Scope: This is a filings test of one company over five fiscal years: Microsoft, fiscal 2014 through fiscal 2018. Microsoft’s fiscal year ends June 30, so “fiscal 2016” means July 2015 through June 2016. Every number comes from Microsoft’s annual reports on Form 10-K and two Microsoft press releases, all linked at the bottom. Each claim is labeled: Fact (a number or statement in a filing), Company assertion (management’s own description), Inference (my arithmetic or reading of the filed numbers), or Editorial (opinion). Nobody is graded. Nothing here is about Microsoft’s stock, AI, or anything after fiscal 2018.
Pull up a chart of Microsoft’s revenue for the first years after Satya Nadella became CEO, and you can get two different pictures of the same years depending on which 10-K you take the numbers from.
Use the fiscal 2017 10-K, and revenue fell: $93.6 billion in fiscal 2015, $85.3 billion in fiscal 2016, $90.0 billion in fiscal 2017. A drop of 8.8% in year two of a new CEO’s tenure is the kind of number that ends careers. Use the fiscal 2018 10-K, and the same two years read $91.2 billion and $96.6 billion, followed by $110.4 billion. No dip in fiscal 2017. Growth every year in the table.
Neither filing is wrong. They’re measuring different things. And if a board, an investor, or a business writer builds a “turnaround” story on one series without knowing about the other, they’ll get the story wrong in one direction or the other.
The question: If you chart Microsoft’s first Nadella years straight from the 10-Ks, does the turnaround survive a change in revenue accounting and a $7.5 billion write-down? Or does the picture depend on which filing you read?
Short answer: The “revenue decline” of fiscal 2016 is almost entirely an accounting artifact. Microsoft’s own footnotes say Windows 10 revenue deferrals cut reported revenue by $6.6 billion in fiscal 2016 and $6.7 billion in fiscal 2017, and foreign currency cut another $3.8 billion in fiscal 2016. Add those back and the decline turns into modest growth. The fiscal 2015 operating-income collapse is mostly one item, $10.0 billion of Phone charges tied to the Nokia deal, which was announced before the CEO change. Strip the identified one-offs and operating income runs roughly flat from about $27.8 billion (fiscal 2014) to about $29.3 billion (fiscal 2017), then $35.1 billion in fiscal 2018. The cleanest growth series is commercial cloud revenue: $9.5 billion, $14.9 billion, $23.2 billion over fiscal 2016–2018. That growth is real. But the filings also show the cloud business was already being run by the man who became CEO, so the filings can’t tell you how much of it the CEO change caused.
Two 10-Ks, One Set of Years
Here are the five-year tables from two consecutive annual reports, side by side. Dollars in millions.
| Fiscal year | Revenue, FY2017 10-K (old basis) | Revenue, FY2018 10-K (new basis) | Operating income, old basis | Operating income, new basis |
|---|---|---|---|---|
| 2014 | $86,833 | $86,833 (not restated) | $27,759 | $27,759 (not restated) |
| 2015 | $93,580 | $93,580 (not restated) | $18,161 | $18,161 (not restated) |
| 2016 | $85,320 | $91,154 | $20,182 | $26,078 |
| 2017 | $89,950 | $96,571 | $22,326 | $29,025 |
| 2018 | n/a | $110,360 | n/a | $35,058 |
Fact: These are the “Financial Highlights” tables in Item 6 of each 10-K. The fiscal 2018 table footnotes fiscal 2016 and fiscal 2017 as reflecting “the impact of the adoption of new accounting standards in fiscal year 2018 related to revenue recognition and leases.” Fiscal 2014 and fiscal 2015 sit in the same table but were not restated.
Inference: That means the fiscal 2018 five-year table has a break between fiscal 2015 and fiscal 2016. Read it straight across and fiscal 2016 looks like a 2.6% decline from fiscal 2015 ($93,580 million to $91,154 million). That’s comparing two different accounting methods. You can’t splice those rows together without saying so, and most charts don’t.
What changed in the accounting
Fact: Microsoft early-adopted the new revenue standard effective July 1, 2017, “using the full retrospective method, which required us to restate each prior reporting period presented.” The 10-K says the biggest change was software licenses: “Specifically, for Windows 10, we recognize revenue predominantly at the time of billing and delivery rather than ratably over the life of the related device.” It also moved license revenue in certain multi-year commercial subscriptions to the time of contract signing. It says “Revenue recognition related to our hardware, cloud offerings (such as Office 365), LinkedIn, and professional services remains substantially unchanged.”
