I’m writing this as a plan for me to follow as I’ve decided on a slightly reckless “business” idea and wanted to share it with you. Case studies are always useful, and I know a lot of would-be bloggers or even business websites are thinking “should I use AI to write articles for me?”
How to write blog articles with AI
Ok it’s stupid simple but there are tools you can use that are free or cheap. I have a big list of Ai writing tools but you don’t need that yet. Stick around to hear my audacious plan (and then read the results!)
Most AI writing software is some versions of GPT3, openai’s model. It’s cheap but the price goes up as the quality of the tool improves; but it’s at Davinci3 now which is about 2cents for a 1000 words, so you can write 1000 articles for $20. Insane.
But now there’s also ChatGPT, which is another version with a better user interface that allows for “threads” which are important: because there’s a limit on characters per session you’re basically starting over all the time or you can’t train the AI on specific data.
With ChatGPT not only can I teach it to write blog posts in my style, with my voice (and avoid all those “Is this AI” checkers which don’t really work anyway); but I can continue long threads or instantly revise, by just typing “can you try that again but make it…” You just keep nudging it to get what you want.
But there are other tools that are better or easier. Jasper is designed for high quality content and the interface is great; I heard recently that chatGPT is already way better than what Jasper is using (because it has 2 million users and is improving faster than the regular GPT3 – though the chat version is also getting hit with lots of restrictions about it’s going to be a pricier model.
ChatGPT is *better* than human
People keep saying AI writing has no soul, or it can’t be truly creative. But that’s not been my experience at all. I just rewrote a whole website in a day, about 50 articles. Previously they were badly spun, barely readable, almost gibberish (I bought the site to rebuild it).
With the right prompts, I got a consistent narrator that was cocky and snarky, which meant quite a bit of humor – seriously he’s much funnier than me, great comedic timing – and also some pretty amazing metaphors. Every single creative metaphor would take me minutes to think up, which would have slowed down the whole project.
There’s already a million free seo tools websites out there. If I just describe each tool, it wouldn’t be unique content, it would just like all the others. So instead I went hard on “humanizing” my voice, even in the preview snippets. The site doesn’t get traffic yet, but the content is already much better, even by just adding more subheadings with keywords. But I’m hoping the fresh, creative voice actually makes the content worth reading. It made me chuckle a few times and I rarely ever do that.
So, have all these associations and metaphors been used before, by somebody, somewhere? All these jokes? You can say chatGPT doesn’t really understand things and is just putting them together. Personally I don’t get why that matters. It’s the difference between a writer who giggles at his own jokes and a writer who just keep producing good content with intent. Maybe after 50 articles, it’s not fun anymore.
We think we prioritize creativity, but we don’t. Most of the internet is either information or entertainment, and it’s very hard to sell entertainment. Entertaining writing can be a BONUS if the core content is good. Otherwise, it’s like those long stories before cooking recipes… everybody hates those, because even though they are personal, creative, vulnerable and human we also know that *human* content is just there to slow us down and keep us away from what we actual want.
I would argue that chatGPT is very entertaining, objectively, to real humans, so it doesn’t matter that he can’t, himself, *actually* get the joke – though that’s a distinction we are forcing on him… because I could talk to him casually as this persona and it would be hard to tell it wasn’t a human (people say they can just feel it, that soulless AI writing. I think that’s already been proven false a few thousand times, people just haven’t figured it out yet).
How to start a blog in a day
This isn’t actually about that… so I won’t go into hosting or wordpress themes or SEO. Maybe, like me, you’ve already got a blog and are just thinking of using AI writing to get more traffic or finish projects. Maybe you want more exposure, a bigger platform, or more awareness for your work. Or maybe you just want to make a bunch of cash and that’s fine too.
The most important thing probably, is the idea or the vision, and the “brandable” name people can say and share easily. You don’t always want a domain name of YOUR name; you probably want to focus on the topic or subject or theme. So a LOT of the work could be research for a great domain. But also, I like to just buy nice expired domains with relevant history. It may not help much, but I think it does – an established brand has a lot of backlinks. Google knows it exists (I’ll share a hack below on how to get Google to rank your site fast).
It won’t work automatically, Google will be skeptical, but if you post similar, good content and get some backlinks, they’ll send a little traffic. If people stick around on your site, they’ll send more traffic.
Go big or go home
I started this blog about ten years ago, didn’t post anything for a year or two, and since then publish maybe 20 times a year. But I have done long, detailed walkthroughs and articles mostly to help artists and authors sell more work. So I get some traffic, but it doesn’t really earn anything.
But I’ve had ideas for businesses for a long time, and recently bought a bunch of domains and built wordpress sites on them, for a bunch of topics that interest me (cats, tattoos, weed, dinosaurs). I also have a few book review sites with almost no traffic.
