How ONE Mistake (Almost) Cost Me $7613.90 in Yearly Income

One Mistake - Oops!Testing is an integral part of successful Internet business.  With a few tweaks you can take a so-so marketing campaign and grow it exponential levels.

On the other hand, testing can also have disastrous results.  Often a single mistake results in a significant loss in revenue.

Today I’d like to talk about how one mistake could have resulted in a $7613.90 decrease in my yearly income.

There’s a lot to be learned from this mistake.

Specifically it’s a classic example of what to not do while testing any marketing campaign…

[title color=”green-vibrant” align=”scmgccenter” font=”verdana” style=”normal” size=”scmgc-2em”]My (Budding) Amazon Kindle Empire[/title]

In the last few weeks I’ve seen some decent results with my two Amazon Kindle books.  At one point the guide How to Make Extra Money Fast Online with Your Computer was averaging 13 sales a day which works about to be $27.13.  Not bad for a simple eBook that I put together as a test project.

Seeing these results, I did what any marketer would do:

Try to increase my sales profits

The idea was simple.   I would improve a few different things to make the book more interesting to prospective buyers.  So I:

  1. Edited the title using a benefit-driven headline
  2. Uploaded a brand new cover image using my favorite Fiverr designer
  3. Changed the tags of the listing

I did all three of these things in a single day.  I felt that the previous version of the eBook had a mediocre eCover and the title was uninspiring.  So changing them would definitely skyrocket my sales numbers..

The result?

3 sales the next day.  Then 3 on the next.  Finally on the 3rd day, I hit the panic button and reverted everything back to the original listing.

In a single day, I took a book that was on track to generate $9902.45 a year (13 sales at $2.09 profit * 365 days) and brought it down to $2288.55 a year (3 sales at $2.09 profit * 365)Or a $7613.90 loss in income.

[title color=”green-vibrant” align=”scmgccenter” font=”verdana” style=”normal” size=”scmgc-2em”]The BIG Mistake…Explained![/title]

The mistake was one of arrogance.  I didn’t slowly tweak my marketing campaign. Instead I got cocky and broke the cardinal rule of Internet marketing:

[quote type=”medium” align=”left”] Only test a single variable at a time. [/quote]

With testing it’s important to know the EXACT reason why a campaign is successful (or unsuccessful).  Tweaking multiple variables makes it impossible to know the exact reason why you’re getting a specific result.

My mistake was to change three different things at the same time (eCover, title, and tags).  When I saw a negative result with my campaign – It was impossible to pinpoint where things went wrong.

In addition, Amazon is tricky when it comes to their ranking algorithmSuccess breeds success.  The higher your book ranks on the “Top 10 Kindle Category Lists,” the more sales it’ll generate on a daily basis.  So I’ve hurt the long-term profitability of this particular book because I tested too many things at the same time.

[title color=”green-vibrant” align=”scmgccenter” font=”verdana” style=”normal” size=”scmgc-2em”]Applying this Lesson to Your Internet Business[/title]

What’s the lesson here?

It’s simple:

[quote type=”medium” align=”left”] Never, EVER Test More than One Thing with your Internet Business [/quote]

(Tweet This if You Agree!)

I know that feeling you get when you see a positive result with a marketing campaign.  You want to take immediate action and maximize the results.  So your first impulse is to immediately “improve” everything about the process.

My advice?

Patience young grasshopper.    Don’t make a boneheaded mistake like me and try to do everything all at once.  Instead make a list of what you want to test and slowly roll them out.

Test one thing and see what happens.  Then make increasing improvements on whatever gets a positive result.  Rinse and repeat.  Focus on slow, continuous improvement till you’re making a lot of money with this income stream.


Originally I titled this post: How ONE Mistake Cost Me $7613.90 in Yearly Income

But I had to add the (Almost) in the middle because I’ve recovered slightly from this mistake.  In the last few days, I’ve climbed back to the 7-10 daily sales mark.  While it’s not at the previous level, I’m still happy to see that this error was somewhat fixable.

Moving forward, I will make absolutely sure to never test more than one thing at a time!  🙂

Take Action. Get Results.

33 thoughts on “How ONE Mistake (Almost) Cost Me $7613.90 in Yearly Income”

  1. Hey Steve,

    Wow. It’s great that you recovered the mistake. And that you’re sharing it. I’m about to publish my first kindle short story, and I need all the information I can get. It’s all new to me.

    How long did it take before the ebooks got approved?

  2. Hey Jens. Thanks for the comment.

    It usually takes less than a day to get approved by Amazon. Sometimes in about 12 hours. That’s the one thing I love about this platform is you can immediately be in business to make money!

    • “Never, EVER Test More than One Thing with your Internet Business” – I’ll remember this Steve… And thanks for sharing the mistakes that you’ve made. This is really inspiring esp. the fact that you got it all back up just in time.

