Paid converts a bit better per session (1.05% vs 0.85% CVR), but Organic's volume + zero marginal cost still wins on aggregate value. Marginal-lead value: +1M organic sessions = ~8.5K leads OT · ~9.3K STV · ~6.5K ATV.
Source: glue.odyn_eu_bi.session_finance_landing_category_agg (sessions) · featurama_store.europe_clean_hydra_* per-source split (leads), Apr 2026, country-scoped. OLX cross-traffic = source_type='organic' AND source_category='olx'. "Other" leads = Other + Organic Social + Internal Communication.
3 / 19
Current State · April 2026 · Organic Search
The Foundation We're Building On · 22M / 528K / 185K · ~30% Discover-Dependent
22.03M
Organic sessions · stack
+44.7% YoY
~15.8M
User-days · stack
avg ~528K daily organic users
184,631
Leads · stack
+43.0% YoY
0.84%
Stack-weighted CVR
OT 0.85 · STV 0.93 · ATV 0.65
Market
Sessions
User-days
Leads
CVR
YoY sessions
YoY leads
🇵🇱 Otomoto · PL
18.03M
12.67M
153,287
0.85%
+45.8%
+44.3%
🇵🇹 Standvirtual · PT
1.91M
1.50M
17,752
0.93%
+57.3%
+51.6%
🇷🇴 Autovit · RO
2.09M
1.63M
13,592
0.65%
+27.3%
+20.8%
Discover-attributed delta: +6.8M sessions (~31% of current organic-search) tracks the Aug-Oct 2025 Google Discover ramp. Pure Discover after subtracting baseline organic growth lands closer to ~24%.
Brand Demand Is in Managed Decline · −50% to −66% Over 5 Years
🇵🇱 Poland · −50%
🇵🇹 Portugal · −56%
🇷🇴 Romania · −66%
🇫🇷 France · La Centrale · −39%
Implication: brand and broad-category search are both eroding. Growth has migrated to specific attributes (body type, fuel, transmission) and specific models (BEV releases, hybrid trims). Non-brand strategy is no longer optional.
Source: Google Trends · 5-yr relative interest index 2021-2026 · ~/Documents/gtrends-tracker/.
5 / 19
Brand vs Non-Brand Split · Apr 2026
Leaders Run 80% Non-Brand · We Have Headroom
🇵🇱 Otomoto
4.76M brand · 4.34M non-brand · 48% non-brand
🇵🇹 Standvirtual
0.63M brand · 0.54M non-brand · 46% non-brand
🇷🇴 Autovit
0.43M brand · 0.85M non-brand · 66% non-brand
Target = ~80% non-brand (the "leader marketplace" healthy ratio). Benchmark: leader marketplaces in mature European auto verticals typically run 70-80%+ non-brand on organic. ATV at 66% has the right mix in OLX stack (partly because brand demand has collapsed fastest there, slide 5); OT at 48% and STV at 46% have 30+ points of headroom. Goal: keep growing brand. Non-brand should grow faster because that is where the search universe is shifting.
Source: GSC seo_search_analytics_by_query · is_branded flag · only source for this split. GSC anonymization (60-70% of unique queries; ~40-45% of clicks). Real non-brand share is materially higher than these floors. Leader-marketplace 70-80% reference from competitive Ahrefs / SimilarWeb scans.
6 / 19
Strategic Conclusion
Brand Is the Base, Non-Brand Is the Ocean
The brand side
▸ Otomoto/Standvirtual/Autovit brand demand down 50-66% over 5 years
▸ Still ~half of OT/STV organic traffic: high-CVR, low-cost
▸ Defend with brand SERP, dealer subdomains, app installs
▸ Don't expect it to grow: manage decline
The non-brand side
▸ Search universe is growing in attributes (body type, fuel, transmission) and models (BEV releases, hybrid trims)
▸ ATV at 66% non-brand share shows the mix is achievable
▸ Non-brand is where the ocean is: taxonomies, specs, leasing, info
▸ Compounding: every keyword won earns sessions for years
Implication: the next 4 slides are the 4 pillars of where non-brand organic traffic lives in motors. Each has a current state and at least one opportunity.
Trends: ~/Documents/gtrends-tracker/. Non-brand search-universe growth: keyword research across 200K candidate queries · DataForSEO + Ahrefs.
