Is It Keto: Month 11

Highlights 🔗︎

• Two of my blog posts reached the front page of Hacker News.
• I may have finally discovered a way to scale my keto site profitably.
• I’m putting Portfolio Rebalancer on the backburner due to lack of traction.

Goal Grades 🔗︎

At the start of each month, I declare what I’d like to accomplish. Here’s how I did against those goals:

Conduct five customer interviews for the portfolio rebalancer 🔗︎

• Result: Conducted zero customer interviews.
• Grade: F

I deprioritized the portfolio rebalancer in favor of Is It Keto. I’ll explain why below.

Implement customer payments for the portfolio rebalancer 🔗︎

• Result: Implemented a Stripe payment flow.
• Grade: A

I successfully implemented a subscription payment workflow for the first time ever. I expected a simple two-day process, but it took me about three weeks to get everything working.

Publish one new blog post 🔗︎

• Result: Published two new blog posts, both of which reached the front page of Hacker News.
• Grade: A

While integrating with Stripe, I noticed that their JavaScript library collected user data from my site. My blog post about the issue reached #1 on Hacker News and prompted a response from Stripe’s CEO. The Register also interviewed me about the story.

Stripe made several changes the week following my post, and I published a follow-up with my thoughts. That post also reached the front page of Hacker News, though it generated a more muted response.

Stats 🔗︎

Portfolio Rebalancer 🔗︎

For its first full month of existence, Portfolio Rebalancer achieved a healthy 1.1k visitors.

Unfortunately, it ended the month with a revenue of $0 and no signs of increasing.

Is It Keto 🔗︎

Is It Keto took a big hit in March amid COVID-19 panic, but it’s been growing consistently every week since then.

Portfolio Rebalancer has lots of visitors but no sales 🔗︎

Lately, I’ve been thinking about how to align my personal blog with my business projects. I’m often successful at attracting readers to my blog, so it would be great if I could use my blog to get customers interested in some of the products I’ve built. This has been hard because most of the products don’t match up well with the people who read my blog. I suspected that Portfolio Rebalancer would be different because I expect there’s a lot of overlap between software developers and passive investors.

While building Portfolio Rebalancer, I paid special attention to which parts of the process would make an interesting blog post. When I discovered Stripe’s library tracking my users, I knew I had found an interesting that directly related to the process of building my app.

The blog post attracted 28k readers over the past two weeks. Portfolio Rebalancer had ~1k visitors in April, and I suspect around 80% of them are the result of my Stripe blog post. But in the end, none of the new visitors purchased subscriptions.

A few reached the point of signing up for an account, and I reached out to all of them. One user replied that they were interested but felt the price was too high. They said they’d feel comfortable paying for it like a mobile app, so $1.99-$4.99 for lifetime access.

I knew $1.99 was far too low to be viable, but I recognized that $20/month didn’t make much sense. Investors who rebalance typically do it once a quarter or once a year, so it felt irrational to pay for the service monthly. I changed pricing to $99 per year with a 7-day free trial, but that didn’t change anything.

I also tried placing Google Ads, but I’ve gotten 99 paid clicks (totaling $240), and nearly 100% of them bounced immediately:

I haven’t given up on Portfolio Rebalancer entirely, but I’ve placed it on the backburner to free up time for Is It Keto. It was meant to be a two- or three-week prototype, but creating the payment flow took longer than I expected.

Venturing into auto-generated pages 🔗︎

Is It Keto is a site I created in 2018 that answers simple questions around the keto diet. The site and I have an on-again, off-again relationship. My pattern has been to work on the site every few months, get bored, then return again when another project doesn’t work out, so here we are again.

One of the most common questions people ask me about Is It Keto is, “Why don’t you simply pull in a nutrition database and auto-generate pages for every possible food?” I’ve always feared the wrath of Google lest I run afoul of their rules against auto-generated content. Google drives almost all of Is It Keto’s traffic, so if they brought the hammer down on me, the business would vanish overnight.