So I was already playing with the idea of hiring writers and paying them a low rate to crank out articles; but then I found services even cheaper on Fiverr, to deliver a bundle of chatGPT articles at crazy prices. The thing is, with the right prompt, they’re already very good, I just don’t want to sit there and do everything manually if I can outsource it, I have too many blogs already, so I’d just avoid it (the best creative projects aren’t profitable right away, so they aren’t *worth* it for anyone to do, unless they can invest the time and resources). Previously, I was just not capable, even of finishing my own beloved creative projects – with chatGPT, maybe it’s not perfect, maybe it’s not my best work, but it’s something that gets done and serves the public by being insanely fun and useful. And if it doesn’t work, no big deal, these are just side projects (given, however, done with intense and extreme intention: I’m not just making random crappy hobby sites; I’m branding things that could be huge and trying to make the best stuff people have ever seen… or at least longer, more-indepth, more useful and better written.
As far as I can tell, I can give specific instructions, and a batch list of 100 articles to write (which I can generate fast in chatGPT, I’ll show you later), and it can be automated and even published to a WordPress blog. There are some WP plugins to use GPT4 straight into your dashboard already, and those would be useful but I haven’t actually found anything that works better than what I can do inside of chatGPT, so I’m using that (chatGPT4, custom voice and style prompts, custom outlines).
My plan to triple website traffic in one month
Ok, it’s a little crazy, but I’ve been at around 35K/month traffic for years, and I’d love to get to 100K/month. I’m going to list a specific action plan down below, and then I’ll go through it, do each one, and keep things updated. And I’d like to do it in 30 days or less.
Is that POSSIBLE?! Who knows. But I have a medium sized platform and I like to run fun and interesting challenges.
Will Google penalize my site for AI traffic?
What they’ve already said, is that they’re open to tolerating AI generated blog content, but only if it’s valuable. The thing is, AI can already right really good, informative articles. Depending on the topic, you may need to check the facts or figures, because it lies (a lot). But the writing quality is good and you can choose any style!
I think Google may flag me if I suddenly start publishing a ton of articles; but maybe not. Several big websites just admitted to doing it for months. It’s a little risky, which is why on *this* website I’ll do it all myself manually, from scratch, (and show you how to do it!) – but I’ll also check the content.
Creating a content strategy
The trick is to make a content MAP first. For example, I have a blog for my fiction writing that doesn’t get much traffic. But since all my novels are based on mythology, I decided it was a good place to post “mythology” content.
1. I googled “mythology” and some common keywords and saw what sites popped up.
2. I put those sites into the Hoth (free tool) to see what those sites are ranking for, and made a list of all those keywords.
3. I put that list into this free “bulk keyword checker” tool, and it organized everything based on search volume.
So now I know “Cleopatra, Prometheus and Athena” get half a million views; Greek Gods are most popular; and individual gods and goddesses get the most search volume. This is typical: more people search for specific things they need more info on; rarely do people search for the very broad stuff.
What I should probably do, is also figure out the competition – there are tools for that. But my blog is already fairly established with a decent domain rank already. I plan to go through my list, focusing on the highest volume stuff first.
But before I start…
First I’ll probably make some listicle roundups where I’m linking to credible sources, influencers and established blogs in my niche or subject. I want to show Google I’m well-informed and pay attention (you can actually write these with chatGPT and it’ll even often put in all the right links correctly, but you need to check because sometimes it randomly puts in fake links and references.
I’m basically signaling to Google where my site belongs, and I’m fishing for backlinks (some of those blogs and websites may notice I linked to them or shared them, and they might link back because I featured them and said nice stuff about them; this works especially well if you add a perfect snippet or review or testimonial, and make some custom graphics that make them look awesome. Provide more value.
THEN: I’ll build a content map; starting with broad posts I can put in my menu (best Greek/Roman/Norse/Celtic mythology… listicles with rough details, in my top menu under “myths and legends” so it’s a subcategory or “SEO silo” – then each of those posts that mention all the individual gods and goddesses, will link to the specific resource as I build them.
Now that I know what I’m doing, I posted the first 10 today and could finish the other 100 in another week or so. But wait, what’s the point?! The problem with building a content blog is, any earnings will be measured against active search intent.
I have a friend with a personal blog where he’s tried to juggle a bunch of random topics, he has good content, but the audience is too mixed and unfocused, or it’s not in the intent-purchase stage. He makes some money with ads, about $200/month for 20,000 traffic I think, which isn’t bad – personally I don’t like ads and would rather offer a digital course or product, but it depends on the kind of site you’re building and your monetization strategies.
People searching for myths and legends, probably won’t buy anything… UNLESS I have something relevant to offer, like amazing art or printable posters of classic gods and goddesses; or my actual YA novels based on Greek mythology. That’s a lot of work to sell books (low cost product) but if done well, it’s free marketing and publicity.
Successful blog in a day?
So – that’s the process. Let’s get back on topic. This site (creativindie) already gets traffic and has a decent domain authority, so my articles tend to rank pretty well for some topics that aren’t oversaturated. But it’s hard to write about “creativity” because it’s too general and it’s not what people search for. This blog has mostly been a personal blog, where I randomly share stuff I’m excited about, but I’d like to focus more on actively helping people, which I can do with more traffic.
I also have several blogs with zero traffic and half-installed crappy wordpress themes that I’ve been sitting on for years. I’ll be building those aggressively as well, because they aren’t really my brand and they’re just fun experiments (that doesn’t mean I’ll be publishing crap, remember, Google wants good, useful content and not random AI summaries of everything else on the internet.)
