7 Technical SEO Market Steps 2026 [Benchmarks Inside]

Technical SEO for an online marketplace is the crawl-budget, pagination, facet, and index work that lets Google discover seller, listing, and category URLs without drowning in filtered combinations. It is not a paid-acquisition mix, it is not a trust-and-safety queue, and it is not a homepage redesign. Screaming Frog, Sitebulb, Oncrawl, Lumar, Botify, JetOctopus, and ContentKing are the seven crawlers in this shortlist. US Tech Automations sits above them as the ticket layer that blocks a facet template from going indexable when canonical, pagination, and log proof have not all passed.

TL;DR: Buy Screaming Frog at £199/year for listing-template extracts and a sample crawl. Buy Sitebulb for owner-readable facet reports. Buy JetOctopus or Lumar for scheduled cloud crawls of category trees. Buy Oncrawl or Botify when logs prove Googlebot lives in ?color= and ?sort= instead of seller pages. Buy ContentKing when a merchandising deploy rewrites canonicals. Do not index every facet combination because "more URLs."

Marketplace technical SEO, defined

A marketplace is a catalog plus a seller graph plus a search UI. Google has to parse all three. The failure mode is combinatorial: 2,400 category URLs × 6 facets becomes a crawl trap, while 50,000 listings sit behind infinite scroll that a raw-HTML crawler never sees. Pagination and incremental loading are not edge cases; they are the product.

BEST_OF earn rate: 15.2% according to US Tech Automations (2026), on 12,514 live pages counted 2026-08-24. '7 Best' earn rate: 25.5% according to the Phase 1 12,514-page count (2026), versus 14.0% for "5 Best." Google Search Central documents how pagination and incremental page loading on ecommerce sites affect discovery of category and filtered URLs — treat that guidance as the spec for this vertical, not as a blog opinion.

ROI context is in SEO worth it for online marketplaces. Answer-engine work is in generative engine optimization for online marketplaces and how to get online marketplaces cited in Google AI Overviews.

Key Takeaways

• Marketplace technical SEO is facets, pagination, listing templates, and logs — not a homepage content score.

• Screaming Frog's public paid licence is £199/year with a 500-URL free cap (checked 2026-09-07).

• Cloud crawlers in this list require vendor quotes for TCO; do not invent a monthly dollar.

• Infinite scroll without paginated URLs is a discovery failure, not a UX win.

• Zapier can ship a sitemap ping; it cannot decide which facet URLs are indexable unless you encode that policy.

Who this is for

This page is for marketplace SEO leads, merchandising engineers, and agencies who inherit a category tree plus a seller storefront template.

Red flags: Skip enterprise crawlers if you have 80 listings you can export by hand. Skip log-file tools if you cannot get CDN or origin logs. Skip always-on monitoring if merchandisers will not accept a freeze when canonicals flip.

Glossary for marketplace crawls

• Facet URL: A category plus one or more filters (/shoes/blue/size-10). Index only when the combination is a real landing page.

• Search URL: ?q= results. Almost never a sitemap citizen.

• Listing template: The HTML (or JS) that every SKU shares. Extract against the template, ticket the template, not 50,000 SKUs.

• Seller storefront: The vendor's mini-site on your host. Often orphaned from the category tree.

• Pagination: page=2 or /c/shoes/2. Required if infinite scroll hides listings.

• Canonical to search: The failure where a category points at an internal search results URL.

• Crawl trap: A link graph that generates unbounded URLs (calendars, sorts, sessions).

• Take-rate: Marketplace revenue share. Not an SEO metric, but it is why listing indexability is a finance conversation.

Pagination and facet risks

Google Search Central's pagination and incremental-loading notes are the working spec: if a listing is only reachable after a scroll event, many crawls will never see it. If filtered URLs are indexable and canonical-to-self, you asked Googlebot to crawl combinatorics. The 7 steps below force a policy: which parameter URLs are indexable, which canonicalize to a clean category, and which are noindexed.

Benchmarks you can actually use (not invented conversion rates): Frog free = 500 URLs; Frog paid = £199/year; Ahrefs Lite = 100,000 crawl credits at $129/mo; Semrush SEO plan = 5 websites and 500 keywords at $139/mo monthly. A 50,000-listing marketplace will not fit in a free Frog crawl. That single number decides the first purchase.

Common facet mistakes: indexing sort orders (?sort=price), indexing in-stock toggles, indexing map views, and letting search-result URLs (?q=) into the sitemap. Each of those looks like "more landing pages" in a dashboard and like waste in a log file.

Criteria

Feature matrix

Paid spider licence: £199/year according to Screaming Frog (2026). Free crawl cap: 500 URLs according to Screaming Frog SEO Spider (2026). SEO plan list: $139/mo according to Semrush (2026). Lite list: $129/mo according to Ahrefs (2026).

TCO

Vendor profiles

Screaming Frog

Best fit: Extract price, SKU, and seller name from a listing template; sample-crawl categories. Limitations: 50,000 listings will not finish on a laptop without a list crawl of a sitemap slice. Implementation: List-mode the sitemap; custom extract; JS on category grids only. Save the .seospider so next week's merch push can use Crawl Comparison. Primary evidence: Screaming Frog. Who should not buy: Teams that need overnight full-index crawls with SSO and a 16 GB box is not available.

Sitebulb

Best fit: Explaining facet waste to a founder who will not read a 40-column CSV. Limitations: Contact vendor for current list price. Implementation: Hint reports on duplicate titles across sort URLs and redirect chains from the last domain migration. Primary evidence: Sitebulb. Who should not buy: Data teams that only want BigQuery exports and already live in Frog.

