Paging Stats Police: Localytics Are Completely Incompetent, Fail Basic Math, Do NOT Buy Their Consulting Or Reporting! (Updated)

I HATE Bad Stats! This is a rough enough industry without the misinformation! Mobile is the fastest-growing giant industry ever (went from zero to 1 Trillion dollars in annual sales in only 29 years, literally the current world record for a new industry). Mobile is also the most complex giant industry, and currently, the most competitive and at least within the handset space, over the past few years, its also been the most volatile industry by its market share measurements. Arguably mobile handsets are the most competitive global industry ever - by measure of how many global Fortune 500 sized rivals are contesting the same single industry at the same time. (not to mention, this industry -mobile - produced the world's wealthiest person and the most profitable company) 

It is difficult to find solid numbers on this industry and we cherish what good info we find. I provide for example a regular free feature on this blog for the industry on the smartphone statistics and market shares. You can find the latest numbers like Q2 data for 2012 here and Q1 data here. What this industry can ill afford is incompetent muddling with silly made-up numbers and utterly false 'stats'. Such as those incompetent fools Localytics out of the USA who suddenly burst on the scene with fantastical stats today about Windows and Nokia 'market share' and unit sales numbers. (I will not link to their page to not give them any more 'credibility' but you can easily find them with simple Google search).

So we page the Stats Police. What is wrong with Localytics numbers? They report 7 monthly 'market share' splits of Nokia, Samsung, HTC and other smartphone makers both globally, and by US share. They do not give any methodology and no sources. But because we know for a fact the Nokia numbers for Q1 and Q2, we can calculate the total of the tables (on the global table, not the US table). If we take the least amount of error in their table, ie using the April-May-June Q2 numbers as the baseline for Nokia - ie Nokia reported 4 million global Lumia sales, then this table by Localytics claims that the April-May-June average of Windows Phone sales was 50% Nokia (46%, 50%, 55%, on a strictly mathematical average, actually 50.3%). That means that when we add the other Windows Phone smartphones that Localytics has managed to find, they report Q2 Windows Phone in total sold 8 million smartphones in Q2. And the same table tells us 3.5 million total Windows Phone smartphones in Q1. And obviously Localytics stats tell us that Windows Phone total smartphone sales more than doubled from Q1 to Q2. That sounds wonderful if you are Microsoft, doesn't it.

Now, lets ignore my stats so we remove any of my possible personal bias in this policing action. Nobody has heard of this fly-by-night USA operation called Localytics who recently wanted some visibility as a Windows Phone analytics provider (that should give us some clue as to their possible motivations to present a biased picture, haha). Lets go to the big 4 - there are 4 big analyst houses who report every single quarter, year-in, year-out, the handset sales numbers, and for many years now, have reported the smartphone sales numbers. They are Gartner, IDC, Strategy Analytics and Canalys. Of the four, three report smartphone operating system numbers quarterly (IDC, Gartner and Canalys). They all tend to have very similar numbers while not exactly the same and they measure slightly different measures (shipments from vendors, or end-user sales to consumers) but they are almost always close to each other. All have reported either Microsoft combined market share (Gartner, IDC) or Windows Phone specifically (Canalys). These guys have the most credibility in the industry on what is the reality to handset market new sales/shipments per quarter.

So lets take the average of the three houses and compare that 'reality' to this new Localytics reporting (and as this is my blog, let me add my own company TomiAhonen Consulting official market share stat also just for reference)


Source . . . . . . . . . . . . . . . . . . . . . . . . Q1 2012 . . . . . Q2 2012
Average of Canalysis, Gartner & IDC . . . 2.8 M . . . . . . . 4.9 M
TomiAhonen Consulting . . . . . . . . . . . . 2.3 M . . . . . . . 4.6 M
Localytics utterly bogus number . . .  3.5 M . . . . . . . 8.0 M
Localytics error vs Average of big 3 . . . . 24% too big  . . 64% too big 

(Note, my number here, is purely Windows Phone, but I also report for both periods the number for Windows Mobile, when that is added to the above number, my number is within 100,000 or 4% of the average of Canalys, Gartner and IDC in both periods, but Localytics only reports Windows Phone number, so to keep this as close to apples-to-apples, I only listed my Windows Phone number).

So Localytics wants to have sensational numbers, with Windows growing phenomenally - more than doubling in size in just one quarter (growth rate of 128%) - and for Q2 they find that Nokia's Lumia Windows Phone accounted for half of all Windows Phone sales globally. So in addition to the known quantity of 4 million Lumia sales in Q2, this fake facts peddler outfit called Localytics claims another 4 million sales mostly from HTC and Samsung. Yeah. And these three big trusted industry analysts only found 870,000 sales outside of Nokia Lumia !!! (and in the case of Gartner and IDC, that includes even soem Windows Mobile sales). Now we have an error of the magnitude of... 359% !!!! That is MASSIVE error.