7 / 19
The Strategic Frame
Four Pillars · One Playbook
Every section, every bet, every action in this deck maps to one of these four. If a pillar isn't here, we don't ship for it.
BUY
Used cars, second-hand, ad-detail, category listings. The transactional core. Where ~90% of organic lands (verified OT/STV/ATV Apr 26).
SELL
Dealer subdomains, post-listing flow, seller pages. Where authority leaks when dealers churn.
LEASE
Long-term rental, financing, consumer leasing. Currently subscale; specialists win.
INFO
Research-intent: specs, prices, reviews, comparisons. The non-transactional pillar we've never owned.
+ GEO overlay across all four. AI engines (ChatGPT, Perplexity, Gemini, Grok, Google AIO) cite content in each pillar differently. Same content + structured data wins both classic SEO and GEO citations.
8 / 19
Pillar 1 · BUY · Taxonomy Expansion
Tip of the Iceberg · +445K/mo from 22% of the Universe
/osobowe?search[filter_enum_fuel_type]=electric Today: parameter URL, no on-page treatment
When a Polish buyer searches "BMW i4 cena", "SUV elektryczny do 200 tys" or "kombi diesel 2.0", Otomoto has the listings — but the URL is a query parameter (left) with no title, no H1, no body copy, no schema. Google can't rank what it can't read. The iceberg is the 49,000+ filter combinations that map directly to BUY intent. The tip is the 5,816 we already rank on. The middle — what we ship in FY26 — is the +445K monthly sessions sitting in pages we own but haven't treated.
Opportunity scorecard · PL
Universe identified
49,240 kws
P1 traffic addressable
2.09M/mo
Run analyzed (P75 cut)
10,722 kws (22%)
Current Otomoto traffic
536K/mo
Monthly opportunity
+445K/mo
Non-brand share of opp
92%
AI Overview presence
33.9% · cuts organic clicks ~38%
Otomoto ranking now
54% (5,816 / 10,722)
Zero presence
4,906 kws
Three moves, sequenced: (1) Expose fuel × body × trim × price combinations via sitemaps + internal links — make them crawlable. (2) Treat each facet with one-attribute title/H1/intro + Vehicle schema — give Google something to rank. (3) Two-tier pagination (canonical 2-9, robots-block 10+) — concentrate authority on pages that matter. Why this works: Pareto holds. Top 10% of keywords = 48.6% of opportunity. Disciplined coverage of the right 22% beats full coverage of everything. FY26 commit: +445K/mo from analyzed slice (54% already ranking). Multi-year ceiling: +700-900K/mo at full universe. Same template ships on STV + ATV.
Build the Research Surface · 1M+/mo Research-Intent
Today: a tech-specs catalog already lives at /news/katalog-samochodow — research content on an editorial path, no per-model live lease offers. The bet: relocate + enhance + monetize. A dedicated /info/<make>/<model> URL space for spec / price / dimensions / range queries. Three surfaces, three intents, all cross-linked:
MARKETPLACE
/osobowe/bmw/i4
Wins transactional · "bmw i4 oferty"
/INFO/ (NEW)
/info/bmw/i4
Wins research · "i4 cena · dane techniczne · zasięg · wymiary"
EDITORIAL
/news/<slug>
Wins opinion · "i4 test · vs comparisons"
Single URL per model, sections anchored: #cena · #dane-techniczne · #wymiary · #zasieg. Stacks Vehicle + VehicleDimensions + Offer + FAQPage schema. ~500 model+variant pages × 3 markets = ~1,500 pages from one template (MVP: top 50 model lines × 3 markets).
autocentrum.pl/dane-techniczne/bmw/i4 629K/mo on thin specs · no real prices
Why this wins: manufacturers have specs but no real prices; autocentrum has thin estimates at scale. Otomoto has real cross-brand data + live inventory + OtomotoLease offers. One template, 4 data sources, ~1,500 pages, addressable ~1M+ mo on cena + dane techniczne combined.
Autocentrum data: Ahrefs Site Explorer · May 2026 · autocentrum.pl/dane-techniczne/ path. /info/ concept: separate URL space displacing manufacturer + autocentrum + AIO citations. Existing seed catalog: otomoto.pl/news/katalog-samochodow (make/model browse, technical specs, no per-model monetization hooks today).
