How To Start a Blog Without Writing: The Jarvis AI + Surfer SEO Method

What if you could write a full blog post on a topic you know absolutely nothing about, in under 20 minutes, without typing more than a handful of words yourself? That is exactly what Alston did in this video, and it is not a trick or a highlight reel. He walked through the entire process on camera, from picking a random niche on Amazon all the way to a Surfer SEO content score of 96 out of 100, with Jarvis doing the writing the whole time.

This is the Jarvis Boss Mode method. Conversion.ai, now branded as Jarvis, rolled out Boss Mode with a critical upgrade: the AI can now look back 2,000 characters into your document, up from the old 600-character limit. That change keeps the writing coherent across longer pieces. Paired with Surfer SEO and Ahrefs for keyword targeting, and Grammarly for the final cleanup, this four-tool stack can take a brand-new blog from zero to 30 posts faster than most people write one article by hand.

What You’ll Walk Out With

A niche-finding method using Amazon product categories that takes about five minutes
How to use Ahrefs to find keywords with difficulty scores of 0 to 5 that a new blog can actually rank for
A Surfer SEO workflow that hands you ready-made H2 and H3 headings before you write a word
How Jarvis Boss Mode generates 1,600 words while you watch, using a 2,000-character memory window
The Grammarly and plagiarism-check final pass before you hit publish
A domain, hosting, WordPress, and theme checklist for anyone starting from scratch
How to figure out which online income model actually fits your situation at finder.platformproof.com

What Jarvis Boss Mode Actually Changed

Before Boss Mode, Jarvis had a memory problem. After you clicked “Compose,” it could only look back 600 characters to decide what to write next. That meant a 1,500-word post could drift off topic halfway through, repeating itself or pulling in content that had nothing to do with the heading above it. Boss Mode extended that memory window to 2,000 characters and introduced a long-form assistant interface where you build your outline first, then let the AI fill each section in sequence.

The practical result: Alston typed in a few headings pulled from Surfer SEO, hit compose, and watched 800 words appear with almost no drift. He hit compose again and pushed past 1,000 words. By the time Grammarly ran its final pass, the post was at 1,600 words and holding a Surfer SEO content score of 96 out of 100. The topic was how to refill a barbecue lighter. Alston said plainly on camera that he did not know the first thing about it.

Step 1: Pick a Niche Using Amazon

The first tool Alston opens is Amazon. Not to buy anything, but to scan product categories for niche ideas. He goes into “See all departments” and scrolls the list. Kitchen and Dining? Storage and Organization? Plenty of options. What pops into his head while he is scrolling? Grills. That is the whole selection process. He picks grills because the idea shows up first and feels reasonable.

The logic behind Amazon as a starting point is straightforward. When someone clicks through your blog post to Amazon and makes a purchase, you earn a commission on everything they buy in that session, not just the product you linked to. A visitor searching for a barbecue lighter might end up buying toilet paper, coffee, and a set of spatulas in the same cart. Those small commissions stack up. That is why low-ticket niches like grilling accessories can still be worth building around, especially when your goal is traffic from a long list of low-competition keywords rather than a single high-commission product.

Step 2: Keyword Research With Ahrefs

Once you have a niche idea, Ahrefs Keyword Explorer is where you validate it. Alston types “outdoor grill” into the search bar and gets a keyword difficulty score of 58 out of 100. Ahrefs labels that “hard.” For a brand-new site with no backlinks and no domain history, competing for a keyword at 58 is not realistic. So he digs into the matching terms and filters down to anything with a difficulty of 5 or below.

Here is what he finds in that low-difficulty range within the grilling niche:

Outdoor grill table: keyword difficulty of 0
Best smoker grill combo: keyword difficulty of 4, roughly 1,700 searches per month
Best burger press: keyword difficulty of 0, roughly 500 searches per month
Barbecue lighter: keyword difficulty of 0, roughly 1,100 searches per month

The goal at this stage is to collect 30 keywords that all sit at difficulty 5 or below. That becomes your publishing queue for the first phase of the blog. You are not chasing big-volume keywords on day one. You are going after specific, overlooked searches where the existing results are thin and the competing sites have weak domain authority. Every one of those 30 keywords is a post. Every post is a shot at passive traffic.

Step 3: Competitor Keyword Research in Ahrefs Site Explorer

One of the sharper moves in this workflow is using Ahrefs to pull keyword lists from smaller competitors. Alston finds a site already ranking for low-difficulty grill keywords, checks its domain rating in Ahrefs, and confirms it is a weaker site. The competitor he looks at is pulling over a million visitors a month but has a domain rating that suggests a new site could compete with it. The key metric he is scanning for: sites with an Ahrefs rank of one million or higher, which signals weaker authority.