Oh, and how does that square with all reported facts from HTC and Samsung that they were reducing their involvement in Windows Phone after Nokia joined in? And how does this square with ALL reported end-user surveys from the USA to Europe to Asia about actual end-user market shares of Windows Phone? And how does this square with Microsoft's own statements about Windows Phone sales being disappointing?

There is some truth to the reporting that yes, Nokia is now the biggest Windows Phone maker, and by a wide margin. But that did not happen now in June, as per these bogus fake 'statistics' pushed by the utterly non-credible con-artists Localytics, no. The point when Nokia passed all other Windows Phone makers happened - according to all of these three big industry analyst houses (and according to my stats) already in Q1, and by a wide margin. The average Nokia leadership in Windows Phone of these 3 big analyst houses was 71% in Q1 and 82% in Q2. These fake charlatans Localytics don't 'find' Nokia passing the 50% point until in June, at the end of Q2. 

Then lets look at Q1 - ie their months January, February and March. When we add those sizes together - remember, their chart clearly is labeled to count "total global Windows Phone handsets" as its measure, not a percent of market share, but actual units - so we now find Localytics only measured 1.2 million Nokia Lumia smartpones in Q1 ??? WTF ? Nokia itself reported 2 million. What is wrong with these guys? Don't they know the industry or did they fail basic math in high school. That table is utterly completely wrong. They have undercounted Q1 Nokia sales by 40% !!! But they have somehow 'discovered' an otherwise unreported huge surge of non-Nokia sales - compared to the experts - they found 187% more non-Nokia global Windows Phone sales in Q1 !!!!  This is so so so so SOOOO fake. These clowns simply haven't even done their basic homework. Anyone with the faintest overall view to the industry knows these numbers cannot be true.

By the way, that is the LEAST error we can find in this utterly crazy statistics they now publish. If we use the Q1 data as their starting point, the errors get even bigger and I won't waste my time on such foolishness. If you report data that quarrels so badly with all reported stats, it can only mean one of two things - either you are utterly incompetent, or else you are corrupt (ie paid for by Microsoft - did you notice that two of the five bosses are ex Microsoft execs, haha, so this is again part of the Microsoft war against the truth). Yes this falls perfectly into that pattern of Microsoft-generated lies about Windows Phone that we've chased on this blog, like that in Britain somehow the consumers had grown tired of iPhones (remember that) or in China Windows Phone outsells the iPhone (haha, how did that turn out) and the other dirty tricks against the truth the friendly folks from Redmond like to run, like the fake blog comments they were caught generating by employees of Microsoft and Nokia, etc. Yes. Yet another in the fake facts file. This Microsoft Windows Phone campaign starts to resemble Mitt Romney's campaign against facts in the US presidential race right now..

But back to these crazies at Localytics. Now note July! These incompetents now include a bizarre July statistic too, by which - get this - HTC shows INCREASED sales of Windows Phone over Q2, and Samsung shows INCREASED sales of Windows Phone over Q2 - and Nokia shows.. a massive 56% increase in Lumia sales over Q2 (suggests Q3 sales rate to 6.25M just in that one month, or if the trend were to continue, a level of 9.7 million Lumia sales just in Q3, utter total absolute bat-shit crazy nonsense, considering what we have heard from every source including Nokia CEO himself that currently Nokia Lumia sales are collapsing). So this bizarre July 'finding' is AFTER Microsoft announced that all current Windows Phone sales are Osborned and ALL published reports worldwide say currently Windows Phone sales are falling with many carriers bailing totally out of the instantly-obsolete platform - to the degree Nokia lowered US sales prices from 99 dollars with 2 year contract, down to 19 dollars; and in Germany T-Mobile refused to even launch the Lumia 900, and Nokia is giving refund checks etc. Yet these incompetent or corrupt clowns called Localytics claim that Windows Phone sales are currently surging and Nokia is somehow already up 56% from Q2 levels - even after Nokia's own CEO said his company smartphone sales this quarter will be down, not up. 

Liar Liar pants on fire! These Localytics are utterly fake bogus unreliable incompetent or corrupt charlatans!

The story has now been reported on several reputable sites with claims like Windows Phone has grown 312% globally since January or related, 273% in the USA. That would be awesome growth, if that was anywhere near the truth. But just by using the big 3 houses, the actual growth of new sales from Q1 to Q2 for Windows Phone was only 72% as Nokia got fully into swing, not 3x growth.

