Tag Archives: stats

Overthinking stats

Manuel Riess has been writing about why he didn’t stick with previous blogs. On the topic of paying too much attention to stats:

When using WordPress, it’s easy to get the basic numbers of interest: how many visitors, from which part of the world etc. The next level is Google Analytics… what a plethora of settings and numbers! I stared at them all the time after a new post went up, it was exhausting.

I highlighted Manuel’s microblog in an email to Kickstarter backers recently, as an example of using custom domain names. You can follow him on Micro.blog.

I still haven’t looked at stats for Timetable. And I’ve resisted adding follower counts and page view stats to Micro.blog for the same reason. If all that drives you is the number of likes on a tweet, or subscribers to your podcast, it’s easy to get discouraged when the numbers don’t pan out. Or worse, overthink your writing when you know a bunch of people are paying attention.

Everyone has something to say. Write because you love it, or to become a better writer, or to develop an idea. The stats should be an afterthought.

iPhone SE sales potential

John Gruber runs down the list of yesterday’s Apple announcements. On the iPhone SE, he recognizes that it’s a great device especially in the short term, before the iPhone 7 is released:

If you listen to my podcast, you know how ambivalent I remain about the physical size of the 4.7-inch iPhone 6 and 6S. I was really hoping that the iPhone SE would effectively have iPhone 6S specs — CPU and GPU performance, and similar camera quality. That seems to be exactly what Apple delivered. I honestly think this is the phone I’m going to use for the next six months.

Jason Snell follows up on the Dan Moren post I linked to by also covering the new products. Highlighting the sales potential for the iPhone SE, Jason writes:

In the past year, Apple has sold 30 million 4” iPhones, out of around 230 million total. That’s only 13 percent of the total, but it’s still a very large number of phones—and this, during a year when the most modern four-inch iPhone Apple sells was introduced in the fall of 2013. Is there room for the iPhone SE to be 20 percent of Apple’s total iPhone sales? I think so.

I keep thinking about the iPhone SE price: $399 for essentially the power of a 6S, which is $649. That’s just a great value. I’ve said on Core Intuition recently that while the 6S and upcoming 7 will always remain the most popular phone, I think the SE could hold its own with the 6S Plus in units sold. Now I wonder if it could even surpass it.

According to David Smith’s stats, the Plus versions represent about 15% of active devices 4-inch or bigger. That share goes up to about 20% if you exclude older devices no longer for sale, like the 5 and 5C. That seems about right to me. If you sat around an Apple Store and watched 10 people buy iPhones, I’d be surprised if more than a couple were the Plus. Starting next week, a couple of those iPhones could be the SE, too.

Ignoring follower counts

I’ve said before that there’s something about the 140-character limit that brings out both the best and worst in people. Nick Harris hints at this while writing about taking a break from Twitter:

“But largely ignoring the Twitter Noise Machine – particularly when my timeline becomes the Twitter Hate Machine – is going to be good for me.”

He also talks about the obsession with stats and follower counts, which Brent Simmons picks up on and carries further:

“I did have Google Analytics for a few months in 2014 when I was doing sponsorships. I spent too much time looking at the numbers and trying to make them go up. But no amount of going-up is ever satisfying: I just wanted more.”

When designing my new microblogging platform, I made a conscious decision to not even show follower counts. You can get the followers from the API, but I didn’t want to have the numbers right in your face when viewing someone’s profile. It’s too easy for us to make a judgement based on how many followers they have, and so miss out on whatever they have to say.

How I use Twitter

“Twiterrific 3.0”:http://iconfactory.com/home/permalink/1887 is out, with a new price of $15 or free to use with ads. The ads are very effective and difficult to ignore, but really they don’t take anything away from the Twitter experience. The new version is great, though, and I’ll be sending my $15 to Icon Factory sometime in the next few days. “As Fraser Speirs said”:http://speirs.org/2007/11/02/twitterrific-3/, it’s a small price to pay to be connected to friends and colleagues.

I post to Twitter much more often than I blog now, and I think I owe some of my followers an explanation. I made a rule for myself early on to only follow people who I have met in real life. I’ve only made a couple exceptions to this, and none recently. It keeps the flow of tweets easier to manage and relevant.

So if you follow me on Twitter and wonder why I don’t return the favor, that is why. You probably have interesting things to say, so say hi to me at some future conference so I can add you to my list. I’ve actually been thinking about taking it one step further and protecting updates, because I tend to post about what (to this blog) have been traditionally private matters. Jury is still out on that decision.

The following numbers are interesting, though. I only do very basic Mint stats for this blog (I just care about referrers, not number of readers), but it does make me wonder how many people read this blog. If you’re reading this, add a comment to this post. (Haha, gotcha! I don’t have comments.)

Foreign sales market

From time to time on the “MacSB list”:http://tech.groups.yahoo.com/group/macsb people ask about the value of localization and what percentage of sales come from foreign customers. Since day 1 of “Wii Transfer”:http://www.riverfold.com/ I’ve always been surprised at how many sales are from Europe. At times it felt as if over half of sales were outside the United States, so I finally ran the numbers to know for sure.

My homegrown customer database doesn’t actually include the physical address, so I grabbed the last 500 sales from PayPal and wrote a quick script to group the countries. Here’s the chart:

Wii Transfer countries chart

The United States represents just over half. If you add up the other English-speaking countries, it hits 70%. Still, this is a purely English-only piece of software. I’ve resisted the push to localize until I feel the codebase is better prepared for it, and the UI more stable.

At “VitalSource”:http://www.vitalsource.com/ last year I wrote a custom Rails web app to manage localization resources for both the Mac and Windows products and deal with the outsource translators, and the takeaway from that experience was definitely to go slowly. It’s easy to end up with a foreign language version that makes compromises and is potentially less useful to customers than the English version. Depending on the size of the product, localization could take weeks or months, time that might be better spent adding features.

Back to the real stats. Why are the foreign numbers so high? I think the weak dollar combined with an already relatively inexpensive price makes Wii Transfer even more of an impulse buy in Europe.