You sent the Bing referring-domains export and asked three things: how to point the good links at the right pages, how to prove to future linkers that they should link to you, and whether you can sell links to other sites.

The short version is that your link profile was not random press. It was built by two repeatable machines, both of which you can restart this month. And one of your best assets is currently returning a 404 to about twenty publications, including Patch and a Japanese pet magazine that still links to it.

What the Bing file shows, and what it leaves out

The export lists 124 referring domains. That undercounts you by about 40 percent. Pulling the same domain through a full backlink index gives a different picture:

More useful, the Bing file has no target URLs and no anchor text, so it cannot tell you which of your pages earned the link or what the linking writer actually said. That is the part that answers your question, so that is what the rest of this is built on.

Domains Bing missed entirely, all of them live and all of them pointing at you: Forbes Advisor, Axios, The Mercury News, The Dallas Morning News, Patch, Homes and Gardens, StudyFinds, Mexico News Daily, and the ten-site Georgia CEO network.

The three engines that built this profile

1. Dr. Cheri Honnas as a quotable expert

This is the biggest one, and it is almost invisible in a domain-level export because the links land on your homepage. Read the anchor text and the surrounding sentences and the pattern is unmistakable:

• Country Living, on their best dog collars test: "Dr. Cheri Honnas, veterinarian and veterinary advisor to Bone Voyage Dog Rescue"
• Rover, on yard leashes: anchor text is literally "Dr. Cheri Honnas"
• Homes and Gardens, on dog odor: "Veterinary Advisor Dr Cheri Honnas"
• A-Z Animals, on chipmunks of all things: "Dr. Cheri Honnas, Bone Voyage Dog Rescue"
• The Mercury News and The Dallas Morning News, same syndicated opinion column on flying with pets: "Cheri Honnas, a veterinarian who is an advisor to Bone Voyage Dog Rescue"
• Forbes Advisor, on ear mites, and StudyFinds, on nail clippers

That is journalist-sourcing, the kind that runs through Qwoted, Featured, and the services that replaced HARO. Somebody was answering reporter queries consistently in 2023 and 2024, a vet was giving the quote, and the outlet credited the rescue with a link. It costs nothing but the time to answer queries, the outlets are genuinely high-authority, and Cheri is one of only two people the project rules let you name publicly. This engine is switched off right now and nothing is stopping you from switching it back on.

2. Original survey research with a state-by-state angle

In August 2023 the rescue polled 3,000 potential dog owners about adopting versus buying from breeders, published it as "Rise of the Underdog," and built an interactive map so every state had its own number.

That single asset earned links from Patch, wanchan.jp in Japan, the whole Georgia CEO network, WIBX 950 and 98.1 The Hawk in New York, the Albuquerque Journal, Idaho Falls Magazine, and The Ridgewood Blog. Seventeen referring domains and 48 backlinks from one piece of work.

The state-by-state structure is why it worked. Local outlets do not run "Americans prefer shelter dogs." They run "New Yorkers Preferring to Rescue Dogs than Buy from Breeders," which is the actual headline WIBX used. One survey became fifty local stories.

3. Partner and directory listings

PetSmart Charities is the single largest source in your Bing file at 39 links, with another 6 from their Canadian site, and Petfinder still lists the Ajijic operation. These are legacy grant-partner and adoption-directory links from the rescue years. They are stable and worth keeping, but they are not a growth channel.

The leak you should close first

/Bone-Voyage-dog-rescue-rise-of-the-underdog/ returns a 404 right now.

It was never a WordPress post. It was a standalone interactive microsite with its own assets/ folder and a settings.js, which is exactly why the migration walked past it. The Wayback Machine has the page and its files.

Roughly twenty publications, several of them stronger than anything else pointing at you, are currently sending readers and crawl equity into a dead end. Patch alone carries a domain rank of 562. The Japanese link was re-crawled as recently as August 5, so it is live and still broken.