Recently, I had an illuminating conversation with a friend who’s had significantly more experience than I have in search engine optimization. He suggested that my fears were unfounded. In his experience, Google only brings down harsh penalties on sites that egregiously try to manipulate search results. I might get in trouble if I used machine learning to auto-generate lots of garbage page text, but if I’m simply pulling nutritional data and presenting it in a way that adds value for users on the keto diet, that should be fine by Google.

My friend predicted that if Google didn’t like my auto-generated content, it’s more likely that they’d simply downrank those particular pages rather than applying a penalty to my entire site.

Chasing the long tail 🔗︎

My discovery about auto-generated content got me excited about Is It Keto again. It’s been successful at attracting visitors, but I was never able to find a way to make growth profitable. Hiring writers costs me $50-100 per article, but articles typically earn less than $1/month in revenue, so I could never figure out how to scale the site.

Page generation would be a gamechanger because it substantially lowers my costs. The cost would be so low that I could produce articles in the “long tail” of Google queries that nobody else would bother creating. For example, if you Google “are pringles keto?” none of the results directly address the question.

According to Google’s keyword planner, there are only ~70 searches per month for searches like “are pringles keto?” Whoever took the #1 spot would only get ~50 clicks per month. Is It Keto earns ~$0.01 per visitor, meaning the Pringles article would bring in about $6 per year. It’s hard to justify the time or cost of crafting a dedicated article for so few readers.

The equation changes if I can generate entire batches of articles with roughly the same effort other authors take to generate one article. There are probably 25 different brands and varieties of potato chips that have similar search traffic. If I make a template that covers them all, that’s $150 per year in extra revenue, which is a lot more appealing. I can likely re-use most of the content for chips for 1,000 other snack foods, so maybe that’s an extra $6k year in revenue. And then what happens when I can do that with 10 other food categories?

Other examples of long-tail SEO 🔗︎

Zapier is a successful product that provides plumbing to connect different apps that are not aware of each other. They used a similar SEO strategy to build their business.

Zapier generates a dedicated page for every possible combination of their partner apps, so if you search “the gift goose + connectwise manage,” the top result is Zapier’s auto-generated page:

According to Ahrefs’ backlink checker, there are zero links to that page, yet it’s the #1 search result. Why? Because search volume is so low that nobody else would bother to create a page. But Zapier has templates set up so that it costs them next to nothing to generate a page that perfectly matches the query.

I finally have an unfair advantage 🔗︎

Startup is one of my favorite podcasts, and I often think back to this moment in one of the early episodes where the podcast’s host and founder was seeking investment from Chris Sacca, a well-known venture capital investor. Sacca asks the host:

I want to invest in companies that have an unfair advantage. What’s your unfair advantage?

One of the reasons I keep abandoning Is It Keto is that it’s never played to my strengths. I write well, but too many other keto sites have good writers for me to win on that front. I consider myself a good developer, but software has never played a key role in the site. A non-developer could have produced essentially the same thing with WordPress or Squarespace.

Pivoting to auto-generated pages gives me an advantage over other keto sites because none of them seem to have developers on staff. They probably have developers that help them maintain their WordPress setups or to create one-off tools like a keto calculator. Those things are in a different category of software engineering than what I’d be doing, which is creating and maintaining a data pipeline that runs smoothly for years. And that’s great because that’s the type of development I do best.

Another advantage is that I already have a database of nutritional information. The USDA offers a free, open database of food data, but their search is terrible. A query like “apples” yields 17,000 results, so you have to pick through to find the most appropriate match. I went through this process for Zestful, so I already have coverage for most foods.

What it will look like 🔗︎

I’ve already begun auto-generating content for Is It Keto. I’ve created 13 new articles about fruits I had never covered. Fruits are the easiest articles because they require no analysis of artificial ingredients, so the article is usually a straightforward, “No, it’s not keto-friendly.”