PS – not absolutely sure this works, but I *think* you can get Google to trust you a little better if you run some low-cost ads to your content for a few weeks (you’re driving the initial traffic; but Google can measure how long they stay on your site – then they have data to decide what to do with you). I also think if you have ads on the display network, which means a bunch of sites are “pointing” back to you, it should help boost SEO or train Google. This seems to be how I got a completely blocked, spammy site to get indexed by Google again… the trick is you need something for sale, so for an example, a smart thing would be a review of something in your niche or subject that you can earn an affiliate commission for. Write a long, detailed review, and then advertise that post with Google. Tons of people search for software like Quillbot (specific name products, not general); but most of that traffic is just trying to go to the Quillbot website – not read a review. Still, if your link shows up first and they click it, you’ll get some traffic, and it might convert just enough to pay for the ads.
THE ACTION PLAN
Here’s what I did tonight. I started 3 new profiles for writers on this site. I made fake photos that look like real people in Midjourney and fake bios for them. I’m not hiding it (I’m announcing it here for example) but by using pennames, I can be less precious with the content. There are a lot of things I wouldn’t want to write about as “me”; when you build a personal blog it’s like you’re “on” all the time and it can be exhausting, trying to be real and personal, but actually people just want real useful content and quick solutions.
3 writers x 100 articles…
300 articles in 30 days, is about 30 articles a day (ten for each of my new writers). Ten articles will probably take an hour, so maybe 3 hours of work a day. I’ll make pictures in midjourney, and I’ll generate all the content with chatGPT3. I’ll record this whole process as I do it.
1000 AI articles from Fivver…
These are for those other sites I told you about. I’ve already done a lot of the keyword research, it’s just manual labor at this point, so I’ll outsource it. I’ll do 4 or 5 sites and split up the articles between them, probably hire different people so I can build them all at the same time.
EDIT: nope I didn’t do that, I’ve just gotten used to using chatGPT – I could use the openAI playground and pay much less, but I prefer working in chatGPT.
How to train AI to write in your own voice!
This part is cool; in chatGPT I can add in a few of my articles, tell it to learn my writing style, and then use it to generate more content. There are some specific prompts you should use so I’ll make a video and post it here when I do it. The sneaky thing is, you can also write in anybody else’s style. For example, my 3 fake personas on Creativindie (my highly respected, well-paid trained writing assistant robots) will each be given a unique personality and writing style, that fits their particular area of expertise.
I’ll do something similar for all my other sites, a fake persona, personality and writing style. I want to make the content unique (and helpful) and that starts with specific prompts so your blog content isn’t just the same as everybody else’s. I can, for examle, tell chatGPT to pretend they’re an anime fan in Tokyo obsessed with otaku culture; or a jaded barista grad student in the New York literati; or a burned out stoner slash Cali-boy surfer… and then I give a detailed outline for the article I want and tell it to start with a funny or charming first person anecdote.
What topics to write about (SEO & Traffic)
Here’s the fun part, we’re going to use chatGPT for all the SEO research as well. There are a bunch of free and paid SEO tools I use, and I’ll probably use those a bit also. Basically I’ve already googled lists of relevant keywords, see what sites kept showing up, then see which of their blog articles are bringing in all that traffic.
I want to take the top 3 articles on google for my keyword, list ALL of their headings, merge them into a list and remove duplicates, rewrite them a bit with a killer headline, then use AI to write all those heading topics. But that’s too much work and I’m lazy. So instead, I’ll ask chatGPT something like “imagine you’re an SEO expert copywriter with deep knowledge of Google rankings, seo algorithms, and website traffic. Give me a list of the highest-traffic keywords and articles I should write about for more traffic, related to this topic/subject” I’d use prompts like that until I get a big list (though I might cheat a bit with other tools to confirm).
Then I’d confirm his role or persona. I could even add a style like: “imagine you’re an SEO expert copywriter with deep knowledge of Google rankings, seo algorithms, and website traffic. You’re a bit of a slob but casual and authentic, smart and personable with a dark sarcastic edge. I want you to write an article about “XXX” but make it the absolute best article with perfect SEO that will rank in Google quickly for the keyword “XXX”” – or I could paste a particular article with the writing style I want it to emulate.
Sometimes it helps to repeat yourself in prompts. I’ll test prompts like that until I get one that works pretty well. I’ll try to limit it, to confirm that it understands the directions, before actually writing the full article. I might ask it to come up with a killer hook or headline first; then I’d ask for a list of all the questions, key points or keywords need to be in the list of heading topics. Then I’d ask it to write the article; then I’d ask to revise it in a different style.
My workflow isn’t fleshed out yet, I’ll make a video when I’m actively doing this. Also, chatGPT has custom prompts now, which means you can add a bonus prompt field that applies to everything in the whole thread, so it’s more likely to remember and stick on task and keep the style (this is what makes Sudowrite so good for writing books).
Predictions: honestly no idea… but I’m pretty certain I can write better than average, interesting copy about interesting topics, to beef up this site and a bunch of smaller side-project blogs. All the small blogs have high domain authority, so all I really need is a bunch of “base” content so Google kind of knows what my site is about, and then a handful of “money” articles which link back to a few pages here…
Before you start linking to yourself though, I like to do those “outreach posts” first linking to everyone else in my field, topic or subject. Establish myself as credible, reliable, professional with quality content – then use those posts to boost things here.
