of the places against 32% for business websites; on informational questions it's the AI Overview, which appeared in 6 cases out of 6 in our test of 6 October 2026.
Yes, and above all slower if you're starting now. According to Ahrefs, the page in first position is 5 years old on average, against 2 in 2017, and only 1.74% of new pages reach the top 10 within a year, against 5.7% in 2017.
Yes, on specific searches. In our test, on queries below 10,000 searches a month the sites of individual businesses held 58% of the places on page one, and on Google UK nine out of ten results for "divorce solicitor manchester" were law firms.
No. An error, such as a page set to noindex or a wrong redirect, gets fixed and the problem disappears. A challenge, such as competition from Yell or the drop in clicks caused by AI Overviews, stays: you choose how to work around it.
No. They can only be recorded and read: between March 2025 and October 2026 Google released 10 core and spam updates, for about 119 days of rollout in total according to the Search Status Dashboard.