He pastes that competitor’s URL into Ahrefs Site Explorer, clicks “Organic Keywords,” and filters to keyword difficulty of 5 or less. That single move surfaces a list of content gaps: topics the competitor is already getting traffic for, where a new site could realistically show up. This research approach originally came from a blogger named Marcus, known as the Affiliate Marketing Dude. The tip about looking for sites with high Ahrefs rank came from the Caffeinated Blogger. Alston credits both of them directly in the video.

While scanning those competitor keywords, Alston also checks the word count of the pages already ranking. The top result for barbecue lighter he pulls up is close to 3,500 words. His rule: if the top-ranking page is over 2,000 words, he passes on that keyword and moves to a shorter-format opportunity. Writing a post that competes with 3,500 words, even with Jarvis doing the work, takes more effort than the modest traffic is worth when you are at ground zero. He looks for keywords where the top result is 2,000 words or less.

Step 4: Domain Name, Hosting, WordPress, and Theme

With a keyword list in hand, you need somewhere to publish. Alston keeps the setup checklist short.

Domain name: He uses instantdomainsearch.com to check availability. He types in “grillman” and the tool immediately shows every variation that is still open: grillman.shop, web.grillman, grillman.design, and others. The obvious .com is taken, as expected. The variations usually land under 15 dollars. His advice is to pick something related to your niche and not overthink it. The domain name is less important than publishing your first post.

Web hosting: Alston mentions GoDaddy and Bluehost as examples. Any reliable shared hosting plan works at this stage. A new blog will not generate enough traffic to strain basic shared hosting for months, so price is the more relevant factor than server specs. Buy a plan, connect your domain, and move on.

WordPress: This is the content management layer that sits on top of your hosting and controls how your posts look and how your site is organized. It is free to install and the most widely supported blogging platform available. WordPress also handles SEO basics well when paired with plugins like Yoast or Rank Math. Almost every established blogging tutorial defaults to WordPress for a reason.

Theme: Alston recommends ThemeForest for premium themes. The most important filter is speed. A bloated theme with excessive animations, built-in sliders, and bundled plugins slows your site load time, and Google uses page speed as a ranking signal. When you are trying to rank on free organic traffic alone, a slow site is a real liability. He says to search for “blog” on ThemeForest and look for themes described as lightweight. Before purchasing, paste the theme demo URL into Google PageSpeed Insights and check the mobile score. Anything above 80 is workable. Below 60 is a warning sign.

Step 5: Build Your Content Outline in Surfer SEO

This is where the workflow starts to feel like a shortcut that actually works. Alston pastes his target keyword, “how to refill a barbecue lighter,” into Surfer SEO’s content editor. Surfer spends a couple of minutes scanning the top-ranking pages for that keyword and then returns a recommended structure: target word count, H1 options, H2 subheadings, H3 sub-points, and a list of target keywords that need to appear somewhere in the finished article.

The heading suggestions Surfer surfaces for this keyword include topics like:

How do you know if a lighter is refillable?
Bleed your lighter
Purchase a can of quality butane
How do I refill a long lighter?

Alston admits on camera that he does not know what “bleed your lighter” means. That is the point of the workflow. He does not need to know. He copies the headings from Surfer’s Brief tab and pastes them into Jarvis as an outline. The AI handles the explanation. Surfer also displays a live content score as the article fills in: keywords that appear turn green, keywords close to the target turn orange, keywords not yet mentioned stay red. The score climbs as Jarvis writes.

Step 6: Let Jarvis Boss Mode Write the Post

Inside Jarvis’s long form assistant, Alston pastes the Surfer headings, marks them with double pound signs to signal heading level to Jarvis, and clicks “Compose.” The AI starts generating text immediately. After the first compose run, the post sits at close to 800 words and the Surfer score is at 61 out of 100. He clicks compose again. The post passes 1,000 words. He repeats the process, watching the word count and Surfer score climb together with each round. By the end, the post is at 1,600 words with a Surfer score of 96.

One thing Alston flags honestly: Jarvis does not always deliver clean output on the first pass. Sometimes it restarts a section mid-paragraph, repeats a phrase, or takes a tone that does not quite fit the rest of the article. That is normal. The editing work required to clean up a Jarvis draft is still a fraction of the work required to write the same content from scratch, particularly on a topic you have never touched. The ratio Alston puts on it is about 95 percent hands off, with some minor cleanup at the end.