Fact: The restatement table in Note 1 shows the size of the change:
| $ millions | As previously reported | New revenue standard adjustment | As restated |
|---|---|---|---|
| FY2016 revenue | 85,320 | +5,834 | 91,154 |
| FY2016 net income | 16,798 | +3,741 | 20,539 |
| FY2017 revenue | 89,950 | +6,621 | 96,571 |
| FY2017 net income | 21,204 | +4,285 | 25,489 |
Fact: The same note says adoption “had no impact to cash from or used in operating, financing, or investing in our consolidated cash flows statements.”
Inference: That one sentence is the key to the whole puzzle. The restatement moved revenue between years. It didn’t add a dollar of cash. Customers paid the same amount on the same days under both methods. So when you’re trying to decide whether a business recovered, the restatement is a timing question. It isn’t about performance.
The old basis already told you this, if you read the footnotes
Here’s the part most retellings skip. You didn’t need to wait for the fiscal 2018 restatement. The old-basis 10-Ks disclosed the distortion as it happened.
Fact: The fiscal 2016 10-K footnotes that year’s results as reflecting “the impact of the net revenue deferral from Windows 10 of $6.6 billion, which decreased operating income, net income, and diluted earnings per share (“EPS”) by $6.6 billion, $4.6 billion, and $0.58, respectively.” Its MD&A says: “Revenue decreased $8.3 billion or 9%, primarily due to the impact of the net revenue deferral from Windows 10 of $6.6 billion and an unfavorable foreign currency impact of approximately $3.8 billion or 4%.”
Fact: The fiscal 2017 10-K footnotes fiscal 2017 as reflecting “the impact of the net revenue deferral from Windows 10 of $6.7 billion, which decreased operating income, net income, and diluted earnings per share (“EPS”) by $6.7 billion, $4.4 billion, and $0.57, respectively.”
Now do the bridge. Dollars in millions.
| Bridge | FY2016 | FY2017 |
|---|---|---|
| Old-basis revenue (as reported) | 85,320 | 89,950 |
| + Windows 10 net revenue deferral (company footnote) | +6,600 | +6,700 |
| = Old basis before the Windows 10 deferral (my arithmetic) | 91,920 | 96,650 |
| New-basis revenue (FY2018 10-K restatement) | 91,154 | 96,571 |
| Gap | −766 | −79 |
Inference: Adding back the deferral Microsoft disclosed at the time gets you within $0.8 billion of the later restated figure for fiscal 2016 and within $0.1 billion for fiscal 2017. The small gap makes sense: the new standard also changed the timing of some multi-year commercial licenses, so the restatement adjustment ($5,834 million and $6,621 million) isn’t identical to the Windows 10 deferral ($6,600 million and $6,700 million). The point is that an attentive reader of the fiscal 2016 10-K already had the right picture in 2016.
Inference: Take the fiscal 2016 old-basis drop of $8,260 million. Add back the $6.6 billion deferral and the roughly $3.8 billion currency effect, and what’s left is about $2.1 billion of growth in constant currency, before the deferral. That’s an estimate on top of two company estimates, and it still includes the shrinking Phone business. It isn’t a precise organic growth rate. But it’s enough to flip the sign. The fiscal 2016 “revenue decline” was not an operating decline.
The $10 Billion Year: Separating the Nokia Charges
The other number that makes fiscal 2015 look like a crisis is operating income: down from $27,759 million in fiscal 2014 to $18,161 million.
Fact: The fiscal 2015 10-K says fiscal 2015 “Includes $7.5 billion of goodwill and asset impairment charges related to Phone Hardware, and $2.5 billion of integration and restructuring expenses, primarily costs associated with our restructuring plans, which decreased fiscal year 2015 operating income and net income by $10.0 billion and diluted earnings per share (“EPS”) by $1.15.”
Fact: The detail: “we recorded a goodwill impairment charge of $5.1 billion, reducing Phone Hardware’s goodwill from $5.4 billion to $116 million, net of foreign currency remeasurements, as well as an impairment charge of $2.2 billion related to the write-down of Phone Hardware intangible assets.”