  3. WOW! That is truly shocking. No wonder you were panicking!

    Thanks for explaining AMAZON’s ranking algorithm … have not experimented with Kindle books yet so your post will definitely come in handy when I eventually do.

    I’m guilty of doing this too (i.e. testing more than one variable). Not with Kindle books but with Google Adsense. You gotta be really patient and wait it out before you test another variable.

    • Marcia,

      First off, Kindle is definitly worth it in the long run. It is a model that makes sense and I think more and more people will be transitionig that way, so better to get started sooner rather than later…

      But yeah, I have also experienced “jumps of logic” with adsense too. Just never so dramatic. Testing is essential, but figuring out what the data REALLY means is just as important!

  4. Steve

    Kindle seems to be the new ‘gold rush’ at the moment in the IM world. courses spring up all the time. I’ve stuck my nose into a few of these courses to see what they talk about….and there’s one thing that seems to be missing that should be a prerequisate for Kindle success – and that’s WRITE A GOOD BOOK.

    I plan to get some books on Kindle and in POD via Create Space later in the year – I’m really, REALLY looking forward to seeing how they go. (And to seeing my name(s) in print!!! Ha ha ha – that’s something I wanted since I was 12!)


    • Paul,

      It is pretty cool seeing your name in print. Kindle is certainly becoming the “rage” but I think the more that people go to that as a format, the more people will ONLY be looking to make purchases in that format. So it feeds itself!

      But I agree, it is essential to have something worth reading. A steaming pile of crap will just get a boatloads of returns and an abyssmal ranking. This is an area where word of mouth means everything.

      There is ALSO a marketing aspect still though. It can be gold but people have to start finding that gold before they know it is gold!

      Anyhow, I look forward to seeing you on Kindle too!

  5. Hi Steve,
    you are right – A/B tests are very important. Thanks for sharing your tip (with Amazon – I think I should try it 😉 )

    In Google AdWords and Facebook Ads you can define a lot of versions of almost the same ad – you should change only one thing comparing to the base ad and you can see which is the best 😉

    BR, Chris

    • Thanks Chris,

      For sure spilt testing is essential. I have made a lot more off of doing it “right” than I aver had from my mistakes. But making the right calls on what you see and not getting “cocky” really helps too! 😉

      You certainly should dip your toes in Amazon. Like anything I think it takes time (and quality) to reach a critical mass, but more and more people are going that way (and looking for their info products there)

  6. Great reminder about split testing Steve.

    This is something I need to do more of. Of course, split testing only works is you have the numbers to draw the true picture out of the stats.


    • Thanks for the comment Mark.

      For sure it is GIGO (Garbage in Garbage out);

      Or maybe in this case GIMSO (Gold in My Stupidity Out) 😉

  7. Hope you are enjoying your trip, Steve!

    I agree with you, Steve. We should never test more than one variable at a time – I have done this in the past, it cost me lot of my time ad effort, because one variable might change the things I was experimenting with another variable.

    Thanks for the tips, Steve. This would certainly help me when I publish my Kindle Book.

    I understand that changing those three values had an effect on your sales, but do you know why? Is it because you changed the title?

    • Jeevan,

      Thanks for the comment.

      I am still 100% unsure as to why it changed the ranking so drastically. But yeah, my guess is either title or tags. SInce with amazon you begin to rank in categories and if you rank high you will show up in peoples “other people have looked at” and “if you like this book you may like”. My guess is whatever i did wrong dropped me out of those queues..

      I am going to proceed slowly and do individual changes to come up with a firm reason why in the future.

      I will let you know if i come up with a definite answer.

  8. Hi Steve,

    Saw your book in action 🙂 Very impressive! As you pointed out, only test 1 thing at a time, which is absolutely true for splitting testing. Split test is essential for the the business, so it only makes sense to test 1 at a time because there is risk involve.

    Thanks for the sharing your experience. Kindle is absolutely a gold mine and Amazon is a marketplace that worth to tap in. My 2nd kindle book is almost ready and just need to beef it up. I’ll most likely to publish it by next week 🙂


    • Ming,

      Awesome! that is great to hear man! Glad you are jumping in. I think soon everyone will (who isn’t already) so it is good to get in there early. Let me know how it goes!


  9. Glad to see that you were at least able to fix the mistake to some extent. Definitely a lesson for us to all learn about split testing. After further testing, did you ever figure out which one it was that resulted in the big sales drop?


    • NOt a hundred percent sure on “exactly” what did it, but I am pretty sure I know why. One of the changes must have dropped me out of the internal linking system within Amazon itself. The “if you liked this book, you may like….” section.

      Going forward (once I am back) I may make individual changes and see if I can nail down a firm answer.

  10. That’s a nice stream of income coming in from your ebooks. I’m also looking to have some books written soon and get into the Kindle rush.

    Nice advice about only changing one thing at a time.

  11. Hi Steve,

    Without testing variables, you may not be earning the maximum potential. The key to remember as your post demonstrates is to test variables ONE AT A TIME!