One thing worth noting is that Boss Mode’s extended 2,000-character memory is what makes longer posts stay on track. The older 600-character limit was enough to finish a single paragraph but not enough to remember the introduction by the time you were halfway through the body. The upgrade is a meaningful one for anyone trying to generate consistent 1,500-plus-word articles.

Step 7: Grammarly Review and Plagiarism Check

Once Jarvis finishes the draft, the next tool up is Grammarly. Alston runs it as a Chrome extension directly inside the Jarvis editor. He is not reading every sentence carefully at this point. He is accepting Grammarly’s suggestions and moving through the document quickly. Grammarly flags grammar issues, word choice problems, and readability concerns. There is a free version that catches grammar basics and a premium version that goes deeper on tone and clarity. Either one is better than skipping this pass entirely.

The step Alston emphasizes most before publishing: run the finished article through a plagiarism checker. In earlier videos covering conversion.ai, some viewers raised concerns that Jarvis was generating content that closely resembled existing articles online. He does not dismiss that concern. AI writing tools work by learning patterns from published content, and those patterns can sometimes produce output that sits too close to existing sources. His approach is to confirm the content is unique before uploading, rather than discovering the problem after the post is live and indexed.

Once the article clears the plagiarism check and Grammarly is done, the post is ready to publish. Remove the pound-sign heading markers used inside Jarvis, paste the content into WordPress, drop in your Amazon affiliate links at the relevant product mentions, and hit publish.

Not sure which online income model fits your situation?

Answer a few quick questions and get a personalized recommendation at finder.platformproof.com.

Honest Drawbacks of the AI Blog Method

This workflow is genuinely fast. It is also not a guaranteed shortcut to search traffic. Here is what Alston does not oversell.

Google knows AI content exists at scale. At the time of this video, Alston noted that hundreds if not thousands of people were already using Jarvis Boss Mode to generate content in bulk. A Surfer SEO score of 96 means the article is structurally similar to what ranks. It does not mean Google will automatically reward it. Structure is a starting point, not a guarantee.
The plagiarism risk is real. Jarvis generates content based on patterns in published web writing. Even when the output is not technically copied word for word, the sentence patterns and phrasing can resemble existing articles closely enough to trigger a plagiarism checker. Running that check before publishing is not optional.
Low-difficulty keywords attract modest traffic. A keyword with difficulty 0 and 500 monthly searches is easy to rank for because the search demand is small. A barbecue lighter article might generate 200 visitors a month at peak. The Amazon commission on a lighter is under two dollars. Volume is the only way this model generates meaningful income, which is why Alston starts at 30 posts, not one.
Results take months, not days. Alston planned to check results at 60 days and expects a fuller picture closer to six months. Brand-new domains typically sit in a search sandbox period for the first three to four months, where even well-optimized posts barely show up in results. This method requires patience alongside the tool stack.

The 30-Post Experiment: What Alston Is Testing

Alston closes the video with a clear experiment plan. He is going to build a real blog in a niche he has not revealed publicly, write 30 posts using this exact four-tool workflow, add Amazon affiliate links throughout, and report back on the results. The signal he is watching for: a meaningful rise in Amazon affiliate earnings from that specific site. If the commission numbers climb two to four months in, the method is working. If they stay flat, the experiment will tell him that too.

He is transparent that this is not an affiliate promotion for Jarvis. He says directly on camera that he is not getting paid to recommend it. He just wants to find out whether AI-assisted content can rank and convert under real conditions, and he plans to share whatever he finds. That kind of public, data-driven test is worth tracking whether you are considering this approach yourself or simply want to see where AI content production is headed in practice.

The Four-Tool Stack: What Each One Does

Here is how the tools divide the work in this workflow:

Ahrefs: Keyword research and competitor analysis. Surfaces low-difficulty, reasonable-volume targets. Identifies which competitor pages rank for what so you can fill the gaps with your own content.
Surfer SEO: Content structure and optimization scoring. Recommends word count, H2 and H3 headings, and target keywords. Gives a live 0-to-100 score as the article is written, updating in real time as keywords are hit.
Jarvis (Conversion.ai) Boss Mode: The actual content writer. Takes the outline, fills each section using a 2,000-character memory window, and generates 1,600 words in roughly 20 minutes with minimal human input.
Grammarly: The final cleanup pass. Catches grammar errors, awkward phrasing, and readability issues after Jarvis finishes the draft. Available as a Chrome extension in both free and paid versions.