10 / 19
Pillar 2 · SELL · Dealer Subdomains
Recover the Dealer Equity · Three Tactics, One Template
989K
Stack dealer-sub organic sessions/mo
OT 872K · STV 53K · ATV 65K
8.5K+
Active OT dealer subs/day
vs 2,069 in GSC = ~4× larger
+2.0%
OT organic YoY · stable
STV +2.4% · ATV −22.0%
27.6%
Annual dealer churn · OT
2,346 dealers went dark in 12mo
Tactic 1 · Rank for brand+city
▸ 192 mid-tier dealers · 100+ sessions/day
▸ 52% invisible (zero brand-query demand); 48% rank #1 brand-only but fall to page 2 for brand+city
▸ Fix: locality schema + city H1 + internal links
▸ Lift: +50-100K mo organic
Tactic 2 · Rescue live-but-dark
▸ Paying clients, live subdomain, 0 clicks
▸ Listings removed → thin page → SERP drop
▸ Fix: prompt re-listing + thin-page content
▸ Lift: +20-40K mo organic
Tactic 3 · Reclaim churned equity
▸ Churned subs → 301 redirect to homepage
▸ Historical equity stranded, not dead
▸ Fix: curated landing → similar local dealers
▸ Recovery: ~30K mo organic
Combined OT upside: +100-170K sessions/mo. Ships on STV (+2.4% YoY) and ATV (−22% YoY). AggregateRating schema = star ratings in SERPs (+15-30% CTR lift on existing traffic).
dkimport.otomoto.pl · Akredyt, Biele "Na OTOMOTO od 2011" = 14-year tenure. 114 → 5 daily sessions (~3,400/mo lost). Live but empty: listings removed → thin page → SERP drop.
Source: URL-grain HYDRA (host-LIKE, organic-search only) · UTAD cross-check: 989K = 77% of seller_page total (1.28M) · HYDRA vs GSC gap: Discover lands outside organic-search bucket (+2% HYDRA vs +39% GSC) · Churn: 27.6% HYDRA (full pop.) vs 36.7% GSC (above anon threshold) · Live-but-dark confirmed via HTTP 200 · Brand+city: Airbyte GSC by_query, top-50 dealers Apr 2026.
11 / 19
Pillar 3 · LEASE · Otomoto · Poland
Top 3 Without Replatforming · 11th Today · 1.70% Share
5.3%
Of 2,202 PL lease keywords Otomoto ranks for
2,086 keywords with zero presence
1.70%
Market share · 11th place
superauto 7.93% · vehis 7.75% · automarket 7.65%
+85,729
Addressable visits/mo at #1
98.8% non-brand · 933 kws drive 80%
78K
Current organic-search sessions/mo
−46.2% YoY HYDRA · −75% in GSC web · corroborated
Four moves · enhance, don't replatform
▸ Optimise existing /otomotolease/ URLs: title/H1/intro/content for the keywords today's pages miss (leasing konsumencki, leasing kalkulator, [make] leasing)
▸ FAQPage + Product/Offer schema (88% of lease SERPs have PAA)
▸ Leverage otomoto.pl brand authority: faster to rank for "leasing" than vehis can for "samochody używane"
+ /info/<model> New-Car-Specs play
Slide 9's /info/ template already includes a leasing rate section (live OtomotoLease offer for that exact make/model). Users researching specs see the price they'd actually pay. Captures research-intent users at the spec stage, before they're shopping leases or used cars.
Compound effect: /info/ ranks for "[make] [model] cena" → user sees current lease offer → enters lease funnel. Outranks specialists because we own both the spec content and the live offer.
Keyword universe: Ahrefs 2,202 PL lease keywords · 411K mo searches · May 2026. Sessions: URL-grain HYDRA event data · lease-URL pattern · organic-search only. Decline cross-checked: GSC indexed-URL universe for the same pattern shrank 9 → 3 URLs (-67%) and web-search GSC clicks dropped 406 → 102 (-75%) Apr 25 → Apr 26, confirming the section is in real decline and not a URL-pattern artifact.
