Me: “Can you give me any advice on self-publishing?”

Paul: “So — have you written the book yet?”

Me: “Umm, yea.”

Paul: “Well, there goes my biggest piece of advice.”

I had this conversation with my good friend Paul back in September of 2022, when I was trying to figure out how to turn my messy draft of a book on analytics into something people might actually find and read. Paul had self-published a book on computer networking several years back, but did not recommend that I follow his example.

He was giving me excellent advice, but I didn’t follow it… and you probably won’t either. How come? Despite it all — the amount of work involved, the difficulty in marketing a book, the minimal prospects for making money on books sales — lots of people want to write a book. Last year, I decided to give it a try for myself and wrote “Google Analytics Alternatives” in anticipation of the shutdown of Google’s Universal Analytics.

This post is intended to answer some of the frequently asked questions I’ve gotten since publishing, and also include some of the financial details that would have been helpful to me before I wrote my book. I self-published the book and it was released November 1, 2022 on both Amazon (paperback and Kindle) and Gumroad (PDF and ePub).

The writing industry is full of poorly sourced stats… it’s even worse than digital analytics.  Such as “81% of people want to write a book“, or that “15% of Americans have started (and 8% have completed) a novel”, or that only “.6% publish”. There are many different numbers out there on this topic and very little of agreement. Whatever the “real” numbers are, it’s clear that while many people are interested in book writing — only some start their books, fewer finish, and only a very lucky and dedicated handful find traditional publication.

Yet the world is awash in books. Even before the rush of AI tools came on the scene this year, there had already been an explosion of books published. AI will only open the floodgates further. I got that “81% want to write a book” number from a New York Times op-ed in 2002, which also said that about 80,000 books were published in the US each year at that time. 21 years later, that number is now somewhere around 3 million books published per year (up 38-fold!). While I find it likely that the former number is understated and the latter overstated, it’s clear that the number of books has absolutely skyrocketed. eBooks, self-publishing, and print-on-demand have all contributed to this rise. Self-published books now make up the majority of the total, about 2.3M of that 3M.

There’s about 330M people in the US, so does that mean 1.2% of Americans published a book last year? Decidedly not… non-US residents also publish books here, authors may write more than one book per year (yikes!), and it’s very easy to double-count books when trying to establish how many books were published. For example with my book I used 4 different ISBNs, one for each different edition. Quality book databases like Bowker’s Books In Print® will do their best to de-dupe, but it doesn’t always work. As an example, my book is listed twice on the popular Goodreads site.

While the number of books published per year has skyrocketed, sales revenue has actually decreased (down 38% comparing 2022 to 2000, adjusted for inflation).

Once you get your book out there, it’s very challenging to get it noticed and purchased. Book sales follow the same kind of power-law distributions as most mass-market content does, meaning that it’s a small amount of books that draw the vast majority of sales. While the “average” book might sell 250 copies (source), I would expect the median number of sales for a self-published book to closer to single digits.

Did I convince you not to write a book yet? No?? Let’s continue then, and I’ll describe what it was like to actually go through the process.

How long did it take to write the book, and what was the process like?

It took me about 7 months to go from starting research to publication, April to November of 2022. This would be very fast in the world of traditional publishing, but seems to be reasonably typical for self-publishing. This is part of the reason that I decided ahead of time to not even try to get a traditional publisher since I wanted to turn it around quickly ahead of the Universal Analytics sunset date.

Stage 1: Research
3 months: April, May, June 2022

I started by determining which tools to test. It was important to me to have a consistent methodology for selecting the alternatives. My methodology was primarily based upon popularity of the tools within the top 1M sites. Probably the most common question have gotten about the book is “why didn’t you review product x?”, and it was good to have a methodology to explain why I didn’t include some particular option. Unfortunately my insistence upon actually using all of the tools with live websites led to me not being able to evaluate Adobe Analytics and Piano Analytics, since there was no way to do so without actually buying those (enterprise-level) products.

I started by purchasing my two test domains. I wanted sites that had existing real traffic from different sources, ideally with at least 1,000 users a month. Then I got those sites setup with all the different tools I wanted to test. Which yes, means that I had as many as 15 analytics tools installed on one website… fun times.