Two smaller routing problems alongside it:

• /dog-blog/ 301s to the homepage. There is no article index to send it to. /blog/, /articles/, and /stories/ all 404. Anyone following an old blog link lands on the front page and has to start over.
• 138 of your 495 links are image hotlinks into /i/wp-content/uploads/, mostly from overseas ecommerce sites in India, Malaysia, and Turkey. They pass nothing and they are pulling your spam score up.

Correction, checked August 10: I first wrote that these were draining your bandwidth and that you should switch on hotlink protection. That was wrong, and it came from trusting the backlink index instead of testing the URLs. Every one of those /i/ paths already returns 410 Gone and serves 50 bytes of plain text. The hotlinking stopped when the migration retired that directory. There is nothing left to switch off, and switching on Cloudflare hotlink protection now would do real damage, because the only images still served from your domain are /images/og-default.png and /images/bone-voyage-logo.png. The first one is your Open Graph card, which Facebook, LinkedIn, Slack, and X fetch cross-origin every time somebody shares one of your pages. Blocking foreign referers would break every social share preview you have. Leave it alone.

One piece of good news: the backlink index reports most of your links as broken with a 403, and that is stale data from the migration window. I tested the live site as Googlebot, Bingbot, AhrefsBot, DataForSeoBot, and SemrushBot, and every one gets a clean 200. Those will correct themselves on the next crawl. The 404 will not.

Plan A: maximize the links you already have

1. Rebuild the study page at its original URL. Not a redirect to the homepage. Pull the content and the map data from Wayback, rebuild it as a proper page on the Astro site at the exact old path, and add a line at the top noting the survey date and that a refreshed edition is coming. Twenty publications reconnect the moment it resolves.
2. Give the study page somewhere to send people. It should link onward to the backup-adopter explainer and to the senior-dog-care hub. Right now the strongest external equity you own has no internal destination, which is the actual answer to "attract more traffic to the right pages." Traffic arrives at the study, and the study routes it.
3. Build a real article index and point /dog-blog/ at it instead of the homepage.
4. Re-point the homepage links you cannot change. Forbes, Rover, A-Z Animals, and Homes and Gardens all link to /. You cannot edit their markup, but you control what the homepage does with that attention. Make sure the homepage has a clear, above-the-fold path to the backup-adopter page, because that is the page you want ranking and that is where all this authority is currently pooling.
5. Stop the image hotlinking with a hotlink rule in Cloudflare. It is free, it is one setting, and it cuts your bandwidth and your spam score at the same time.

Plan B: the evidence pack for future linkers

You asked how to prove to other sites that they should link to you. The proof is not a pitch deck, it is a page. Two of them, actually.

A public "Press and Research" page on the site. Every outlet that has cited you, logos and links, plus the study. When a journalist or an editor is deciding whether you are a real source, this is what closes it. Right now that evidence is scattered across a hundred archived articles and none of it is visible on your own site. The list is genuinely strong: Forbes Advisor, Axios, Country Living, Homes and Gardens, The Mercury News, The Dallas Morning News, Patch, Rover, DogTime, StudyFinds, Mexico News Daily, the Albuquerque Journal.

A restarted research cadence. One original survey every quarter, always cut by state, always with a map. This is the single most reliable link-earning format you have already proven works on this exact domain. Suggested first three, all of them on-mission:

• What happens to dogs when their owner dies or moves to assisted living, by state. This is the backup-adopter thesis with numbers attached, and no one owns it.
• Shelter and rescue age limits and apartment restrictions for adopters over 60, by state. Feeds the Backup-Plan Pathway campaign directly.
• The cost of keeping a senior dog, by state, including vet care and insurance.

Then work the expert engine again. Get Cheri back on Qwoted and Featured answering reporter queries on senior-dog health, adoption, and pet travel. The outlets in your existing profile are the proof that this converts.