Missing head terms: leasing (24K), leasing konsumencki (14K), leasing samochodu (12K), leasing kalkulator (3K), auto leasing (1.9K), leasing operacyjny (2.5K).
Fix /home or build /leasing.
+12K traffic/mo
2Programmatic Make × Model × Body
1,795 missing lease pages (3.9% entity coverage).
Template proven at pos 5: /otomotolease/oferta/{make}/{model}/{body}-leasing-…
Scale to every make × model in PL.
+75K traffic/mo
3On-Page Attribute Optimisation
116 URLs already ranking (most at pos 4-10).
▸ Title + H1 around dominant kw
▸ Intro w/ primary entity
▸ FAQ block (capture PAA)
▸ Schema: Product, FAQ, Breadcrumb
▸ Anchor-text consistency
+1.5K traffic/mo (fast)
4Internal Linking Architecture
60 client URLs — authority fragmented.
▸ News / baza-wiedzy → /marka & /oferta
▸ Make × body grid on /home
▸ Fix "otomoto leasing" cannibalization
▸ Push equity from /osobowe/ to /lease/
Multiplier on 1, 2 & 3
Pillars 3 & 4 are the lever · internal linking + on-page attributes: the /otomotolease/oferta template works (hyundai/nowy-tucson at #5) only because it receives links. Pillars 1+2 (programmatic scale + head-term hero) produce thin, unranked pages without on-page discipline and link equity from hubs. Linking + on-page is what turns the other pillars from inventory into traffic.
Sizing: PL-only · Ahrefs Site Explorer May 2026 (head-term volumes) · OtomotoLease site-audit (entity coverage, ranking URLs, internal-link map) · 5-method market-share for #1-capture rate. Multiplier from pillar 4 not counted in FY26 commit.
12b / 19
Pillar 4 · INFO · Editorial + Research Queries
Best 2026 Is the Open Surface
Editorial trend · what's actually happening on OT /news/
Polish /news/ is in real decline on classic Web Search. Google click count fell −39% year-over-year (Apr 2025: ~427K → Apr 2026: ~259K). Google Discover added ~186K clicks for the same pages over the same period — material traffic that doesn't show in our session counter because Discover's in-app browser strips the Google referrer roughly 70% of the time, landing those sessions in organic-direct rather than organic-search. When we filter the session counter to the same scope (OT /news/, web-search only) it reconciles: −39% on both sides. The chart above shows a broader editorial bucket where Discover's offset softens the curve; the article-level decline is real and the bet 5 recovery (next slide) is sized against it.
"Best/Cheapest 2026" surface · May 2026:588 PL kws · +7-11K mo (+530-566%) · 100% non-brand · 87.8% trigger AI Overview (-62% click mult.) · FAQ schema: 582/588 (one template). Top targets: najtańszy samochód elektryczny, najmniej awaryjne auta, najtańszy chiński samochód. Three authority pages anchor the internal-linking spine.
Top OT /news/ URLs by GSC-clicks loss between Jan 2026 peak (677K) and Apr 2026 trough (337K). Green rows = evergreen utility, refresh-recoverable · grey row = aged-out news, not recoverable (excluded from sprint scope).
URL · OT /news/
Jan 26
Apr 26
Lost
Type
/uploads/.../umowa-kupna_sprzedazy.pdf
6,994
466
−6,528
Utility template
/legalne-ledy-zamiast-halogenow-czy-warto
14,463
5
−14,458
Legal/utility
/jaki-wybrac-silnik-do-bmw-serii-3-f30-31
4,812
1
−4,811
Buyer guide
/zmiany-w-przepisach-w-2026-r-kalendarz
2,175
4
−2,171
Annual refresh
/uzywany-peugeot-308-ii-opinie-usterki-silniki
2,317
2
−2,315
Used-car guide
/jak-za-darmo-sprawdzic-wyposazenie-auta-po-vin
5,741
4,059
−1,682
Tool (still ranking)
/najlepsze-suv-y-2025-ranking-oml
2,009
802
−1,207
Rebrand to 2026
/niedrogie-auta-z-4x4-duster-nie-jest-najtanszy
2,730
0
−2,730
Best-of guide
/nowa-alfa-romeo-tonale-z-2023-r-oferta
9,820
0
−9,820
News (decayed, excluded)
Top-25 cumulative loss
~75-80K monthly clicks lost across the top 25 declining URLs Jan → Apr 2026. Most are still indexed but demoted in ranking.