Find Your X

The AI blog method works best when it matches the right person in the right situation. If you are already a fast, confident writer, this workflow saves less time than it would for someone who freezes in front of a blank document. If you are entering a niche where hands-on experience and authority matter, purely AI-generated content may need more human review before it earns trust from readers and search engines. The best tool is the one that fits your actual starting point. Find out what that is for you at finder.platformproof.com.

Frequently Asked Questions

Is Jarvis (Conversion.ai) still available today?

Conversion.ai rebranded as Jarvis and later rebranded again as Jasper AI. The tool is still active and widely used for AI content generation. Boss Mode, the feature Alston demonstrates in this video, is now part of Jasper’s subscription tiers under updated branding. Check Jasper’s current website for pricing and feature availability, as both have changed since this video was recorded in 2021.

Do I need all four tools to start with this method?

No. Ahrefs is the most expensive tool in the stack, and there are free or cheaper alternatives. Ubersuggest and Google Keyword Planner can surface keyword difficulty data at no cost, though with less precision than Ahrefs. Surfer SEO has alternatives like Frase that come in at a lower price point. Grammarly has a free tier that handles basic grammar. The core workflow, building an outline before writing and using AI to fill it in, works at lower cost. What you give up is precision in the keyword targeting and content scoring steps.

Will AI-generated content rank on Google?

Alston was running a live test to find out when this video was made in 2021. Google’s stated position has evolved since: the search engine says it evaluates content on quality and helpfulness, not on how it was produced. In practice, AI content that is accurate, well-structured, and genuinely useful to the reader can rank. Content that is repetitive, vague, or inaccurate tends not to hold rankings. The Surfer SEO score helps with structure. The plagiarism check helps with uniqueness. Human review of the draft before publishing helps with accuracy.

What keyword difficulty score should a new blog target?

Alston’s filter in Ahrefs is difficulty 5 or below. That is a tight ceiling. Search volumes in that difficulty range are usually modest, running from a few hundred to a couple thousand searches per month per keyword. The upside is that a brand-new domain can realistically rank for those keywords within a few months, rather than spending a year trying to catch up to sites with thousands of backlinks. You are trading top-of-funnel reach for the ability to actually appear in results early on.

How long should each blog post be?

Alston’s rule is to match or beat the word count of the top-ranking page, but only if that count is 2,000 words or under. If the top result is 3,500 words, he skips the keyword entirely and finds a shorter-format opportunity instead. The reasoning is simple: generating a 3,500-word post that needs to beat a well-established article on a high-authority site is more work than the traffic payoff justifies, especially when low-difficulty alternatives exist nearby. The Jarvis-generated post in this video lands at 1,600 words, which is typical for the keyword targets he selects.

How do I pick a WordPress theme for a content blog?

Alston recommends ThemeForest for premium theme options. The main filter is load speed. A theme packed with animations, built-in sliders, and pre-installed plugins will slow your site, and Google actively uses page speed as a ranking signal. Before buying any theme, paste the demo URL into Google PageSpeed Insights and look at the mobile score. Aim for above 80. Anything below 60 is worth avoiding. Lightweight, clean themes built specifically for text-heavy blogs tend to score better and require less performance optimization after installation.

How do I check whether Jarvis content is plagiarized?

Tools like Copyscape, Quetext, or Grammarly’s plagiarism checker (available in the premium plan) can scan your draft against published web content. Paste the finished article in and review the results before uploading to WordPress. If you see a high similarity match to any single source, rewrite that section manually or use Jarvis to generate an alternative version of the flagged paragraph before publishing. This step is worth doing every time, not just when you suspect an issue.

How long until this kind of blog generates affiliate income?

Alston’s estimate in the video is that he expects to see initial signals at 60 days, with a fuller picture at six months. New domains typically go through a search visibility crawl period for the first three to four months where even well-optimized posts gain minimal traction. Amazon Associates payments are issued 60 days after the purchase month. Realistically, you are looking at five to seven months from your first published post to your first meaningful commission check, assuming you are publishing consistently and targeting keywords in the 0-to-5 difficulty range.

Sources

Jarvis / Conversion.ai (now Jasper AI): AI writing tool demonstrated in this video
Surfer SEO: content optimization and scoring tool used in the workflow
Ahrefs: keyword research and competitor analysis platform
Grammarly: grammar and readability checker, used as a Chrome extension
Amazon Associates: affiliate program discussed as the monetization model
Marcus Campbell, the Affiliate Marketing Dude: keyword research tip credited in the video
The Caffeinated Blogger: domain authority filtering tip credited in the video
instantdomainsearch.com: domain name availability tool mentioned in the video
ThemeForest: WordPress theme marketplace recommended for lightweight blog themes