Stage 2: Writing
4 months: June-September 2022

I wrote the book in Scrivener, which I highly recommend. Scrivener was very helpful nearly all the way through the process from organizing my research to exporting ePub and PDF files ready for publication. The two cases in which it was not as great was for collaborating with my editors (we did that in Google Docs) and doing figures and tables (which may explain why many tech books are done in AsciiDoc instead). Thanks to Scrivener I also have a record of how many words I netted per month. The “net” is an important distinction there because I added a lot and lots of words that I ended up deleting later:

How many words did I write per day? About 1,5000/day on the days I was able to put aside most of the day for writing, but those days were rare. The more typical scenario was that I set aside maybe 3-4 hours of a day and got 500 words. I found that if I wasn’t able to dedicate at least 3 hours at a stretch for writing that it was not really worth trying to write, as it took a long time for me to really get into the flow of things. There’s a ton of discussion on writer blogs about word counts, but for me it was more important to try to schedule time to write than to think too much about how many words per day.

37,000 words (171 printed pages) is on the short side for a book, but it felt like the right length to get across what I wanted to in a readable way. Going into deep details on the individual products is not something I wanted to do.

Stage 3: Editing
2 months: September, October

I hired professional editors, and I’m glad I did, even though it was the biggest single expense. The book is much more readable and succinct for it. After spending all that time writing a book the absolute last thing I wanted to do was re-read the stupid thing one more time. I tried automated tools like ProWritingAid and Grammarly, but the results just weren’t to my liking. This can be a tough choice if you’re self-publishing on a budget, but I would highly recommend hiring an editor if at all possible. I looked at the Upwork and Reedsy marketplaces for freelance editors, though I didn’t end up hiring from those.

I wanted someone that had experience with Google Analytics, so my editors were not just doing copy editing, but also helping with the larger conceptual issues of the book (typically called developmental editing).

To further improve the content I also was fortunate to recruit a few “beta readers” to give me more general feedback on how I could improve the book. While the content of the book itself wasn’t all that much different after all this editing, it flowed so much better and really communicated what I was trying to get across.

Once all that was done the editors did a proof-reading pass to find anything missed the first time or any additional mistakes I’d introduced. Ideally proof-reading should happen after everything else is done, but I didn’t totally follow that rule and continued to make edits up until release.

Stage 4: Production
1.5 months: September, October 2022

Just because you’ve got the contents of a book doesn’t mean you actually have a book. You still need cover art, an ISBN, copyright, reviews for the cover, an author bio, etc, etc. You also need to go through the medieval sounding process of “typesetting”… which is just formatting your book for display. For me I had to format for 4 different destinations: PDF for printing, PDF for electronic viewing, ePub, and ePub for Kindle. Scrivener let me compile the book for all of those different destinations pretty easily once I had created the styles for each of them.

Kindle Direct Publishing (KDP), which is Amazon’s self-publishing service, does have tools that can help you turn your MS Word document into something that will look ok on their platform, but especially since I wasn’t just publishing to KDP I did not go that route. KDP is a good platform, but it only allows you to publish to Amazon. If you publish on KDP you can e-publish in other places, but if you choose KDP Select (which allows for higher royalties), then you can only distribute digital versions via Amazon.

Stage 5: Marketing & Updates
November 2022 onward…

During the writing process I gave very little thought to how I was going to market the book. During drafting, anything that takes you away from actually putting words on the page can be a dangerous rabbit hole.

Strategically it probably would have made a lot more sense to plan out how the release would work before I got underway, but writing a book is daunting enough without having to think about sales & marketing, which is decidedly not a strength of mine.

The primary place I marketed the book was through LinkedIn, but I tried a lot of different channels. I also spent quite a lot of time during the first couple of months updating the book. All tech books become outdated quickly, but the nature of my book with 15 different products to keep track of meant there were updates every week.

The advantage of a platform like Gumroad is that when I release updates I can send an email to people that have purchased the book a link to an updated copy. Amazon does not give you any contact info for who buys your book, but Gumroad does.

So far I’ve released 4 updated version of the book after the initial release (with a changelog documenting the updates), with the most recent update being in February. I’m not sure how useful updated versions are for those who have already read the book, but it is very helpful to be able to continue to update the book. Because the paperback version is print-on-demand I can keep it current without having to dump a bunch of old out-of-date copies. In March I decided that it wasn’t worth the large amount of time I’d been spending keeping up with the constant changes in all 15 products and ceased publishing updates.

How many books did I sell, did I make any money?

How much did it cost to do?
The primary cost was of course my time spent writing the book. Beyond that (notice how I’m going to gloss right over that part…), the expenses were:

I’ll be honest, I went over-budget by more than I’d have liked. It’s also a bit embarrassing to admit that writing this article was the first time I’d explicitly summed all these items up. In retrospect there was some things I spent money on that I didn’t need to (advertising, some of the software and data, cloudflare, etc). As a learning experience I’m glad that I did spent money on things like advertising so I could see how it worked.