I’ve set it up so that I can improve my templates over time, then simply re-run my pipeline to update all previously-generated articles.

One of the most popular pages on Is It Keto is “Is Russell Stover Sugar-Free Chocolate Candy Keto?” I wrote that article by hand, but re-reading it today, I could largely re-use the content to build a template for other chocolates and candies:

Legacy projects 🔗︎

Here are some brief updates on projects that I still maintain but are not the primary focus of my development:

Zestful 🔗︎

Zestful had an odd jump in traffic because a Chinese blog published an unauthorized translation of my article, “My Second Year as a Solo Developer,” and I guess it gained enough traction that people clicked through to Zestful. It didn’t seem to translate into any sales, though.

Month-over-month sales increased, but Zestful is historically bursty. Anything from $0 to $100 in monthly revenue is normal.

What Got Done 🔗︎

What Got Done is my weekly work journaling app. I tried to build it into a business last year but relegated it to “hobby project” status after it failed to gain traction. I still use it regularly myself, and I sometimes add features on weekends or evenings.

Recently, I was looking at my user dashboard and noticed that the site had 370 registered users. On average, the site gets one new user every day:

The problem is that, of these 370 registered users, the total number who actually use the site is about… five.

Users sign up and then immediately abandon the app. Sometimes they’ll post updates for a week or two, but they almost always end up fading away within weeks. With so many people interested enough to sign up, I feel like my failure to retain them is a huge missed opportunity to grow the site.

I suspect two main causes:

• Lack of onboarding
• Failure to support habit-building

One of the most valuable pieces of feedback my friend DK gave me after testing out my app was that the signup process left him feeling lost. The first thing a user sees after registering is a giant blank textbox and vague instructions about how to fill it. I could certainly do more to help the user create updates in smaller, incremental steps so that it doesn’t feel like such an overwhelming task.

The other problem is that What Got Done generates the most value for people who use it consistently, but it offers no tools to help people build a habit of using the site. I could fix this by introducing options to get email reminders or calendar events. I could also reward users for consistent participation via points or badges.

Interesting discoveries 🔗︎

Ahrefs Academy: Blogging for Business 🔗︎

In response to COVID-19, Ahrefs released their $800 course on content marketing for free. I generally prefer to learn by reading rather than watching online videos, but I’ve been impressed with this series.

One lesson from the video that hit home for me is that traffic is a vanity metric. If you’re blogging as part of your business, the number of visitors you attract is only an intermediate goal to something else, like selling products or acquiring new customers. I experienced exactly this with Portfolio Rebalancer; it was great to attract 1,000 visitors in its first month, but the volume itself isn’t much to celebrate without accompanying revenue or signups.

The series itself is a fantastic example of elegant content marketing. Throughout the course, they’re suggesting strategies of planning your content to attract search traffic, and then they always show Ahrefs tools that can help. They build credibility by sharing useful information; so, when they show you their tool, you trust that it’s helpful.

Companies often get this backward; they start with their own product and show you all the ways it can help you. That turns people off because the company hasn’t yet earned the viewer’s trust, so the content just feels like advertising.

Ahrefs is undecided about how long the course will remain free, so I’ve archived my copy for future reference.

Wrap up 🔗︎

What got done? 🔗︎

• Published two blog posts that reached the front page of Hacker News.
  • This continues my lucky streak on Hacker News with five frontpage articles in the last four months.
• Launched the paid version of Portfolio Rebalancer.
• Put infrastructure in place to generate new Is It Keto articles automatically.

Lessons learned 🔗︎

• Traffic is a vanity metric.
• Auto-generating pages in a useful way gives developer-authors an advantage over authors who use generic publishing platforms like WordPress.
  • This is still a theory, so we’ll see if I can make it work.

Goals for next month 🔗︎

• Add 100 new articles to Is It Keto (a ~50% increase in the current corpus size).
• Publish one new blog post.