So - please please PLEASE do not report on these hoodlums on anything other than being utterly unreliable, fake and con-artists who peddle fake stats to confuse the industry. They are either incompetent, or they are corrupt (or both). This industry needs analysis and facts, but good analysis and good numbers. Please do not buy anything from this company, and do not ever recommend them to anyone. Localytics - shut down your fake operation and crawl back under the rock where you came from. Shame on you!

And individually. You are a disgrace to the mobile industry Raj Aggarwal, Henry Cipolla, Brian Suthoff, Andrew Rollins and Will Pryor. I hope you never work another day in this industry and don't anyone ever give these clowns a good recommendation either. Shame on you!!!

UPDATED - About 6 hours after I posted this, I found that Localytics had revised its graphics at its blog, re-labeling them but not indicating it had made those changes/updates. They now state 'Global Windows Phone App Usage Share' whereas earlier when they were collecting lots of press attention, the graphics stated 'Global Windows Phone Market Share' - a huge change in meaning! (the graphic is otherwise identical) Here is the original screen shot of how they had their blog for the record:




And note, all early articles reporting on this story was talking about handset market shares, not a word about 'app usage' shares. Their press releases etc did not talk about this graph. The blog article (currently) says nothing about the change which is huge in meaning. They make no mention of the faulty reporting now linking to this newly-defined graphical information. They now have a short blurb of their methodology. I am totally ok with the company if they issue a clear apology on their blog that they published accidentially wrong information and that this is the correct info, but if they just go quietly swiping graphics without acknowledging the huge change in what they claim as their data as reported, then that is further proof of highly unprofessional behavior - still - SHAME ON YOU LOCALYTICS !!!  Confess !!!

What changed? If we are not now talking of handset market share, but apps used on those phones, then roughly speaking, the overall scale in Q2 is about in line with the installed base of Windows Phone smartphones globally (13.8 million Windows Phone handsets shipped cumulatively, and Nokia Lumia 6.6 million shipped cumulatively, which would be 48% of all Windows Phone smartphones ever sold/shipped. This would be reasonably close to the 50.3% that Localytics reports as the Nokia 'market share' in the graphic for the months April, May and June ie Q2. But if that now, suddenly, is the new meaning of this 'market share' graphic, then Q1 doesn't match up with reality..  Now their Q1 number suggests global Windows Phone installed base would have been 6.0 million ??? When in reality by Q1 the Windows Phone cumulative shipments were 9.0 million. An error instantly of 50%. No, I do understand you Localytics have a mission to push the fake news headline that Windows Phone is massively exploding globally, but you can't invent this stuff! The Windows Phone base did not suddenly more than double in one quarter. It grew by 53%, not 128% !!!!

Confess Localytics, do not try to relable your lies. This math STILL doesn't work. But here is truly the sad part, if you want to take a 'lets assume this is reality' angle to this as an installed base argument. If the Nokia 6.6 million Lumia sales cumulatively truly account for 55% of all Windows Phone handsets in use as of June ie end of Q2 of 2012, as this latest variation of Localytics data suggests - then get this. The other manufacturer made Windows Phone handsets obviously account for 45% which is 6 million. What does that mean?  That means that of the 13.8 million cumulative sales of Windows Phone since its launch two years ago, already 1.8 million Windows Phone handsets by Samsung, HTC and others - have been tossed into the garbage can as useless. They are not used anymore! That is yes 25% of these 'premium' smartphones that should be still in the prime of their lives. Note, this data was before the July Osborning of Windows Phone series by Steve Ballmer. And again, if we take the reported news, of supposed 312% global growth for Windows Phone this past six months (or 273% USA growth) - when measuring installed base, that only grew 53%, not by more than 3 times. If these Localytics fools were professionals, and they saw that their initial data was severely mis-reported, they would issue a statement about it, with clear clarification - and ask those news and tech sites to correct the faulty data. But, if Localytics is just a Microsoft stooge, then they of course love it that their data is so wrongly reported.

REAL BAD NEWS IS..

So if you really want to believe these numbers, then we know that of the non-Nokia made Windows Phone handsets that were all supposed to be magnificent and new and iPhone-killers, 25% have been totally discarded already! Hardly the stuff of any 'third ecosystem' where iPhones go to kids and relatives when new models are bought, and remain in use for years after the first owner gives them away (or sells them). No wonder the Yankee Group study found that of new Nokia Lumia Windows Phone buyers, 4 out of 10 rate it such a rotten smartphone, they gave it a 1 out of 5 on their satisfaction rating (5 being best, 1 being worst). Four out of ten who used Lumia. This is your Windows Phone right here. Failling as we speak. And if this data now is the installed base, it only goes to show, even some months old Windows Phones from HTC and Samsung are being tossed away as useless, at the rate of 25% already. That is no way to win your third ecosystem, Steve Ballmer, is it? 