Sprint scope (recovery): top 20 evergreen URLs · 1 content writer × 4 weeks · ~100 hr total. ROI: ~35-45K mo clicks recovered. + "Best 2026" surface (parallel workstream): 588-kw universe · +7-11K mo traffic at #1 capture · FAQ schema on 99% of queries. Editorial CVR not validated; both bets sized in clicks/sessions. Same patterns replicate on STV + ATV (STV needs a 360-URL legacy /news/ → /diarioautomovel/ 301 review from Aug 2024).
Source: GSC seo_search_analytics_by_page Jan 2026 vs Apr 2026 · OT /news/ URL pattern · clicks delta. URL-level historical analysis only available in GSC; HYDRA can't replay URL clicks.
13b / 19
The Emerging Channel · GEO / AI Visibility
Named Everywhere. Sourced Rarely.
200
PL prompts tracked · Otterly
28 comparison · 172 informational + transactional
~40%
Mention rate · Otomoto in PL AI answers
vs ~50-60% in transactional prompts
0.9%
Citation SoV · Comparison prompts
Named ~40% · sourced 0.9% · ~44× attribution gap
8%
Overall citation SoV · PL
Strong in transactional, collapses on comparison
Pillar mapping · what AI engines want
▸ BUY: structured product data → Otomoto ad-detail with Vehicle/Offer schema
▸ INFO: spec + comparison data → /info/<model> with Vehicle + VehicleDimensions + Review schema
Why the comparison gap matters
AI engines preferentially cite structured content with extractable answers: autocentrum's spec pages get 2,700+ Grok citations because each query has a discrete factual answer.
Otomoto's ad-detail pages don't have comparison content. The /info/ namespace fixes this: one URL per model with explicit comparison sections + schema = the structured passage AI engines quote, with attribution back to Otomoto.
Shipping the protocol stack does not guarantee AI citations. What it does: makes us legible + cheap-to-fetch for systems already deciding whether to cite us.
Citation leverage is still a content + authority play. /info/ namespace + editorial recovery (slides 10, 13) are the structured corpus that earns the quote; the protocol layer just lowers the cost of finding it.
Standards landscape May 2026 · IETF / W3C / Anthropic agentskills.io / Google A2A · OT/STV/ATV roadmap pending eng scoping. Frame inspired by suganthan.com/blog/how-to-make-website-agent-ready.
14b / 19
Summary & Ask
Build the Moat · Five Bets · +838-938K Sessions FY26
Organic search delivers ~22% of stack sessions, growing +44.7% YoY. Brand demand is structurally declining (−50 to −66% over 5 years), so the play is non-brand. The five bets below: one platform fix per pillar, shipped once, benefits all three markets.
Bet
FY26 SEO opportunity
Multi-year ceiling
What ships
1 · BUY · Taxonomy
+445K mo
+700-900K mo
Expose fuel × body × trim × price via sitemaps + treat each facet (title/H1/intro + Vehicle schema)
2 · BUY · /info/
+200-300K mo
+1M+ mo
MVP: top 50 models × 3 markets on /info/<make>/<model>; full ~500 models at scale
3 · SELL · Dealers
+60K mo
+130K mo
FY26: mid-tier rank lift (192 dealers with own brand+city). FY27: post-churn landing + recovery
20-URL refresh sprint (~100 hr, +35K mo) · "Best 2026" surface = 588 verified PL kws, +7-11K mo at #1, FAQ schema on 99% · STV 360-URL legacy 301 review
Sum
+838-938K mo
~+1.9-2.1M+ mo
~7-8K extra leads/mo at stack CVR (editorial excluded; CVR not validated) · subject to eng scoping · GEO upside on top
FY26 = analyzed/verified addressable base · final scope subject to eng capacity · Multi-year ceiling = full-universe at same capture rate · GEO lift (slide 14) compounds with /info/ + editorial. Next slide: how we ship this — four parallel buy-ins.
15 / 19
Next Steps · How We Ship This
Four Parallel Buy-Ins to Unlock Delivery
The 5 bets above are the what. The four buy-ins below are the how. Bets become tickets only after all four converge — they run in parallel, the longest path determines launch.
