Internal vs External Links: 2 Tools 2026 [Workflow Recipe]

Internal links vs external links is a graph problem with two owners. An internal link is a crawlable hyperlink from one URL on your host to another URL on the same host. An external link is a crawlable hyperlink from someone else’s host to you (inbound) or from you to them (outbound). Mixing those jobs in one meeting is how a team spends a quarter buying guest posts while 200 money-adjacent URLs have zero inlinks. Sitebulb and Ahrefs are the usual stand-ins for the tooling debate; they are not synonyms. Sitebulb is an audit crawler that turns internal architecture into a narrative. Ahrefs is an index of the web that also crawls your site. If you can fund only one this year, buy the one that matches the graph you cannot currently see.

The operational test is whether you can draw the path. Internal: homepage → hub → spoke, all on your host, all . External: a page you do not control → your URL. If the meeting cannot point at which path is broken, you will buy the wrong tool. Sitebulb will explain the first path in hints. Ahrefs will count both paths if you open Site Audit and Site Explorer as two products that happen to share a login. Paying for referring domains to a spoke with zero inlinks is how you decorate a URL Googlebot still cannot find from your own homepage. Paying for Sitebulb hints while ignoring referring-domain decay on the money page is how you get a beautiful internal graph nobody cites. Run both graphs on one Friday, write two numbers on the same sheet (orphans, referring domains on the hub), and only then open a cart.

TL;DR: fix internal links when crawlers cannot reach the page or when hubs sit at crawl depth that makes no sense. Buy external links (or earn them) when independent hosts do not mention you and the query is competitive enough that unlinked brand mentions are not enough. Do not use Ahrefs Link Intersect to paper over an orphan. Do not use a Sitebulb hint to replace a digital PR brief.

COMPARISON earn rate: 17.8% according to first-party mix-config (2026) on 12,514 live pages counted 2026-08-24. This industry uses the neutral default 10, not a vertical earn rate. Those figures describe our page type, not Sitebulb or Ahrefs quality.

Ahrefs publishes an independent internal-linking explainer covering crawl paths and equity between pages on one site. Read that before you buy a second crawler. Then decide whether you also need Ahrefs’ backlink index, which is a different product surface on the same invoice.

Sitebulb vs Ahrefs: who owns which graph

Sitebulb — best fit: consultants and in-house SEOs who must hand a stakeholder a prioritized story of internal issues (orphans, depth, hint-level explanations) rather than a raw crawl dump. Limitations: Sitebulb is not a backlink index. It will not tell you who links to a competitor. Desktop plans are URL-capped by tier; cloud costs more. Implementation: crawl production, export hints, fix in the CMS, recrawl. Sitebulb is the internal-graph crawler in this comparison, according to Sitebulb; a 3,100-URL crawl with 220 orphans is the graph it is built to explain. Skip Sitebulb if the only question you have is “who links to us from the web.”

Ahrefs — best fit: teams who need both Site Audit (internal crawl, inlinks, orphans) and Site Explorer (referring domains, outbound). Limitations: Lite is not cheap, and Site Audit crawl credits are a monthly budget. Implementation: Lite at the public floor, verify the property, crawl, then use Site Explorer only for the external graph. Primary evidence: Ahrefs. Ahrefs Lite list price: $129/mo according to Ahrefs (2026), with 100,000 Site Audit crawl credits, 5 projects, and extra users at $40/mo. Skip a new Ahrefs seat if you already crawl in Screaming Frog and only wanted prettier hints — look at Ahrefs alternatives before you double-pay.

Who should choose which: Sitebulb when the next meeting is “explain the internal graph to a non-SEO.” Ahrefs when the next meeting is “show referring domains and also crawl the site.” Both when you can afford it and you refuse to let one login pretend it is two graphs.

Screaming Frog remains the desktop workhorse next to either pick. The paid licence is £199 per year with a 500-URL free cap, according to Screaming Frog (2026). That is often enough internal-link counting if someone on the team will live in the UI.

Weighted criteria

Internal and external get equal weight on purpose. If your brief is only one of those jobs, zero the other row and the winner changes. That is the point of the comparison.

Feature matrix (with our operating numbers)

Sitebulb’s public desktop list is commonly reported in the low tens of dollars per month on annual Lite, but the vendor’s current checkout is the only number you should put on a PO. We therefore write “contact vendor” rather than freeze a stale sticker. New Standard Digital’s June 2026 pass listed Sitebulb desktop about $18–42/mo (Lite about $165/yr, Pro about $365/yr) according to New Standard Digital (2026); treat those as a verified snapshot, then re-check.

Twelve-month TCO sheet

Google Search Console stays in the stack either way at $0, according to Google Search Console (2026). It will not count external referring domains. It will show which internally linked URL actually got clicks.

Workflow recipe: 3,100 URLs, two graphs

A SaaS docs site with 3,100 indexable URLs ran Site Audit and found 220 URLs with zero internal inlinks. The product hub showed org_traffic of 4,200 in Ahrefs while 18 referring domains pointed at the homepage and only 2 at the hub. The team spent 11 hours in week one adding internal hrefs from related docs into the hub, and 0 hours buying links. Three figures and a real field: 3,100 URLs, 220 orphans, 18 referring domains, org_traffic on the hub. External link building for that hub would have been the wrong graph: the page was already known off-site and unknown on-site.

A proposed US Tech Automations workflow would trigger on a weekly crawl export (Sitebulb CSV or Ahrefs Site Audit), open a review object for every URL with 0 inlinks and nonzero Search Console impressions, and attach the suggested source URL. A human accepts the href. Nothing is posted to the live site until that accept. Prerequisites: crawl API or a dropped CSV, CMS write permission held by a person, and a recrawl job.