SEO and Backlinks
Real quick: search engines analyze the links pointing at you to determine your site’s value – it’s the only metric at all, but it’s still a good benchmark and it is super important. You can make great content but it still might not get seen unless you’re actively reaching out and getting people to share it. That’s tedious, so I haven’t done much of that. However recently a bunch of big sites have started reaching out and asking for backlinks, which means:
- big companies are actively hiring staff to do this kind of outreach
- I can ask for backlinks in return, without putting in much work.
How to actually make money with a blog?
You might be asking, why do all this work? Why waste all this effort? You’re not really adding value, you’re just spinning out more recycle content like everybody else, the internet is an infinite copy of itself until humanity has lost all truth and meaning.
Yes, I feel like that, definitely. Most of this content, I don’t care about. I’m not excited about. But I do have things I really care about, that I built with love and passion, that genuinely help people… and I want more people to discover them.
Someone fancy (and rich) said your earnings reflect the amount of value to the amount of people, or more specifically, the difficulty of the problem you solve and the number of people who need a solution.
“Blogging” is usually a form of entertainment, where you be cool and interesting, but for introverts who don’t like talking about themselves to strangers on the internet (like me) it seems pointless to crank out AI generated blog content, especially if anybody can do it (and they will!)
But blogging doesn’t have to be your passion, and in fact, being passionate about blogging can be a huge waste of time and money, if you don’t focus on the #1 thing: helping people achieve what they want to do. It’s not always about you. Branding and relationships are great after you’ve provided value and as a way to win a new lifetime customer, if you have something to sell.
Most bloggers or potential small business owners don’t really know what they want to sell (or even what value they provide!) but luckily there are lots of ways to make money on the internet. I haven’t talked about blogging or business for a long time, but “AI blog writing” is still generally new. There’s actually not a lot of search volume traffic for chatGPT yet because it’s a brand new thing; people are only vaguely, generally search for relevant terms but nobody’s really sure how to use it yet or if it’s even OK to use!
So, by taking a big risk and putting out a bunch of content around this (and using AI tools on my old and new blogs) I might have the benefit of ranking well when a lot more people are searching for things. People who want to start a blog. Who need hosting, websites, branding, SEO help, even article writing… all things I could do, or recommend to affiliates. I mostly talk about writing and I’m really excited to share some new writing tools; and I talk about book design and publishing and have tools for that too.
I can either get more affiliates to promote my products or I can get more traffic to show up in results; the more people I help, the more things they might buy (things I like, trust and use). Traffic = income, even if I’m just running ads – something I don’t like and will probably never do on this website. I prefer making my online courses, getting lots of email signups with free books and things, trying to get some sales in my email funnel.
It works a little bit, not great. But if I triple my traffic, without doing anything else, I’ll triple my income too (something I desperately need to do, because my income and traffic has tanked in the last few years, probably because I never update my blog.
But I’m actually more excited about the new, faceless blogs where I can just focus on great content and not building a personal brand; I can write lists of best AI writing tools and earn commissions; maybe I’ll put a blogging course together or an old book project I never finished (but have a cover for; I’ll probably have AI write most of that too.)
CASE STUDIES & RESULTS
I’m not actually sure if I’m going to post this before or after I finish; so maybe I’ve already done this stuff (if so, you’ll see more details and videos). I’d like to keep a daily record of the thing I do every day for 30 days, but let’s be honest, I’ll probably procrastinate and do it all over a binge weekend at the end of the month).
Here’s the cheatsheet of steps involved.
1. buy a domain and set up wordpress (done).
2. if necessary, create fake people or profiles. You can be honest and say it’s your chatbot assistant if you want (done)
3. set up a basic about page with your brand statement and purpose, choose a great hook or tagline (I’ll probably use chatGPT for help with this; and actually I’ll probably do this at the end, after I’ve generated my list of articles, I’ll go back and clean up the style before it starts getting traffic but after I know exactly what all the topics are
4. SEO research: figure out what your main topics are, and the 100 most popular articles that 5 of your biggest competitors are ranking for (or trust chatGPT to do an OK job at this).
5. Have AI rewrite the headers and make them snazzy; then use AI to generate heading lists for each topic, then write each topic. These will take awhile. Revise the style a bit (or set up a default tone and writing style); post them all to your blog. SurferSEO makes this a lot easier if you want to get it just right; but I also think there’s a disadvantage to just using tools to make a “perfect” post… as opposed to a long, messy, disorganized, sprawling but authentic post like this.
6. Add nice images (midjourneyAI)… there are WP plugins that pull from openAI and Dalle2 but the quality isn’t there yet. Tag your images with seo keywords in the title and description. Use 5 images per post (not necessary but let’s be thorough, it’ll keep people on the page longer).
7. create interlinks (there are plugins for this), and a “silo” menu of topics, so your menu reflects your main categories. Try to have less than 10 categories and a clean, simple menu; organize all your blog content with these categories and make them easy to find. You might want to set this up before you post all the content; but probably choose topics after you’ve done your keyword research.