    Sharing this with us lets us know just how easy it is to loose if you aren’t doing things right.

    As you state we must have patience. The positive side is we can get answers to all the different variables to better out marketing efforts.

    We can test everything, just changing one word in a head line can make a big difference! Changing the color of your sales piece. Changing the format, etc..

    As marketers we are also information gatherers of our sales pieces. Like everything else in life there is a right way of doing things and a wrong way.

    Thanks for sharing,

    Terry Conti

    • Thanks Terry,

      I absolutely agree with you. I have always tested the crap out of everything I did. But you know sometimes you get too big for yourself and need to be reminded to take things slowly and carefully.

      Hopefully it is a good reminder to us all. You make your biggest mistakes when you know what you are doing but think you are somehow above the the rules. We aren’t! 😉

  12. Nice post Steve. I’ve made the same mistake before too. I started changing one thing on a niche site and then noticed something else I’ve been wanting to change and then another. I pretty much made sure I couldn’t keep track of all the changes and what worked and what didn’t.

    With niche sites, ebooks, or any product offering they are rarely considered “done”. The creator will always want to make one more tweak to make it even better. As you’ve pointed out, that can have disastrous outcomes.

    Thanks again,
    ~Johnny Bravo

    • Thanks Johnny,

      It is always good to know that I am not the only one to make this kind of mistake. For sure, change is essential doing split tests and makign the “Right” cages can turn the ordinary to the extraordinary, but it is important to get it “right”

      Thanks for dropping by and commenting, have a great day/weekend!

  13. Great post Steve!

    It is certainly a best practice to test / change on variable at a time. On the flip side of your outcome, it could be that one makes multiple changes at once and experiences a successful outcome, but then you don’t know which change actually caused the desired outcome.

    Again, great tip!

    All the best,

  14. I have learned this lesson long time ago, not from my online business. Since I work on numerical methods, whenever I simulate a model and seek results I should not change more than one parameter per simulation. If I do, and if I get a different result from my previous simulation, I will never know which parameter change caused that change. So yes I am lucky to have known this little crucial trick 🙂

    I always apply it to my business 🙂 But it is all for good Steve. Hope you are having a nice time!

    • Jane,

      YOu are right, this goes far outside the parameters for JUST a/B split testing online. IN general it is a very bad habit to make assumptions when basing them on multiple variables. Just a good way to get yourself a big mistake

  15. What do you think about the strategy of offering free Kindle books initially (to gain popularity among the readers) and then start charging from the 3rd or 4th book? Provided the content is good, do you think this will work better? Just want to know your thoughts on the same…

    • Actually I think quite highly of it.

      The ebooks I have up there now I started as “free” (though just for a short period) This rewarded the people who were regular reader and were interested and got that nice boost to downloads.

      Of course, the material HAS to be good…that is the most important point, but I think haveing at least 1 totally and forever free ebook can be a great idea.

      In fact I have been kicking around the idea that my “basic” affiliate marketing book I may at some point transition to free.

      So… I like it. So far I have only done the “start” free. But I think having one totally free ebook in your arsenal is a good way to let people see what you are about and make a decision before buying.

  16. Interesting how you broke it down this way Steve. Testing should be a cardinal rule but not a lot of us take into consideration the small things that’s making the big stuff stand. I’m guilty of this all time because I’m a hardcore multi-tasker.

    Hence, slow as a turtle and stick with one. Learning more and more each time Steve.

    On another note, how’s it going with the road trip so far?

    Thanks for the share and also a peek into your Kindle experimentation which was what I was curious about previously.

    Talk soon!

  17. Hey Steve,

    Not sure if this was the affiliate book I purchased from you from Amazon (which I read on a kindle while traveling a few days ago) but man, I think the book was fantastic as it was.

    If you did hurt your sales by modyfing three aspects of it, well, a hard lesson learned then.

    Normally this scenario will happen when you are expert enough at something and I could use a ton of different examples where people would tweak several different things at the same time after a ‘test run’ (F1, motocross, skateboard, sail boat, fishing reel, hunting rifle, etc, etc)

    The trick with Internet Marketing relies in that you just never know 100% as each part that you tweak, does not depend entirely of another part of the same thing (we’re not talking about a machine, if we were, then it would be easier) so yeah man, I understand your experience and the good thing is that you managed to correct the problem right away.

    Honestly I wouldn’t be too afraid to modify it again but maybe I wouldn’t start with the headline, I would probably start with the tags and leave the rest alone as it is good enough already (even the tags right now but if you insist on moving something, you could do those).

    Great lesson man, thanks for sharing it and thanks for the reminder!


  18. This is great lesson to learn Steve, I’m happy that your ebook sales are recovering.
    It happened to me also, but with ppc advertising and banner, we’re all humans and we make mistakes. But imagine what would happen if testing did work, I still use testing because we need to risk from time to time in order to increase sales.
    thanks for sharing

Comments are closed.