Comments

[PeterElgin]

tomi,

I predict nokia will sell 1 M Lumia units across the world in Q3 from 4.6M in Q2.

plus 2M Symbian units in Q3.

Completely disaster. Nokia is BK in Q3.

What's your prediction ?

Thanks,
Peter

[Earendil Star]

Tomi, don't get so upset.
You know too well that this is the usual MS tactics.
And they will continue. We all know that.
Yet, thanks for highlighting these situations.
And please continue unveiling their lies. We all need that.
Thanks

[Interested to know]

I'm glad you're out here to call BS on the Microsoft-sponsored War Against the Truth.

[John Waclawsky]

The Microsoft game plan is just like Zune and history is repeating itself. Please take the time to read the "entire" reference belwo. There are direct parallels to what is happening today! It's an excellent read:
http://www.roughlydrafted.com/RD/Q4.06/D0BC712B-7DBA-46CA-AA44-19376E64FBA6.html

You have a great blog. Keep up the great work

[eduardo]

"Liar Liar pants on fire! These Localytics are utterly fake bogus unreliable incompetent or corrupt charlatans!"

Come on Tomi. Don't hold back, tell us what you really feel.

[midwaslll]

Android 4.0 Android Mini PC & Android TV Box (Cortex A5, HDMI, USB, 4 GB)! Precio: $ 55.00! http://redir.fr/ghgqb

[B95]

OMG - tempest in a dew drop. Localytics is a start up company focused on instrumenting apps for analytical reporting. They have a few apps that they/their customers instrumented across platforms and were reporting how those apps are being used.

So they mislabeled one of their graphs - big deal. Their sample is still too small anyway.

Anybody that knows the company, knows what the graphs were all about.

[John Waclawsky]

Tomi, I posted earlier and I suspect you are busy. But I think, given what you know, it would be really fascinating to see your analysis of the current WP7 and WP8 industry or eco-system "activity" (or hype) compared to how in the past Microsoft approached Zune, which is explained at

http://www.roughlydrafted.com/RD/Q4.06/D0BC712B-7DBA-46CA-AA44-19376E64FBA6.html

There is an emerging "big picture" when you compare WP7/8 activities and how Microsoft consumer products struggle (an understatement) when exposed to real competition. Explained at:

http://www.roughlydrafted.com/RD/Q4.06/2E6D9BB2-FE1B-4556-8389-67BD581FBCCC.html

Wouldn't you agree the whole (unfortunate) Nokia partnership is just a part of this reoccurring pattern. Why didn't anyone on Nokia's board find this obvious history of consumer failure before they decided to bet the company?

[foo]

New numbers from comScore.

Japan’s Smartphone Surge: 1 in 4 Mobile Consumers Now Owns a Smartphone

Smartphone Audience Grows 43 Percent since the Beginning of 2012

Nearly Two-Thirds of Japan’s Smartphone Population Uses an Android-Operated Device

Microsoft market share drops from 3.7% to 3.2%

http://www.comscore.com/Press_Events/Press_Releases/2012/8/Japan_Smartphone_Surge

[vladkr]

Tomi, please, don't tell us you're discovering that we're surrounded by misinformation (not only about mobile technology).

Not you ! You know about Stephen Elop's lies, you know about restrained (or bugged) devices to force people buy another product (let's say Lumia for instance), you know about fake reviews on resellers, you know about Ballmer's policies, so I'm surprised you're surprised.

[vladkr]

Two of the three co-founders of Localistics are from Microsoft so there shouldn't be any surprise.

http://www.localytics.com/about/

[N9]

anti wow :) and will it sell as many as L900 sold ???

[anti]

@N9

let's see, at least someone is not happy about it ;-)

[canada goose outlet]

thanks a lot for this good site view

[Moncler jacket]

Ladies decide to choose http://www.jacketsbrandsuk.com because doing so accompanies all closet any kind of secure plantar to . Better pigmentation is usually coveted, hued shrewd. These characteristics insure that it is a lot http://www.jacketsbrandsuk.com/moncler-jackets_c20 more running.

[Beats By Dre Australia]

slowly let thisWhere not only food is also nervous tension, grass, the else Handonglayue can go to cattle? These cattle raising, but come.The cattle are livestock