This is all stuff I’ll do myself, manually, for this website, so I’ll be a little more controlling about quality and probably add some unique introductions or conclusions – even a touch of wacky eccentricity – so Google can’t quite figure out whether it’s real or not (though, I think if I train it in MY writing style, which is random, endless, “look-a-squirrel!” sentences, tell it to repeat itself and change its mind halfway through ideas…. what I’m saying is, I think AI could probably already write like I write, but maybe it shouldn’t. Maybe I should just let it write good, clean articles and keep the weird stuff to myself.
8. For the OTHER sites, it depends on what I can hire out, and what cost and what they’re willing to do. It’s a lot of work if I want them to do it all well, but it’s probably worth it: everybody is going to be pumping AI garbage now and it’ll be really hard for Google to catch everything. The sites that have traffic, will probably keep the traffic. Unless you have something cool that draws attention…
9 *Mostly* I’ll write those 100/1000(?) AI generated posts, and make them good and long and useful. Maybe they won’t have a “soul” but they will be clear and helpful. But then I’ll write some fun, weird stuff. I may make very specific, slightly ridiculous things; things that people would share or enjoy. I may also add some simple tools (you can actually build a simple website tool, like a calculator of some kind, in chatGPT, so maybe I’ll play with that or hire something out. Tools work great for traffic.
10. and then mainly, if those blogs get traffic, I’ll sell products, and if they don’t, a few nice backlinks to some of my *main* content on this site, or a few of my other sites for book design templates, for example, they’ll push my rank up a bit in Google. I’m on page 3 right now for a lot of things (around position #25) which means almost no traffic. If I can push any page up to the top ten, it could be a lot of traffic, depending on who is searching.
The crazy thing is right now, the term “ai writer” gets about 12K hits a month. That’s nothing. All my writing site competitors are getting 60K traffic for random erotic literature about giantess vore (don’t look it up). But it’s growing quickly. ChatGPT went from zero to 2 million users in a month, it’s (I think) the fastest growing company anyway. Because it’s amazing.
PS. I was already ranking for “free chatGPT” and “free AI image generator” – and getting traffic, which shows you how early we are… but I was using a simple WordPress plugin and it was costing $100 a day just to run the free tools – and of course that specific audience isn’t interested in buying paid AI tools (duh!). All traffic is good traffic… but not if it’s losing money.
With AI writing tools, everyone is skeptical at first, then amazed. I was out with a developer I told about this stuff months ago, he just started playing with it this week and his mind is blown at the possibilities. It allows everybody to do so much more, to get so much more work done with less effort. Of course it’s a problem, because all that high-pay, high-skill work you used to do can now (or soon) be outsourced to anybody with internet.
But creative people, with a vision, with a realistic and profitable business plan, have a big advantage!
*OK sidenote tangent, but far too many creative people start exciting projects with no idea how they will actually make money, and then they spend years “working on their business” and nothing happens. It’s exhausting.
I was a starving artist for a decade. I’ve paid my dues. But I figured out creativity is a gift to be shared, not an indulgence to be nurtured privately. If you want to make a difference, add value, use your skills to help others – or just help yourself and your family by generating extra income with a successful blog (a rare thing!) – blogging might be for you, and this might be the time. There’s a slim window of opportunity, before all the big players get on board, where small, nimble, self-employed businesses can leap ahead of the competition in ways that were never possible before a month or two ago.
Everything’s going to be shaken up for good. You’ll either get buried or rise to the top.
PS. I’ll probably post this when I have the first few videos up, but I don’t promise it’ll be all clean and polished. It’s mostly a roadmap for myself. But if you’re new here, you can sign up on my list and I’ll let you know when it’s all done (and the results! though I think it’ll probably take six months to actually triple my traffic (and income), I plan to triple my content in 30 days and see what happens).
DAY ONE:
I got started testing prompts with a big, long article. I do want 100+ long-tail keywords for everything related to my niche but I also want to rank well for the big, obvious keywords. There’s not actually a ton of competition for my target keyword (book cover design) and I already have some blogs that rank well-ish. So I asked chatGPT to write me a detailed outline for a 5000 word article.
It refused, saying it can’t write that long. So I had to prompt it by breaking up the outline into sections saying “write this section with these headers; then write this section with these headers. It took about an hour probably. I got 6146 words and it’s probably the best, longest article on the internet. But it’s rough and needs polishing; there’s a lot of repletion, and it definitely “sound like AI” (formal, a lot of in conclusion or moreover…) I could run it through some style rephrasing tools but that doesn’t really make sense for a lengthy how-to article.
I *will* go through and revise, edit, add a ton of pictures, add a ton of resources and links, and write in some real human content. For a big, main anchor post like this, I want it to be awesome so I’ll spend more time. Then at the bottom I’ll break it all down into genres “how to design a scifi book cover” etc and post about 20 of those with links to individual articles. I’ll use a bunch of cover design examples. On this particular site, I probably won’t do a hundred things, but I want to rank for that one keyword because I’ll be rebuilding the site with some great cover design tools and launching soon, and I’ve never really added good content to it.
