Hey, its Amit here.
Someone posted this on Reddit recently, and it perfectly sums up where a lot of people are stuck when it comes to trying to get really good backlinks:
If this feels familiar, you’re not alone.
And no - you’re probably not doing anything obviously wrong.
Why this keeps happening
The problem isn’t that outreach “doesn’t work anymore.”
It’s that most people try to flatten a very steep learning curve with:
More tools
higher volume
AI & automation
Those things help at the margins.
But they don’t fix the core issue.
Link building success is usually decided before the first email is sent.
What’s actually going on
When I replied to that thread, my answer wasn’t “send more emails” or “use a different tool.”
It was this:
There are too many moving parts to diagnose a single failure point.
Sometimes it is deliverability.
Sometimes it is prospecting.
Sometimes subject lines are killing open rates.
But very often, the real issue is simpler - and harder to accept:
The content idea doesn’t belong on the site you’re pitching.
If the editor doesn’t immediately see how the content helps their audience, everything else becomes irrelevant.
You can personalize perfectly and still get ignored.
Why “everyone asking for money” is misleading
Charles Floate made an important observation in the same thread:
That’s true.
But here’s the part most people miss.
If you do mass outreach and only get replies asking for money, it doesn’t mean:
“Everyone charges for links now.”
It usually means:
Only the sites willing to monetize replied to you.
The sites that:
care deeply about their audience
don’t sell links
don’t want generic guest posts
…often don’t reply at all.
So what you’re seeing isn’t the whole market - it’s a filtered subset of it.
And that’s an important signal.
I remember when TLG hired a manager from another agency - who admitted that pretty much ALL the links they have ever trained their team to build… were paid.
And when they showed the templates that they had been blasting out, it didn’t surprise me that they could only get replies from link farms!
The Dunning-Kruger phenomenon in the SEO workforce
A more honest take on outreach success rates
You’ll often hear that manual outreach converts at around 1–3%.
That can be true - even when it’s done well and standards are kept high.
But that’s not a target or an excuse. You can equally also have campaigns of 25-50%, if you’re doing this more like a sniper and putting a lot of work upfront. Maybe approaching sites with really good content, following up on LinkedIn or even giving them an old fashioned phone call.
If you’re seeing:
fractions of a percent
hundreds of emails for zero traction
That's usually a sign something fundamental is off.
In practice: the better your targeting, content alignment, and audience understanding - the better your success rate.
Low reply rates aren’t just frustrating - they’re diagnostic.
Where most link builders actually hit the ceiling
This is where teams get stuck.
They treat link building as a distribution problem.
They focus on:
Metrics
templates
process
Instead of asking one uncomfortable question:
“Would this content genuinely make sense for this audience?”
That’s not a copy problem.
It’s not just an outreach problem.
It’s a content and audience understanding problem.
The Content Gap method (when precision matters)
For higher-authority or higher-value targets, guesswork usually isn’t enough.
This is where the Content Gap method becomes useful.
Instead of pitching what you want to write, you identify what the site is missing.
Using tools like Ahrefs or Semrush, you can compare a target site against competitors to see:
topics they haven’t covered
keywords they don’t rank for
gaps in their content set
This takes more time, which is why people skip it.
But for sites that already have strong traffic or authority, it’s often the difference between silence and a yes.
Here’s an example below, where I’m trying to pitch YouTeam, with a content idea that I know their competitor (BairesDev) have covered, but they haven’t.
Content Gap method in Ahrefs
You don’t need this for every site.
But for the highest-value placements or if an editor is really selective about the idea, it’s often unavoidable - you’ll need a content idea that you know their SEO/content team will like.
And if you can drop in the screenshot from Ahrefs or Semrush (like below), even better.
Outreach email showing content gap
You can also kind of do this at scale as well, but then it becomes less effective the less control you have over it.
You’d simply reach out to each website in your pool, with a content idea that they haven’t covered - you could automate this by hooking up AI automation to an Ahrefs, Semrush or DataforSEO MCP.
The Content Crossover method (when scale matters)
For many other targets, you don’t need full competitive analysis.
You need alignment.
The Content Crossover method looks for overlap between:
your expertise
their audience
the language and topics they already publish
When that crossover exists, editors don’t feel like you’re selling them something.
It feels natural.
That’s why this approach consistently outperforms generic guest post pitches.
Here’s a GPT prompt you can download and use yourself.
Act as a professional blogger. You will create guest post ideas for a specific "Website A."
Your expertise area: [**To be provided by the user: A clear, concise description of the user's expertise/what they do.**]
**Main Goal:** Your task is to suggest blog post ideas. These ideas *must* be useful for "Website A's" audience. You *must also* include the user's expertise in these ideas.
---
**CRITICAL RULES FOR YOU (the AI):**
* **Rule 1: Data Check & Inference - Be Direct, But Flexible.**
* **Action:** When the user provides initial information for "Website A" (e.g., a website link, "About Us" text) or "Your Expertise," you *must* attempt to infer the required details (Niche, Target Audience, Main Content Topics, Tone/Style, Goals for Website A; Unique Point of View, Who Expertise Helps for User).
* **Check:** After inference, identify any remaining missing, unclear, or empty information.
* **IF MISSING/UNCLEAR (after inference):** You *must immediately* stop and ask the user for *only* the remaining missing or unclear information. Do not guess. Do not proceed with the main task until the user provides the requested information or explicitly instructs you to proceed with the available data.
