Barbehow Case Study 1, Year 2, Quarterly Update 3: Are Niche Websites Still Profitable in the AI Era?
Dim
Welcome back! Did you hear that I started a newsletter called Publetise? Subscribe to get my best strategies and techniques for affiliate site owners, bloggers, and content creators delivered to your inbox every week.
It’s summer.
People all around the world are firing up their grills and cooking delicious meals for their families and friends.
As they’re grilling, they come up with all sorts of questions. And whenever they pull out their phones to search for answers, Barbehow is one of the sites they most frequently come across.
The Progress So Far
When I first started building this site, I told you I’d begin by focusing on:
- Informational content
- With keywords researched in LowFruits
- Monetized with display ads
Once the site had established itself as an authority in the niche and gained ground on Google, the editorial team and I would slowly but surely work our way up to commercial content monetized with affiliate programs.
This is exactly what the team and I did when we published a bunch of commercial articles in Q4 of last year and Q1 of this year, rounding up the best grills, meat smokers, and accessories for different purposes in anticipation of BBQ season.
Our patience has paid off, and LowFruits’ suggestions were once again right on the money.
Today, the commercial articles consistently rank on the first page of search results—and for the second consecutive quarter in a row, the site is earning more revenue from affiliate commissions than display ads.
The site earned $1,847.57 in April, May, and June of this year:
- 59% came from affiliate programs (mostly Amazon Associates)
- 41% came from display ads (exclusively Ezoic Ads; I am a member of the Ezoic Premium program)
The fact that I’ve only published 16 new pieces of content since the beginning of the year, and yet the site has already generated $2,778.53 in revenue in H1 2023, speaks plenty about the passivity of this business model.
What’s Next for the Case Study Site?
I’m split between two options, and I haven’t made up my mind yet.
One, I could sell the site this year and refocus time, talent, and capital on growing a bigger asset in my portfolio.
Or, two, I could continue investing in it and building it up for an even stronger BBQ season and sale next year.
If I decide to keep the site, I’m already thinking about several ways to grow it.
Many of these ways involve experimenting with new supplementary channels and revenue streams, as well as using generative AI in creative ways beyond just content creation.
Until I’ve made up my mind, the focus is on improving the existing content, whether that’s:
- Adding additional sections to articles based on Google Search Console data,
- Creating illustrations that make it easier for readers to digest the information, or
- Tweaking the design and content for Conversion Rate Optimization (CRO)
Either way, I think that, ultimately, selling this site is a conclusive and somewhat poetic ending to a case study gone right.
As you’ll find out in the programmatic SEO case study update later this month, they don’t always go as well as planned.
What do you think?
Hit me up in the comments below.
P.S. If you’re a serious buyer interested in the site, hit me up at [email protected].
Is Building Websites Still a Good Business in the AI Era?
Lately, I’ve been getting a ton of questions from friends and coaching clients about whether building websites is still a good business model on the web that’s increasingly flooded with AI content.
The honest answer is: I don’t know.
(And if anybody claims to know for sure, take what they say with a pinch of salt.)
So let’s talk about what I do know!
Barriers to Entry
Before generative AI, creating a niche website required tech savviness and an investment of a few thousand dollars (or the equivalent amount of time) in content targeting the right keywords. Then you had to wait for 6 to 9 months for that content to mature and start getting organic traffic from Google. Eventually, you’d monetize that traffic with display ads, affiliate programs, and/or selling products.
Thanks to ChatGPT, the cost of content production has dropped to $0 (or if you’re using ChatGPT Plus, like myself, $20 per month). Literally anyone can spin up a site, start prompting, and pump out tens—even hundreds—of articles within a few weeks.
Don’t get me wrong; the content that comes out of any generative AI tool is shit content. It needs editing, fact-checking, and a good amount of formatting to become factual and legible for human readers. If you rely heavily on generative AI, you also need good prompts.
But that doesn’t negate the fact that the barriers to entry into this business have been demolished.
In response, brand-new sites need more time to rank high and receive organic traffic. More time without traffic means more time to see revenue and break even.
Aged domains can help you reduce that time to days, weeks, or months. However, they’ll only get more expensive as SEOs revert to them—and the more people in the industry use them, the more likely search engines are to crack down on them sooner rather than later.
They are a shortcut.
However, shortcuts can, and very often do, backfire.
Generative AI-Powered Search and Search Experiments
Then there’s generative AI-powered search.
As if it wasn’t enough that anyone can flood the web with shit content, search engines have started experimenting with chatbots and generating long-form answers by themselves.
We have yet to see how this experiment turns out.
I’ve seen experiments of the same magnitude fail before.
Many years ago, Google tried to launch a social network called Google+. They aimed to compete with Facebook and Twitter, and it was arguably one of their biggest experiments since they changed search. (For those who remember, Google+ ultimately failed and faded into obscurity.)
Voice search was also supposed to revolutionize search. And yet Siri still struggles to do much more than set an alarm or reminder, and I continue to use my Echo Dot as nothing more than a talkative kitchen timer.
Will the same happen to Bard and/or the Search Generative Experience (SGE)?
If it doesn’t, and these things stick around, how will they affect the traffic that Google is sending to publishers like you and me today?
Will informational queries get more clicks? Will commercial queries get fewer clicks?
How much traffic will we gain and lose?
Once again, no one knows.
Bing claims their Bing Chatbot is actually increasing CTRs to publishers in the SERPs.
However, we’re still in the early days of generative AI and large language models—and everyone, including search engines, is learning and adapting. Within a few years, how we interact with computers “tomorrow” may become vastly different from today.
Or it may stay more or less the same.
Not everyone knows that, in 2012, I built a media company on the Bulgarian web (although I no longer live there, Bulgaria is my home country). My partners and I grew it to a portfolio of several sites, YouTube channels, and a tech-themed show on a music TV channel. I eventually sold my shares and cashed out.