Oncrawl

Best fit: Log proof that Googlebot prefers ?color=. Limitations: Contact vendor; needs logs. Implementation: Segment listing vs facet vs search; ticket if facets win 7 days in a row. Primary evidence: Oncrawl. Who should not buy: Marketplaces without log pipelines or without an engineer who can join those logs to the sitemap.

Lumar

Best fit: Staging vs production before a merchandising relaunch. Limitations: Contact vendor. Implementation: Extract canonical on category templates; fail when extract empty or points at search. Put that fail in the same CI lane as the storefront build. Primary evidence: Lumar. Who should not buy: 100-listing pilots with no staging host.

Botify

Best fit: Crawl-budget control on large URL spaces. Limitations: Contact vendor. Implementation: Alarm when filtered URLs out-crawl seller stores for a week. Primary evidence: Botify. Who should not buy: Pre-launch catalogs or teams that cannot get logs. Do not buy it to "get more SKUs ranked."

JetOctopus

Best fit: Cloud JS crawls of category trees. Limitations: Contact vendor; confirm crawl-credit math on the quote. Implementation: Schedule after each facet-policy change and after each grid redesign. Primary evidence: JetOctopus. Who should not buy: Teams already in Lumar CI with no dual-run reason.

ContentKing

Best fit: Always-on when merch can edit canonicals. Limitations: Contact vendor; confirm Conductor SKU on the quote. Implementation: Alert on /c/ canonical changes and page a human before the next merch freeze. Primary evidence: ContentKing. Who should not buy: Teams that ignore alerts during launches or have no merch owner.

Merchandising and SEO will fight over "more pages." The crawl is the referee. If logs show Googlebot on sort URLs, the extra pages are a cost. If URL Inspection shows seller pages as Discovered-not-indexed, the extra pages are also a cost. Put both screenshots in the same ticket. Do not debate in a slide.

A 500-URL free Frog crawl is still useful as a template sample: pick 50 listings, 50 categories, 50 sellers, 50 facet URLs, 50 search URLs. That 250-URL sample tells you whether JS hides cards, whether canonicals point at search, and whether seller pages are orphans. It will not tell you crawl budget. Logs or a cloud crawler tell you crawl budget. Do not skip the sample because "we are too big for Frog." The remaining 250 free URLs can be a second-pass list of known-bad parameter patterns from the first export. Two sample crawls in one afternoon beat a six-week Botify procurement with no policy.

Worked example

A 50,000-listing marketplace with 2,400 category URLs, 6 facets, and a $24 take-rate on a $62 average order ships a new grid in React. Shopify-style products/update webhooks fire for 1,100 SKUs overnight. A 4-hour JetOctopus cloud crawl plus a Frog list-mode sample of 2,000 sitemap URLs shows 38% of Googlebot-equivalent hits on parameter URLs and 9 category pages with canonical-to-search. Search Console on a top seller returns inspectionResult.indexStatusResult.coverageState = Discovered - currently not indexed. The ticket stays open until the 6 facets have a written index policy, the 9 canonicals point at clean categories, and a human merchandiser signs that infinite scroll still exposes paginated links. US Tech Automations would ingest the crawl CSV and the policy table, open tickets per category template (not per listing), retry failed inspections, and stop for human review before anyone noindexes the entire /c/ tree.

Seller storefronts are a second template. They often inherit noindex from an old vendor experiment, or they canonicalize to the marketplace homepage. Crawl a sample of 50 sellers the same week you crawl categories. If sellers 404 when a shop is suspended, that is correct. If sellers 404 when a shop is live, that is a ticket. Do not mix suspended-seller 404s into the same count as category canonical bugs; they are different owners.

The 7 steps: write the facet allowlist; expose paginated links; crawl JS grids; extract listing cards; join logs; sample URL Inspection; ticket templates not SKUs. Add an eighth only after those seven have names on a roster.

When NOT to use US Tech Automations: If the catalog is 80 listings and a founder can click them, stay on free Frog. If Botify already owns crawl budget and merchandising honors the alarm, do not add a parallel queue. If n8n already pings sitemaps after products/update and a human checks coverage weekly, staff that check.

Zapier, Make, and n8n can retry sitemap pings, store run history, and branch on errors. They will not invent a facet policy, idempotent ticket keys, or access control around who may noindex /c/. A proposed ticketed design would key on category template IDs, collapse 500 listing 404s into one template ticket, and require merch + SEO dual approval before a robots change.

See the ticket layer on pricing. Other workflows start on the company homepage. Do not buy a second cloud crawler until the facet allowlist exists as a dated document with an owner, not a Slack thread. Print the allowlist. Put it next to the sitemap. If they disagree, the sitemap is wrong and the next crawl will prove it within one business day, before paid ads or email campaigns start sending those URLs to buyers this week without a crawl pass.

Questions

Which technical SEO tools do online marketplaces need first?

A crawler that can render category grids, plus logs if you have them. Frog for extracts; Oncrawl or Botify for log proof; ContentKing if merch can edit canonicals.

Should every facet URL be indexed?

No. Index combinations that have search demand and unique content. Canonical or noindex the rest. Prove the choice in logs, not in a slide.

How does pagination guidance change infinite scroll?

If listings are only reachable after scroll, add paginated URLs or crawlable links. Incremental loading that never changes the URL is a discovery risk.

Is Ahrefs Site Audit enough at 50,000 listings?

Lite's 100,000 crawl credits can sample, not always full-index. Standard at $249/mo raises credits to 500,000. Still pair with a specialist spider for extracts.

Can we stitch facet rules in Make?

Make can apply a spreadsheet policy to a robots or parameter map if you maintain that sheet. You still own the policy, the approvals, and the rollback.

When do we buy Botify instead of JetOctopus?

When crawl-budget modeling on logs is the job and procurement can finish. If you only need scheduled JS crawls, JetOctopus or Lumar is the smaller conversation.