How many books did I sell, and how much did I make?
280 copies for a total of just about $4,000 (so $14 in royalties per copy).

Before anyone starts feeling bad for me being over $5k in the hole for writing a book, I’ll state that I closed that deficit via consulting gigs that I can directly attribute to the book. Establishing yourself as an expert in the field and raising your profile is undoubtedly the real reason to consider writing a book like this, monetarily speaking.

For the three different formats:

For example, a Kindle sale in the US would net me $8.73 and a Gumroad eBook sold in Germany would net me $17.55. In particular, the Kindle royalty percentage was frustrating to me — making only 35% of the sales price vs. 90% on Gumroad for a nearly identical ePub file. Amazon does offer a 70% royalty program, but if you opt-into that you can’t set your price higher than $9.99 and additionally Amazon will charge you for download bandwidth. The larger size of my book (because of all the screenshots) meant the most I could make on the 70% royalty program was $4.60 per book.

As you can see, the percentage I kept from sales with Gumroad was much much higher than Amazon. This allowed me to do a couple of things:
1. Purchasing Power Parity (PPP) pricing on Gumroad.
This dynamically adjusts the price based on the location of the buyer. The US price ($25) is the standard price, and then the theory is that countries with lower wages get discounts. For example, a purchaser in Bosnia would be able to purchase for $10 and in France it’d be $19.50. (No, people in Luxembourg didn’t pay more than $25…). PPP pricing is certainly not a perfect solution, and it didn’t seem to drive significant additional sales, but I like the fairness it tries to introduce.

2. Lots of discount codes.
Nothing draws people in like a sale, and about 25% of the books I sold on Gumroad were with some kind of discount code. I also offered a “site license” version of the book which was sort of a discount, 20% off for 5 licenses bought in a bundle. Even though I didn’t see a large amount of the $99 version of the book, it still made a significant impact on my bottom line and I’m glad I did it.

I didn’t run any sales or discounts on Amazon, and especially at first I primarily promoted the Gumroad edition. Overall most of the books I sold were on Gumroad, but as time passed that has shifted more to Amazon.

Gumroad eBook sales included a package of both PDF and open ePub versions together

Includes Gumroad sales only, and original attribution on “website” sales were typically lost.

Amazingly, the sales breakdown quite closely matched a poll I did before I wrote the book asking for preferred format.

That survey was what made me choose to release on Gumroad as well as Amazon, so I’m glad people actually did buy in a similar distribution as to what they said they might.

As far as the sources of sales, I’d guess that a lot of those “Direct” sales were from various Slack and Teams group chat. I’d consider those to be word-of-mouth. For someone that claims to be an attribution and SEO practitioner I didn’t do a great job on either of those fronts. I never made it anywhere near the 1st page on Google for “Google Analytics Alternatives”. Part of my issue was that I had 4 different URLs for my book:
1. The https://gaalternatives.guide alias, which redirected to…
2. The book’s homepage on my website: https://www.quantable.com/analytics/google-analytics-alternatives/
3. The Gumroad book page: https://quantable.gumroad.com/l/ga-alternatives
4. The Amazon book page: https://www.amazon.com/Google-Analytics-Alternatives-Navigating-Options-ebook/dp/B0BN2JB7NZ/

At one point I had 3 of those URLs on the 4th page of search results pages, but that doesn’t drive very much traffic. If I were to do it again, I might try to have everything on a domain like #1, combining my numbers 1-3 above and consolidating my organic efforts. I’m not sure that would have done much better, since that would have been a new domain without any links, competing against domains like moz.com, searchengineland.com, semrush.com, etc. which are on the first page now.

Where’d I make the most sales?

Tim Wilson did a bunch of stuff to help me with my book and I repay him with graphs like this?

I sold books in over 30 countries, which is kind of amazing.

Since I and much of my network are in the US it’s unsurprising that was my biggest source of sales. I was a little surprised (though I probably shouldn’t have been) that the book was nearly as popular in Europe as it was in the US. Also interesting to see Canada as #2 as I would have bet Germany or the UK.

Before this post ends up being a book itself, I should wrap this up. I hope providing some of the details of my particular journey will help someone else with theirs. Writing a book is undoubtedly a challenging project, and one that isn’t likely to reap you a lot of direct financial reward, but it can be really rewarding and occasionally fun as well! I’m happy to chat about the process with anyone who has more questions, please leave a comment here, email me (jason [at] quantable [dot] com), or message me on LinkedIn or in the #measure slack.

Share this post: on Twitter on Facebook on LinkedIn