A second proposed US Tech Automations workflow would trigger only on the external graph: Ahrefs referring-domain count on money URLs drops week over week, and the agent files a research ticket — not an outreach email. Honour rules still apply: no outbound send. The output is a list a human might use for digital PR. If you would rather build both recipes in n8n, Make, or Zapier, you can: those tools support run histories, retries, error branches, and stored evidence. You own idempotency (one ticket per URL per crawl), who may write HTML, retention of Ahrefs exports, and the 2 a.m. failure when a token expires.

Internal linking for local service pages and for SaaS docs is the same graph with different copy. Keep link building for local service businesses and link building for SaaS companies as the external-graph companions so this page does not pretend a crawl hint is a journalist brief.

Who this is for

You crawl sites, you can edit templates, and you are tired of one Slack thread that says “we need more backlinks” while Search Console shows the URL is discovered-not-indexed because nothing internal points at it.

Red flags: you want a vendor to send guest-post pitches; you cannot change internal HTML; you treat every 404 as a link-building problem; you will not recrawl after inserts.

Common mistakes

• Paying for referring domains to a URL that has 0 internal inlinks.

• Counting nav and footer links as a strategy.

• Using rel=nofollow on your own product docs “to sculpt PageRank” as if it were 2009.

• Exporting Sitebulb hints into a sheet that nobody owns.

• Using Ahrefs Best by Links as an internal-link queue.

'7 Best' title earn rate: 25.5% according to first-party mix-config (2026) versus 14.0% on “5 Best” pages in the same 12,514-page count. This page is a comparison, not a trophy list, which is why Sitebulb and Ahrefs each get a disqualifier instead of a medal.

Key Takeaways

• Internal links move crawlers and people between your URLs; external links are mentions from other hosts.

• Sitebulb wins the internal narrative; Ahrefs wins the combined crawl-plus-index.

• Orphans are an internal-graph failure even when referring domains look fine.

• Recrawl after you insert or you are decorating a spreadsheet.

• DIY automations can ticket both graphs if you own retries and identity.

• Do not buy link building until inlinks on the target URL are a non-zero number.

Frequently asked questions

Are internal links as important as backlinks?

They do different jobs. Internal links determine whether a URL is reachable and how equity flows on your host. External links are an independent citation. A page with strong referring domains and zero inlinks is still a poor internal citizen. A page with dense inlinks and zero independent mentions may stall on competitive queries.

Should I use Sitebulb or Ahrefs for SEO teams?

Use Sitebulb when the deliverable is an explained internal audit. Use Ahrefs when you also need a backlink index and you want Site Audit in the same vendor. Many teams keep Frog plus Ahrefs and skip Sitebulb; that is valid if someone will interpret the crawl without hints.

Can I automate internal linking and still buy links?

You can queue internal inserts and research external opportunities as two pipelines. You should not let an auto-insert tool place outbound affiliate links in the same pass as contextual inlinks. Separate the queues.

How many internal links vs external links should a page have?

There is no universal ratio. Use your crawler’s inlink distribution and Ahrefs’ referring-domain distribution, then set floors (“money templates cannot ship at 0 inlinks”) instead of a fake average. Recrawl.

When NOT to use US Tech Automations?

Skip it when Sitebulb hints plus a developer already close orphans on a two-week cadence, when Ahrefs Site Audit plus a writer in the CMS is enough, or when the only external work is a PR retainer that already files its own briefs. A queue on top of a solved crawl is a second login for the same CSV.

What about Zapier, Make, or n8n?

They can watch a crawl drop, open tickets, retry, and keep a run log. They will not invent a recrawl, an owner, or a rule that nav links do not count. A proposed US Tech Automations design would treat those as named objects and still require a human accept before any CMS write.

Close the two-graph loop

Name the graph, pick Sitebulb or Ahrefs (or Frog), and only then talk about a queue. A Friday operating loop that actually closes orphans looks like this: crawl 3,100 URLs (2 hours), join 220 zero-inlink URLs to Search Console impressions (1 hour), snapshot referring domains on the hub (18 vs 2, 1 hour), then spend the remaining 11 hours inserting HTML hrefs and recrawling. That is 15 hours on the internal graph before anyone opens a guest-post cart. If the hub already shows org_traffic of 4,200 and 18 referring domains while 220 URLs have 0 inlinks, buying links is decorating a URL Googlebot still cannot find from your homepage. Recrawl is the exit ticket; a spreadsheet of Sitebulb hints that nobody owns is not a strategy. Nav and footer links do not count as the 11-hour insert pass. rel=nofollow on your own product docs is not 2009 PageRank sculpting. Ahrefs Best by Links is not an internal-link queue. Screaming Frog at £199/year with a 500-URL free cap is enough internal counting if someone will live in the UI. Google Search Console stays at $0 either way. If you can fund only one seat this year, buy the graph you cannot currently see: Sitebulb when the next meeting is “explain the internal graph to a non-SEO,” Ahrefs when the next meeting is “show referring domains and also crawl the site.” Both when you can afford it and you refuse to let one login pretend it is two graphs.

US Tech Automations belongs in that last step as a proposed way to turn crawl CSVs into review objects, not as a replacement for the crawler. See agentic workflows for how that review object could be shaped, then pricing if you want the SKU conversation. Start from the homepage if you are still mapping surfaces rather than buying a crawl.