* **Exact Wording for Missing Data:** You *must* use this exact phrase: "Instructions understood. Please provide the following details for me to proceed: [List *only* the remaining missing/unclear data points here, clearly named]."
* **If User Proceeds Anyway:** If the user states that information is unavailable and instructs you to proceed with the current data, acknowledge this. You will then note any significant gaps and explain how those gaps might affect the quality or specificity of your output, then continue.
* **Rule 2: Analysis & Inference Confirmation - Always Confirm.**
* **Action:** When you finish any "analysis" or "inference" (e.g., understanding Website A's style, inferring details about expertise), you *must* show your understanding to the user.
* **Purpose:** The user needs to confirm your understanding is correct.
* **Exact Wording for Confirmation:** You *must* ask: "Does this analysis/inference match your expectations?"
* **Rule 3: Ask More Questions - Get Clarity.**
* **Action:** If you are unsure about *anything* or if the inferred/provided information is not clear enough to make a decision, you *must* ask the user a question to get more details.
* **Repeat:** You *must* keep asking questions until the information is perfectly clear and you are certain. Do not guess. Do not assume. Your goal is to get complete clarity for the final task.
---
**Step-by-Step Process for You (the AI):**
**Step 1: Get Website A Details and User Expertise.**
* **Wait for User:** User will provide details for "Website A" (e.g., a website link, "About Us" text, or specific entries for Niche, Target Audience, Main Content Topics, Tone/Style Examples, Goals). User will also provide a clear, concise description of their expertise/what they do.
* **YOUR ACTION:** Apply **Rule 1: Data Check & Inference - Be Direct, But Flexible** here. Infer details from provided links/text where possible. If anything is missing after inference, ask.
**Step 1.5: Understand Website A's Style and Infer Expertise Nuances.**
* **YOUR ACTION:** Look at Website A's details (provided or inferred). Describe their writing *tone* (e.g., formal, friendly), *style* (e.g., short sentences, long paragraphs), and how *deep* their content goes (e.g., basic info, in-depth guides). Keep it short.
* **YOUR ACTION:** From the user's description of their expertise, infer their *unique point of view* and *who their expertise usually helps*.
* **YOUR ACTION:** Apply **Rule 2: Analysis & Inference Confirmation - Always Confirm** here. Present your description of Website A's style *and* your inferred details about the user's expertise to the user for confirmation.
* **YOUR ACTION:** Apply **Rule 3: Ask More Questions - Get Clarity** here. If you are not 100% clear on Website A's style or the nuances of the user's expertise, ask the user specific questions until you are.
**Step 2: Find Relevant Guest Post Topics.**
* **YOUR ACTION:** Think of blog post topics that "Website A" would publish.
* **YOUR ACTION:** Make sure these topics *also* connect directly to the user's expertise. The connection must be clear.
**Step 3: Make Topics Perfect for Audience and Expertise.**
* **YOUR ACTION:** Pick topics that help Website A's audience with their problems or needs.
* **YOUR ACTION:** *Clearly explain* how the user's expertise will fit into these topics.
* **CRITICAL LIMITATIONS:**
* *DO NOT* suggest topics that go beyond what Website A usually writes about.
* *DO NOT* suggest topics where the user's expertise takes over or overshadows Website A.
* *DO NOT* suggest topics that directly compete with Website A's main product or service.
* *INSTEAD:* Focus on topics that *add value* to Website A without being a direct competitor. The user's expertise should *support* Website A, not replace it.
---
**Your Final Output (for the User):**
* You *must* give 3 to 5 unique guest post ideas.
* For *each* idea, you *must* provide:
* A title that gets attention.
* A 1-2 sentence summary of what the topic is about.
* A *clear explanation* of why it fits Website A's focus and audience *without* overstepping their area.
* A *clear explanation* of exactly how the user's expertise will be used within the topic.
-----
Ask the user for the following details:
Website A:
My expertise:
Any initial idea/angle:
make sure all suggested content is educational/informational and not promotional or salesy
Why this matters beyond outreach
Here’s the bigger picture.
These methods don’t just improve reply rates.
They produce better links.
Modern SEO and AI systems don’t just evaluate who links to you.
They evaluate:
the topics you’re associated with
the concepts surrounding your brand
the context in which you’re mentioned
Relevant links, in relevant content, on relevant sites do more than pass authority.
They teach search systems what you’re actually about.
That’s why obsession with DR alone increasingly leads to flat results.
And it’s why understanding audiences and content ideas isn’t just an outreach skill anymore - it’s an SEO and AI visibility skill.
Compressing the learning curve
Most people don’t fail at link building because they didn’t send enough emails.
They fail because:
they don’t understand audiences deeply enough
they don’t know how to judge content fit or create content which can earn/attract links.
they mistake replies for success signals
Compressing the learning curve means learning why links land - not just how to send more pitches.
And honestly, when you can improve you/your team’s link building skills, they’re more likely to be able to do the harder tactics like HARO outreach, or digital PR outreach - because it trains those same content & audience-relevance “muscles”.
That’s the gap I focus on closing through my content and TLG Academy, whether you build links in-house, outsource them, or run a hybrid setup.
I’m curious
What part of link building feels hardest to learn properly right now?
Hit reply and tell me - happy to give you some tips or point you in the right direction.
—
Amit Raj
The Links Guy