Does SEO work in 2026? The data, not the opinions.

of trackable website traffic comes from organic search. Paid search brings 15%.

Source

BrightEdge Channel Report, 2019 ↗

Sample

thousands of domains, tens of billions of sessions

But: 2019 data — BrightEdge's last full channel report. Shares vary widely by industry (B2B: 76% from search in total).

of searches in Bulgaria go through Google. 58% of them happen on a phone.

Source

StatCounter Global Stats — Bulgaria, 2026 ↗

Sample

June 2026, all platforms

But: Measures traffic to sites running StatCounter code, not every search. Bing is ~5% on desktop.

higher conversion from AI-assistant referrals (Perplexity 10.5% vs 1.76% from Google organic) — low volume, high intent.

Source

Seer Interactive; Similarweb, 2025 ↗secondary source

Sample

~25M impressions (Seer); other studies report 2–4×

But: AI platforms are ≈0.3% of all web traffic (SE Ranking, 2026). High conversion, small volume. A hedge, not a main channel.

of all pages on the web get zero visits from Google.

Source

Ahrefs — Search Traffic Study, 2023 ↗

Sample

14 billion pages

But: Ahrefs' three reasons: nobody searches the topic, no backlinks (only 1 in 6,671 pages without any reaches 1,000 visits/mo), wrong search intent. All three are checkable before writing.

of newly published pages reach Google's top 10 within their first year.

Source

Ahrefs — SEO statistics, 2026 ↗

Sample

Ahrefs index

But: Most published content has no strategy behind it. The number argues for research before writing, not against SEO.

until the first measurable SEO results. For a brand-new domain Google says: up to 12 months.

Source

Ahrefs poll · John Mueller, Google, 2024 ↗

Sample

3,680 LinkedIn/X responses — a poll, not a controlled study

But: Our own domain: 13 indexed pages and 1 Google click in month one. We say it because it's normal — and because anyone promising results "in weeks" hasn't looked at the data.

click on a classic result when Google shows an AI summary, versus without one. 26% of AI-summary sessions end with no click at all.

Source

Pew Research Center, 2025 ↗

Sample

900 US adults, 68,879 searches, March 2025

But: Google only, US only. An AI summary appeared in 18% of searches — mostly informational, not "buying now".

organic clicks when an AI Overview is shown — in a randomized experiment. Paid clicks: roughly unchanged.

Source

SSRN field experiment, 2026 ↗secondary source

Sample

1,065 Chrome users, randomized

But: The strongest causal evidence so far. The effect is largest on informational queries; transactional ones ("price", "near me") are barely touched (AWR, 2026).

organic CTR on AI-Overview queries — bottoming out (−61%) in 2025 and partially recovering by February 2026.

Source

Seer Interactive, 2026 ↗secondary source

Sample

42 → 53 brands, 3,119 informational queries

But: Two different samples — not an apples-to-apples comparison. The message: the game changes but doesn't vanish; pages the AI cites win.

cost per new customer via SEO, by approach and segment (B2C thought-leadership $298 · B2B basic $1,786). PPC: $290 B2C / $802 B2B. Facebook Ads B2C: $230. Email: $287 / $510.

Source

First Page Sage — CAC by Channel, 2025 ↗

Sample

~120 firms, 2022–2024, three-year average, US

But: One agency's data from its own campaigns. Includes a 4–6 month learning period with much higher CAC. Directional, not your niche.

average Google Ads click (2026). Conversion 8.18%, lead $66.69. Lawyers pay $9.87 per click, dentists $8.00, restaurants $2.05.

Source

WordStream / LocaliQ — Google Ads Benchmarks, 2026 ↗

Sample

medians, US, Google + Microsoft Ads search campaigns

But: Bulgarian CPCs are usually lower, but the industry proportions hold. We use them to compute "how many clicks your budget buys", not as a promise.

median cost per lead from Meta (Facebook/Instagram) lead campaigns. Click $1.92, conversion 7.72%. Restaurants $3.16/lead, lawyers $18.17, dentists $76.71.

Source

WordStream / LocaliQ — Facebook Ads Benchmarks, 2025 ↗secondary source

Sample

Apr 2024 – Jun 2025, lead objective, US

But: A lead ≠ a customer. Meta leads are cheaper but colder — they close at a lower rate than someone googling "dentist near me".

average return on email marketing — to existing customers and subscribers.

Source

Litmus — State of Email, 2025 ↗

Sample

self-reported by marketing leaders

But: Litmus sells email software. Wide spread: 35% get $10–36, 30% get $36–50, and 21% don't measure at all. Only works if you already have a list.

higher lead-to-customer conversion for referrals than any other channel. Referred customers are worth 16% more over six years.

Source

Marketo; Wharton (Schmitt, Skiera, Van den Bulte), 2013 ↗secondary source

Sample

Marketo: 10 channels; Wharton: German bank, 10,000 customers

But: The cheapest channel, but you can't "turn it on" with budget — it needs happy customers and a process for asking.

of Bulgarian companies with 10+ staff have a website. Only 16.2% sell online. 41.5% use social media.

Source

НСИ — ИКТ и e-търговия в предприятията, 2025 ↗

Sample

non-financial enterprises with 10+ employees

But: Half your competitors have no website at all. That is the strongest argument for organic search in Bulgaria: the niche is often empty.

of Bulgarian enterprises have "very low" digital intensity — the second-lowest level in the EU after Romania.

Source

Eurostat — Digital Intensity Index, 2023 ↗secondary source

Sample

enterprises with 10+ employees, EU-27

But: Weak online competition means a well-built site with structured data often wins the niche with modest effort. Doesn't hold in saturated niches (Sofia real estate, loans).