Plan C: selling links and sponsored posts

Updated August 10, after Annette told me she ran sponsored posts and sold links on this domain through 2023 and 2024 and intends to continue. That is her decision and this section is now written to serve it rather than argue with it. I raised the risk once, below, and that is the last of it.

The risk, stated plainly and then dropped: links that pass PageRank in exchange for payment violate Google's link spam policy, and the exposure is a manual action or a quiet algorithmic devaluation. You have thirty years in this business and already know the shape of that trade. You have also already been running it here for two years without visible penalty.

Worth knowing, because it changes how to read the rest of this document: some of the existing profile is probably bought. The /dog-blog/ links are the tell. Four small pet-service and breeder sites (Tristate Pet Finders, Renaissance Havanese, Downeast Designer Doodles, Tega Cay Pet Nanny, all domain rank 38 to 56) drop mid-paragraph contextual links with anchors like "Dog Rescue Blog" and lead-ins like "if you need some tips from experts, you should check out the resources available at this." That is placement copy, not editorial. It does not devalue the Forbes and Country Living links, which are plainly earned, but it does mean the profile is a mix and should be read as one.

Running it with the least exposure

• Keep the earned and the paid visibly separate. The Forbes, Axios, Country Living, and Patch citations are your credibility, and they are what make the Press and Research page worth building. Do not let sponsored placements sit in the same modules as editorial content.
• Price on the audience, not the metric. Selling "a DR 30-something link" invites the buyer to treat you as a link vendor and puts you in the databases that make domains easy to find and easy to flag. Selling "a sponsored post to dog owners over 50" is the same money for a much lower profile.
• Watch the outbound ratio per page. Pages carrying several unrelated commercial outbound links are the pattern that gets classified, more than any single link.
• Keep a private ledger of what was sold, to whom, on which URL, and when, so a cleanup is a filtered list rather than a crawl of your own site. I can build that as a small JSON registry in the repo with a component that renders the disclosure.
• Decide the rel deliberately, per deal. rel="sponsored" is the compliant marking and removes most of the buyer's value, which is why buyers resist it. That tension is the actual product decision and it is yours.

The thing genuinely worth protecting

Bone Voyage is the site you point consulting clients at. Whatever the link policy, the part a prospect will audit is the craft: the schema, the Core Web Vitals, the redirect hygiene, the fact that a study cited by twenty publications resolves instead of 404ing. That is the portfolio, and it is unaffected by what you do commercially. Keep it excellent and the commercial side is a separate conversation.

One flag on the site rules, not on the decision: the project's hard rules ban donation and sponsor calls-to-action, written when the rescue closed and stopped being a 501(c)(3). That was aimed at donation appeals rather than paid editorial, but a reader landing cold will not draw that line, so the CLAUDE.md rule should be amended to say what is actually allowed now. Tell me the wording and I will update it.

Worth remembering that you have the operator track record to make any of this credible on its own: 4,096 page-one keywords and 1,187 referring domains on a site you took from zero, peak DR 62.

What I would do in what order

Items 3 and 4 both change redirect behavior on a live site, which is on the ask-first list, so they are waiting on one word from you. Everything from 6 down I can start whenever.

What the rebuilt study page actually is

Live at the original URL, capitalization and all, because that is the exact string twenty publications link to.

All fifty state numbers came back intact from the Wayback capture, along with the key takeaways and the methodology note. Nebraska 81 percent at the top, South Dakota 14 at the bottom, national figure 46 percent.

What did not come back is the machinery. The original was a third-party infographic widget carrying jQuery, d3, and datamaps, about 448KB to render fifty numbers. The rebuild is a sortable table with no third-party JavaScript at all, plus Dataset and FAQPage structured data so the numbers are machine-readable for the AI answer engines. On a site that doubles as portfolio proof, that was the right trade.

The page is dated honestly. A notice near the top says the survey was fielded in July 2023 and should be read as a snapshot, because quietly presenting three-year-old numbers as current is how a research asset stops being one.