1 · Business buy-in
C-suite · Country leadership
Ask: sign off on the FY26 +838-938K sessions / ~7-8K extra leads/mo target as a tracked OKR. Why now: without this, the bets stay "ideas" not commitments. Blocks every downstream buy-in.
2 · Product + Engineering buy-in
Product + Eng leads · Consumers, Sellers, Lease
Ask: sprint capacity for BUY (taxonomy + /info/ template), SELL (dealer-sub lift, schema), LEASE (URL enhancement, FAQ/Offer schema). Why now: bets 1-4 are platform work, not SEO work. SEO can ship copy/schema, not URL templates.
3 · Marketing buy-in
Marketing / Editorial
Ask: staffing for 20-URL evergreen refresh sprint (~100 hr) + "Best 2026" editorial program (588 PL kw universe, FAQ schema). 1 content writer × 4 weeks to start. Why now: editorial recovery is sized in clicks, not eng capacity. Bet 5 stalls without writers.
4 · JIRA slicing + delivery
Tech lead per pillar · SEO team
Ask: translate each bet into discrete tickets w/ scope, acceptance criteria, and sprint allocation. One epic per bet, dependencies mapped. Why now: tickets are the only way bets become shipped code. Without slicing, capacity allocation is theoretical.
Sequence: Business (1) unblocks Product+Eng (2) and Marketing (3); JIRA slicing (4) runs in parallel from week 1. The longest path determines launch timing — push all four in parallel from this week.
200 PL prompts × ChatGPT + Perplexity + Gemini + Grok + Google AIO · May 2026
Caveats: (1) Slides 2-4 + editorial chart use UTAD-aligned HYDRA aggregate. Dealer sub (11) + lease (12) use URL-grain HYDRA event table (~30-40% over-count vs UTAD; Discover gains land outside organic-search bucket). (2) GSC used only where HYDRA can't answer: Discover, brand split, parameter clicks, pre-Apr-2024, editorial Q23 scope. (3) HYDRA organic-search = source_type='organic' AND source_category='search'. (4) GSC anonymization: 60-70% of queries; non-brand shares are conservative floors. (5) Editorial CVR not validated; recovery sized in clicks, not leads. (6) La Centrale: own stack; Trends-only.
18 / 19
Annex
How Google Discover Counts in Our Organic-Search Bucket
Our headline +44.7% YoY lives in HYDRA's organic-search bucket = source_type='organic' AND source_category='search' (sessions arriving with a Google/Bing search referrer). Discover sessions do not arrive that way uniformly. They split between buckets depending on whether the in-app webview preserves or strips the google.com referrer. Here's how the split lands today.
→ ~30% land in organic-search (preserved); ~70% in organic-direct/other_referrals (stripped)
GSC total ad-detail clicks
21.28M
HYDRA organic-search ad-detail sessions
7.31M
= Web (1.44M) + Image (0.24M) + ~30% of Discover (~5.6M)
Why "partial" not "0% or 100%"
Discover's in-app webview behaves differently across Android versions, Google app builds, and OS-level referrer policies. Some variants preserve google.com as the referrer; others strip it. Empirically (this month, OT-PL ad-detail) the split sits near 30/70 preserved/stripped.
Referrer-policy risk
If Google tightens referrer-stripping in a future Android / Google-app update, our visible Discover-attributed organic-search would compress. Roughly 5-6M monthly sessions could shift from organic-search to Direct/Other overnight. Same users, different bucket.
Implication for the FY26 bets: Bets 1 (BUY · Taxonomy) and 2 (BUY · /info/) win on web-search ranking regardless of Discover-policy changes and protect us against the referrer-tightening downside. Bets 3-5 (SELL · Dealers, LEASE, INFO · Recover) are structurally similar. None of the FY26 commit depends on Discover staying as-is. Discover is upside, not the floor.
Sources: GSC seo_search_analytics_by_page (search_type breakdown, /osobowe/* ad-detail URL pattern) · HYDRA session_finance_landing_category_agg (landing_page_type='ad_page'). Reconciliation method: GSC clicks minus HYDRA org-search sessions on the same scope → implied stripping rate.