One of the sites, Smart News, is still around—and it looks very much like how I remember it.
Can you imagine that the way we built websites eleven years ago hasn’t changed all that much today?
Sure, the sources of traffic have. Back then, Facebook was king, and Google was secondary.
Traffic Sources
So next, we have the means of getting traffic.
Before Google rolled out the Penguin and Panda updates, SEO was all about buying backlinks to boost PageRank (Google even had a PageRank toolbar) and cramming keywords.
Even so, the business’s core mechanics—finding a low-competition niche, investing in good enough content, monetizing with display ads, affiliate partnerships, and product or service sales—haven’t changed much.
As long as there’s a Web, there will always be ways for publishers and content creators to reach audiences and make money.
And yet there’s something about generative AI that’s giving me a gut feeling that it could be the most significant change I’ve witnessed since I started my first website back in 2007.
So, I’m cautiously optimistic.
Online publishing may undergo a major transformative shift, for better or worse, and the way we currently do things may no longer be the same way we do things in the future.
If you’re already on board, it’s worth sticking around and keeping to what works to see how things unfold. If you’re just entering this business, keep in mind that you’re entering an industry that just found itself in uncharted waters, and the weather’s getting stormy.
My play?
Focus on quality, optimize properly, scale wisely, and don’t keep all your eggs in one basket.
Tell me about yours in the comments below!
Content
Traffic
Ezoic Big Data Analytics
We’re right in the middle of BBQ season, and the seasonality of the BBQ niche is showing in the site’s traffic.
Even though we only published only 1 post during the entire second quarter, the website received an impressive 68,265 visits and 77,674 page views.
That’s a 57% increase in visits compared to the 43,361 visits we had in the first quarter and a 56% increase in page views compared to the 49,556 page views we had during the same period.
| Month | Visits | Page Views |
| April 2023 | 20,285 | 24,313 |
| May 2023 | 23,762 | 26,518 |
| June 2023 | 24,218 | 26,843 |
Google Search
During Q2, the site received 53,200 clicks and had 1.9 million impressions on Google’s search engine results pages. The average click-through rate (CTR) was 2.8%, and the average position in the search results was 14.5.
Just to refresh your memory if you don’t remember the stats from the previous update, in the first quarter of 2023, the site received 31,000 clicks and had 1.29 million impressions in Google’s SERPs. The average CTR during that period was 2.4%, and the average position in the search results was 18.
| Month | Clicks | Impressions | Average CTR | Average Position |
| April 2023 | 15k | 551k | 2.7% | 15.5 |
| May 2023 | 18.4k | 659k | 2.8% | 14 |
| June 2023 | 19.7k | 691k | 2.9% | 14.3 |
Earnings
In Q2, the site earned $1,847,57, of which $751.60 was from display ads (compared to $413.31 in Q1) and $1,095.97 from affiliate programs (compared to $528.07 in Q1).
Display Ads
| Month | Earnings | ePMV |
| April 2023 | $199.93 | $9.86 |
| May 2023 | $260.63 | $10.97 |
| June 2023 | $291.04 | $12.02 |
Affiliate Programs
Amazon Associates
All “best X for Y” articles have buttons pointing to Amazon and other sites, including direct-to-consumer (D2C) brands or alternative retailers. Even so, Amazon continues to be the best-converting affiliate program.
| Month | Earnings | Conversion Rate |
| April 2023 | $203.73 | 8.64% |
| May 2023 | $434.83 | 9.06% |
| June 2023 | $442.04 | 8.59% |
Other Affiliate Programs
| Quarter | Affiliate Program | Earnings |
| Q2 2023 | Impact.com | $15.97 |
Who I Am
Hey!
My name’s Dim.
Thanks for reading (or skimming) this far.
I started my first site in 2007 after I stumbled upon a blog about making money online. I’ve been buying, growing, and selling sites ever since.
These days, I own an indie media company and run an email newsletter for online publishers called “Publetise.”
Don’t be a stranger: Subscribe to my newsletter at Publetise.com and get my best strategies and tactics delivered to your inbox once a week.
Dim, thanks for the continuous updates. It’s been great to follow along. I don’t remember if you ever did an backlinking or if you’re relying solely on low-competution keywords. Also the EPMv seems to be on the lower end to me. Do you typically see higher EPMv on your other sites?
Thanks for following along, John!
I do link building now and then — HARO and traditional; just did an order with Indexsy, which I’ve been working with lately. But it’s definitely not as big of a priority for me as it should be.
The ePMV is low, but it’s aligned to what I’m seeing on another food site I have with Ezoic. It used to be higher, but this year advertising revenue in the food niche hasn’t been as good as it was during the pandemic lockdown.
Dim
Thanks for this piece Dim!
What do you think of travel blogs? The space itself might feel saturated, but I noticed that most blogs are the same, recommending the top 10 activities to do in a city, what to do 3 days in…., etc.
Do you think is still a good idea to invest time and effort in a travel blog if the content is really unique?
Hey, Alex,
I believe there will always be a demand for content related to travel, especially if it addresses specific needs of certain groups of travelers that have not been well served.
I also know Carl Broadbent, who I follow on YouTube, is investing in a travel website (also a case study) and seems to be getting good results.
Just a word of caution based on what you shared: travel blogs tend to produce similar content because that’s what people are searching for. If the topics you have in mind are unique, it might be worth checking the search volumes for them before investing too much in content to target them.
Consider whether they are worth pursuing compared to the more common types of content found on travel blogs—the revenue’s in the numbers (and conversion rates).
Dim
I too have always been disappointed in Siri’s failure to answer basic questions. Perhaps generative AI is better suited for voice search? A conversational search experience.