I also made a huge, huge course outline for AI writing and AI art tools, and hundreds of really specific, creative and cool midjourney prompts I can use on multiple websites.
DAY TWO:
I got distracted but then suddenly invigorated again: I realized I could use new AI voices to record all my audiobooks and AI writing to finish all my fiction. But that led me to some realizations about how I can use AI to write and publish short stories and poetry on my website(s) that would be narrated by AI voices.
That’s a huge project but also a huge traffic driver.
My *goal* is to test WordPress plugins soon, because I think there’s a way to generate 100 short story ideas, and put them in a plugin to auto-generate a new short story every day, (and then add art and audio manually… there are some that can generate an image with AI but they’re not great because the use dalle2.)
*** got distracted again and decided to buy a dinosaur website. But that became a dinosaur + pirate website. More on that soon…***
PICKING A DOMAIN NAME
This is actually super important so I wanted to let you in one my process. I’ve already done this for most of my other sites, but: I was vaguely interested in a dinosaur site that was already getting 20K/month traffic, but it sold for $5000 and I’m not that interested. There’s another also for $5000 for no traffic. But I did a little research anyway to see what I could find.
I’m looking for clear, short, brandable names with “dinosaur” (main keyword) towards the front. I did the SEO research, “dinosaur” gets about 2.2 million searches a month and all the popular dinosaurs get an average of 100K each (at least 20 of them).
But then I saw “realdinosaurpictures.com” which is fun/surprising, and I *could* make amazing dinosaur photography in midjourney that’s better than anything out there. 30K people search for “dinosaur pictures” and that’s a direct match. I can add dinosaur art/cartoon/comic/cute or how to draw…. all those get traffic and are related to dinosaurpictures. So that’s still a good idea, and a fun name.
But then I found “dinosaurpirate.com” – it’s a little riskier, but also more brandable. This is always the decision you will have to make: in anything you do. The creative, fun thing, or the extremely clear, obvious thing. A homepage with great dinosaur pictures is high value in itself; but it would basically be like a wikipedia with better dinosaur pics.
Dinosaurpirate is more fun but less obvious: “pirates” only get about 100K search traffic, so not worth it on its own, at all. But it gives me the chance to add pictures of pirates riding dinosaurs (something more surprising and unexpected; delightful fantasy art; maybe even a book.)
So then I thought: I can do all the “real dinosaur pictures” on the homepage anyway, and I can have long articles with all the facts and information…but I can also have a story. I can AI-generate a running episode about an actual shipwreck survivor who discovered “the REAL Jurassic park” and then show his field notes, his photography, and his adventures. It’ll be like Lord of the Flies meets Crusoe… with treasure hunting. I’ll get it AI narrated; it could become a podcast and a book series. And it won’t be just another site about Dinosaurs, it’ll be something new and intriguing.
Plus, because pirates and dinosaurs is a fun, fantasy combination, it means I can post about other random popular things like dragons and unicorns (real pictures of mythical beasts!) and it won’t be weird.
So that’s the plan: it’ll take more work and branding and design, but I enjoy that stuff; but then mostly the process will be the same – find the top 100 traffic seo/keywords related to the topic, make the “best articles on the internet” with the “best pictures on the internet” and better, more enjoyable content than anyone else (in a kid-friendly voice; out of those 2.2 million searches, 1/4th of that is “dinosaurs for kids.”
Update: I still plan to do this but got sidetracked.
DAY… 17?
Oops I lost track, got distracted finishing a novel and suddenly got my mojo back for some reason. I used chatGPT to rewrite 50 core articles for one of my new sites (he absolutely nailed the style, so now I’m excited again about some bigger projects.
I was looking for a perfect chatGPT wordpress plugin. I tested a few but they weren’t great. I found one that *worked* but after doing a few articles, I realized they definitely sounded like chatGPT (without prompts, it has a style that’s somewhere between condescending and grammar geek). I do plan to test a few more tools, but I’ve been putting if off because it’s just one thing to learn.
I wanted:
- Autopost hundreds of articles from a list of keywords
- But also add unique custom styles and tone of voice
- And also add an outline/template/format
- Schedule it to post 3 a day, so it’s not all at once and looks organic.
But I’ve gotten pretty good at using chatGPT, with a sharp concise prompt where I just switch out the main keyword or topic. So, for example… I have a book review site and want to review the top 100 books of all time (put my phd in literature to good use), and maybe another 1000 books. It would be great if that was automatic, and I could enter my keyword list and be done in 30 minutes. That’s probably possible if I find the right tool. But – I don’t really mind just sitting here listening to music and “writing 1000 articles.” I can even do it on my phone while watching Netflix. I’ve been doing maybe 30 an hour, so about a week’s worth of work to have a totally finished new blog with 1000 articles. There are people you can hire on Fiverr who will do that for a thousand bucks, but I wouldn’t trust the voice, tone, quality or format. These have to be the best thing on the internet.
I can use the same formula to some less important, smaller projects like product reviews or information (my other sites are on cats, weed, desks, anime, tech, books… and a couple others). I want great content, the best content, but I also want it relatively fast, because I have a very slim advantage if I get started first and go farthest.
Won’t everyone just write the same content?
