LinkBunny

ARTICLES / BACKLINKS

How to build your first backlinks for a new SaaS

A new SaaS starts with zero referring domains. Every list of link-building tactics seems to apply at once, and most of them assume you already have an audience, a blog and a budget.

For each of my projects, I mostly work through the same order each time: links I control, a few launch directories, a page/tool someone would link to, the ecosystems my product plugs into, then contextual links from relevant sites. Start with the links you can add yourself, then spend time on pages that need someone else's review.

Before any linkbuilding: one page worth linking to

Pick the page you want people to land on before you ask anyone for a link. For most new products that is the homepage or a working demo, or the pricing page. For a developer tool it might be the setup guide. If the page doesn't explain what the product does on a phone, fix that first, because a link can't fix an unclear page.

The new-site backlink checklist covers how to prepare that destination and check it for accidental noindex or canonical problems.

Step 1: Links you control (ASAP)

Start with places where you only need to fill in a field:

  • the website field on your GitHub organization and the product's repositories
  • your own social and maker profiles
  • your other products and your personal site, if the product is relevant to their readers
  • package pages (npm, PyPI, crates.io) that have a homepage field

Some of these links carry nofollow or ugc. They give people a path to the site, and you can add them without waiting for a publisher. And as you build more products, you have more of these that you own and control.

Step 2: A handful of launch listings that fit (ASAP)

For me, launch directories have been the quickest way to get the first outside links to a new product. They are usually indexed fairly quickly. I wrote up the nine launch directories that worked for me. You don't have to submit to everyone, but it hardly hurts and it's made much easier with coding agents now.

Things I've seen while listing LinkBunny in November 2026:

  • Some free listings require a badge or link back on your site. One free tier asked for a small nofollow link in our footer. Decide up front which badges you are willing to keep, or if you want to add badges at all.
  • Some free queues are long. One listing I submitted in early November scheduled the launch for February 2027. At this point you might not even know if our product will still be around.
  • The link you get varies. On November 17, the HiCyou listing was index, follow with a nofollow link carrying a ?ref=hicyou parameter. The InventList listing linked with ugc nofollow noopener on a page marked noindex. Check what each listing gives you after it goes live instead of relying on the directory's description.

Skip directories that sell "DR boosts" or guaranteed rankings. Google's link spam policy lists buying links for ranking purposes as link spam. (Interestingly, LinkBunny is a link exchange site for coding agents. But we only exchange nofollow links and site owners always have the choice of whether to link or not. This fits well within Google's policies.)

Step 3: A page/tool someone would link to (ASAP)

Listings say the product exists. A resource gives another writer a reason to link to you in the middle of their own article. Pick one you can finish in a couple of hours or days:

  • an integration or setup guide that works end to end
  • a free tool that does one job without requiring a signup
  • a small dataset or benchmark you can explain and update

For LinkBunny, the free Placement Checker checks whether a page's HTML contains an expected link and reports its anchor and rel values. It works without an account, even if you never use the core LinkBunny exchange network/feature.

Step 4: Ecosystems the product plugs into (ASAP)

If the product integrates with a platform, that platform often has a marketplace, registry, showcase or "awesome" list. These listings reach people who already use the platform.

Only list where the integration exists and works. Read the contribution or submission rules first; many curated GitHub lists want a specific format and reject promotional descriptions.

For example, for LinkBunny, I build integrations (typically MCPs) for various coding agents like Codex and Claude Code and submit to their directories.

Step 5: Contextual links from relevant sites (ongoing)

This is where the links that read like recommendations come from: a mention inside an article that helps the same readers. There are two ways I go about it.

The first is editorial outreach to named sites. Build a short list of pages where a link to you would help their readers, then pitch a specific contribution. How to build a backlink prospect list with your coding agent walks through the research part. You can even automate this with coding agents.

The second is an exchange network. LinkBunny lets site owners use their own coding agent, such as Claude Code or Codex, to exchange nofollow backlinks autonomously. Each site is reviewed before it can be matched, every placement has to be contextual and nofollow, and the agent asks you before it checks in for a request or publishes a placement. This pairs well with the first approach because you can grow nofollow links more aggressively and autonomously and use the first approach for more targeted outreach.

Step 6: Verify and log every link

Check the published page even after you get a submission receipt or a "done" email. Open each source page, check the link points to the right URL, and record its rel values. The free Placement Checker does this for one expected link without signup.

Keep one file with every attempt, including the ones that went nowhere. A listing that was rejected or is still pending in three months is useful information when you decide where to spend time next.

Check back often. Coding agents are great at such verification and tracking tasks.

Copy the plan prompt

Run this with your coding agent in your site's repository. It drafts the plan and a tracker; it doesn't submit or send anything.

Read this site's repository and public pages. Draft a backlink plan in docs/first-backlinks.md.

Order the plan in these steps and keep each step short:
1. Links I control: list every profile, repository and package field that should point to the site, with the URL to use.
2. Launch directories: list directories whose readers would use this product. For each, note its submission rules, cost, badge or link-back requirements and review time, from the directory's own pages. Mark anything you could not confirm as "unverified".
3. A page or tool someone would link to: propose 2 options I could finish in a few hours or days, based on what the product already does.
4. Ecosystems: list platforms this code integrates with and where each one lists integrations (marketplace, registry, directory or curated list). Only include integrations that exist in the code.
5. Contextual links: leave this as a heading. I'll build the prospect list separately.

Then add a tracker table with these columns:
date, step, source page, target URL, status (planned/submitted/pending/live/rejected), rel values, notes

Do not submit forms, create accounts, send email or change site code. List anything that would need my approval at the end.

Later, ask the agent to recheck every submitted and pending row and update the status and rel values in the tracker.

What I don't expect from the first month

Some listings will still be in a queue a month later. I look at which pages link to the product and whether anyone clicks through. Indexing, rankings, traffic and AI citations don't follow a fixed timeline.

DIRECT LINE TO THE FOUNDER

Questions about fit or setup?

I’m Hwee-Boon, the founder. Ask about LinkBunny, site quality, or setup, and I’ll reply by email.

0 / 2,000⌘/Ctrl + Enter to send