No – because truly creative content is risky, and most people won’t use the right prompts are make sure the tone is really great. Also, what I imagine is that SEO scammer gurus or clickfarms will focus on high-pay affiliate stuff or Amazon shopping for junk. I’m going to tackle things that aren’t that sexy, like books. The literary and artistic circles I’m in, will probably be the last to embrace AI. For my part, I get to produce great content that simply was never worth producing before, because it’s too much work with no pay, but I can do it anyway (the gift of being able to do things because they matter, or they interest me, or just for fun, instead of doing things only to make money).
UPDATE: earlier I said I wanted to get to 100K/month in 30 days… I blew past that with only one article on midjourney prompts, so that’s boring now. Let’s go for 500K! I’m going to stick with my plan of 300 new articles (which is crazy), but maybe I’ll set them up to publish slowly, one a day for a year instead of all at once? Eh, maybe not I’m impatient. Also that idea about 3 different writers? Not sure I’ll do that, too much work and I’m lazy too. Is it more disingenuous to make names and faces for fake robot people; or to put AI writing under my own name and profile?
I actually just wrote a long movie review of The Lesson which is all about literary theft and the inherent violence of the creative process.
PS – you know how to tell Google you’re not just posting AI content? With a big, messy, rough, unfinished blog post like this that has a ton of content but also mirrors the wild and lawless landscape in my creative brain. More entertaining? Maybe. More useful, engaging, practical, helpful? Probably not! I’ll bet all the AI content, in fact, is both more entertaining and more useful… so is there more *value* to this post because it is 100% human? I don’t see it (unless, paradoxically, the humanizing element is for the google robots so they don’t think I’m just a spam farm; and the AI content is actually the good stuff humans like)… in which case we will have all become performatively human.
DAY… 18
I started building content and backlinks from some of the sites I bought a year ago, that has only had terrible ipsum lorem dummy text or badly spun articles. I deleted some stuff or rewrote the content. Now typically, I should build a LOT of content and link about to a LOT of similar blogs in the same field, to build a kind of web so google knows where I belong. Then once I get some traffic, I could link back to my own articles. But I rushed it a bit. Traffic seemed to “freeze” or even go down a little for a few days which is worrying; my guess is Google saw the new links and wants to analyze them to see if it’s spam or if I’m buying links or something. Hopefully as I improve the quality of those sites, the links will at least not hurt rankings and might boost them a bit.
Right now I’m building up content for a gaming girl site (for girl gamers), and chatGPT has been really awesome. The only problem is, the content is a little older so it can’t give me “news” or links to new products. So it’s not really great for a review site that focuses on latest releases. It still does a pretty great job of tone of voice and writing style (seriously, it’s amazing – it’s so high quality) and it can also provide real links to relevant things on the internet. It’s just not very fresh or new.
What I really want is a chatGPT that fully connects to the internet; then it’s game over. I tried Bing but it sucks; it only “steals” examples from the top ranking articles and summarizes them into a new list. I tried Google’s Bard and it’s better; it can give me more recent resources but it does NOT cite the sources… so I’m not sure which is actually worse. But I’ve been trying to use them to get more recent content and then having chatGPT rewrite it in our style and voice for this project.
For now, chatGPT and I have built a rhythm, where she (yes I switched the gender; she is large, she contains multitudes) writes content and then adds a “link to an interesting article” but there’s no real link; so I just have to do a google search and replace that link with a real (relevant, newer) link. It works.
Pinterest: also, I decided, even though my new content is GREAT I need more; I made a ton of midjourney aesthetic artwork with the goal of going viral on Pinterest, to get ton of shares and backlinks. The hypothesis is, even though Google may not love my AI generated content, I’m very confident that actual humans will! I already see this on facebook; one post warning “how to recognize AI art” and another just an image with a thousand gushing comments, about how beautiful and amazing and SKILLED the artist who made it is.
Critics are vocal and outspoken; but regular people just love great content. I mentioned this earlier, but images can be the thing itself – also this month I made some posts about icons or emojis; I’ll be making sites about tattoos and dinosaurs – because there’s huge value in dumping 500 awesome images in a huge gallery when people are searching for those kind of pictures. I can afford to make them AND give them away for free. I can create a branded style (for this blog for example, I generated a few dozen similar images for featured images, and just pick one out for each new post.)
So I need to set up an inspo gallery with a good pinterest plugin to make this happen, and use good keywords and tagging and things (I’m not a pinterest expert). I could even test some pinterest ads just to get the ball rolling and hopefully signal to Google to send some traffic my way.
I’ve lost count of the days…
It’s probably a month later. So my experiment has basically “failed” but only because I haven’t done the work; I’ve only recently started aggressively posting new content on some of these side-hustle blog ideas. I think the whole internet is on lockdown right now with an absolute ton of AI generated spam sites; I’m trying to make mine *better* quality with editorial oversight and in-depth prompts, plus amazing midjourney artwork.
But I’m also feeling discouraged; I should have focused on all my sites that already get traffic but need fixing and improving. These brand new sites – even after I put up 50+ articles or so, are getting basically zero traffic. BUT partly that’s just like a new book; just because you publish doesn’t mean anyone will find it. You need to drive traffic first so Google knows you exist and can measure user engagement.
I should run a contest or even some light google ads just to get the ball rolling, but first I have a rather ambitious plan for all of these sites which include several hundred posts. Using chatGPT I’m limited even on the paid account, because I only get 25 request per hour or something lame; when I’m stuck I switch into the openAI playground but it’s less enjoyable to use and I don’t trust it for long-form content (where it keeps the same style, voice and outline and I just give it new topics.)
I haven’t failed because I haven’t finished… and doing *something* with these sites that I’ve been sitting on for a year feels good because I like accomplishing things, it’s very visually satisfying to see the number of posts grow especially when your whole site finally looks good with great pictures and content.
There’s a lot I can do with the marketing/promotion, but I’m still in the content creation phase. I wish it could be rushed; I haven’t found a bulk-automatic wordpress blog generator that works as well as doing it manually; I’ve thought about hiring someone to do this click-button basic stuff but oh well – it’s really only a week or two of full-time work for myself and then it’s done.
Ultimately even if these sites don’t get traffic, the plan was to build them up with lots of relevant outlinks so google won’t notice or mind a few links back to my own main sites and pages… but that’s a TON of work for a couple backlinks and probably not worth it on its own.
Mostly, I love the idea of one of these sites getting a ton of traffic and making money, so I can stop doing serious things like book design and publishing and tell people my “job” is just a blog about anime, dinosaurs, cats, tattoos and weed. But obviously there’s a ton of competition and I need a unique advantage or twist (OR – for example with dinosaurs – I don’t think there will be enough money involved in some big traffic sites for most internet marketers to risk it, so I can do some things just for fun).
Final Update
If this is your first visit to my blog, I apologize… or maybe I don’t. I’m kind of proud of being messy. This could have been a simple, clean, clickbaity listicle like all the other trash on the internet, with no real value. This post has *dozens* of content marketing and SEO strategies.
As I mentioned earlier, most of my new sites (that I’m trying to recover and rebuild) aren’t getting any traffic – but after abandoning them for years, I can hardly except to post a ton of content and rank them instantly. On this site however, my most trusted platform, out of my top ten best-performing posts, fully half of the traffic drivers (5 out of 10) were pretty much made by AI in about an hour; compared to the other five that are ancient, rough and messy. And that’s not even a fair comparison – because I have hundreds of personal blog posts I’ve written over the past decade, and maybe 10 or 20 articles I made recently with chatGPT – so percentage-wise – the AI posts are driving a huge portion of my traffic.
There is some effort involved: I’m manually choosing fun and interesting content intended to surprise and delight – but I’m not really wasting time trying to research and write it all myself. The impulse will be, to suggest it’s better to take time and do 1 solid human article, because it’s worth 20 spammy AI written articles. That’s a humanistic, ethical skepticism towards artificial intelligence, and it has merit (your fears are justified!) But the evidence suggest that 1 high quality GPT written article is actually better content than the 20 random blog posts I wrote personally and passionately.
I mentioned earlier, right now I’m focused on backlinking, because some big names have been reaching out and I’m happy to take advantage of the opportunity. I started working on that “mythology” site I mentioned earlier because I’m yearning to get back to my fiction writing, which I adore – and I’m pretty close to finishing some series and making it profitable (but I’ve been *really close* for 5 years…) – but also because I’m procrastinating from my actual task:
I have a newly finished online cover design tool, on another site that gets traffic, but I need to make 100+ cover design templates for it to be a really solid offer. That was last month’s project… but now it’s this month’s. I’ve been working on this for literally a decade without ever finishing, but I tend to procrastinate super-hard when I’m within an inch of actual completion. So I’m wrestling with that and making slow progress.
But I’ll keep doing all these things, all these strategies to boost traffic.
And then, crucially, I’ll rebuild my funnels, autoresponders, landing page offers and checkout/thank you pages, figure out some kind of affiliate system, and split test everything until it’s making a good profit (I’ll write more posts like this detailing the process and letting you know which tools I chose to use). The good news is I already get good traffic and I’ll be expecting a lot more soon with these new backlinks; the bad news is it’s converting pretty poorly – but that should be a fixable issue, it’ll just have to be my focus for awhile.
So: my big goals are to *finish* this DIY book design package so I can launch it profitably and set it up to succeed; and then focusing more on content, fixing up all my pages and articles, creating lots of long tail keyword content, until I’m getting all the traffic and making all the money (in that very specific micro-niche).
Once I get my passive income streams back up, I’ll focus on fiction for most of this fall and next year (I need to finish about 5 novels to complete 5 trilogies (I have a bad habit of starting new projects and not finishing). But I’ll also, hopefully, be growing all these new sites and my *main* sites to keep boosting rank and traffic.
It’s a lot of work for a one-man, small online enterprise, and it can be overwhelming… but I also kind of love it. If it sounds exhausting though and you already have an online business, and just need a content writer to do SEO research, craft a marketing strategy, build backlinks and create content for you – just get in touch!
I’m a philosophy dropout with a PhD in Literature. I covet a cabin full of cats, where I can write fantasy novels to pay for my cake addiction. Sometimes I live in castles.
Add Comment