Fact: The deal behind it: “On April 25, 2014, we completed the transaction to acquire substantially all of NDS for a total purchase price of $9.5 billion, including cash acquired of $1.5 billion” (fiscal 2014 10-K; NDS is Nokia’s Devices and Services business). Microsoft and Nokia announced the deal on Sept. 3, 2013, per Microsoft’s press release that day. Microsoft announced Nadella’s appointment as CEO on Feb. 4, 2014.
Inference: Within about 14 months of closing, Microsoft had written off $7.5 billion against a business it paid $9.5 billion for, about 79% of the headline price (the price included $1.5 billion of cash acquired, so the write-off is an even larger share of what was paid for the operations). The deal was signed under the prior CEO and closed under the new one. The write-down came in the new CEO’s second fiscal year.
The people cost, in the filing’s own numbers
Fact: “In July 2014, we announced a restructuring plan to simplify our organization and align NDS with our company’s overall strategy (the “Phone Hardware Integration Plan”). Pursuant to the Phone Hardware Integration Plan, we eliminated approximately 19,000 positions in fiscal year 2015, including approximately 13,000 professional and factory positions related to the NDS business.” Then: “In June 2015, management approved a plan to restructure our Phone Hardware business to better focus and align resources (the “Phone Hardware Restructuring Plan”), under which we will eliminate up to 7,800 positions in fiscal year 2016.”
Fact: Fiscal 2016 still carried “$630 million of asset impairment charges related to our phone business, and $480 million of restructuring charges,” together reducing operating income by $1.1 billion (fiscal 2016 10-K). Fiscal 2017 included “$306 million of employee severance expenses primarily related to our sales and marketing restructuring plan” (fiscal 2017 10-K).
Editorial: Those are stakeholder costs, and they belong in the ledger as costs. About 19,000 eliminated positions in one year, plus up to 7,800 more planned the next, is not a “win” to celebrate. It is the bill for an acquisition that didn’t work. Any honest turnaround scorecard has to carry it.
Strip the One-Offs: What Operating Income Actually Did
Here’s operating income with the charges and deferrals the company itself identified added back. Every add-back is a number from a 10-K footnote; the sums are mine. Dollars in millions.
| Fiscal year | Reported operating income | Company-identified items added back | Operating income before those items (my arithmetic) |
|---|---|---|---|
| 2014 | 27,759 | none in the table footnotes | 27,759 |
| 2015 | 18,161 | +10,000 Phone impairment, integration, restructuring | 28,161 |
| 2016 (old basis) | 20,182 | +6,600 Windows 10 deferral; +1,110 phone impairment and restructuring | 27,892 |
| 2016 (new basis) | 26,078 | +1,110 phone impairment and restructuring | 27,188 |
| 2017 (old basis) | 22,326 | +6,700 Windows 10 deferral; +306 severance | 29,332 |
| 2017 (new basis) | 29,025 | +306 severance | 29,331 |
| 2018 (new basis) | 35,058 | none added back here | 35,058 |
Inference: Once you take out what management itself flagged, the “crash and recovery” shape disappears. Operating income runs about $27–29 billion from fiscal 2014 through fiscal 2017. That’s a plateau, not a collapse. The real step-up comes in fiscal 2018, to $35.1 billion, a 21% increase over fiscal 2017 on the new basis. On the new basis the operating margin goes from 28.6% (fiscal 2016) to 30.1% to 31.8% (fiscal 2018).
Inference: So the honest headline for the first three Nadella fiscal years isn’t “collapse, then rescue.” It’s “flat underlying profit while the company absorbed a failed acquisition and changed how it sold Windows, then a clear gain in year four and five.” That’s less dramatic. It’s also what the documents say.
Net income has its own one-off
Fact: Fiscal 2018 net income of $16,571 million “Includes a $13.7 billion net charge related to the Tax Cuts and Jobs Act, which decreased net income and diluted earnings per share (“EPS”) by $13.7 billion and $1.75, respectively.”
Inference: Read net income without the footnote and fiscal 2018 looks like a 35% profit drop from $25,489 million. Add back the tax charge and it’s about $30.3 billion, a record for the period. Same lesson again: in this five-year window, nearly every “bad year” headline is a footnote you didn’t read.
The Cleanest Series: Commercial Cloud, and Its Trap
Fact: The fiscal 2018 10-K: “Our commercial cloud revenue, which primarily comprises Office 365 commercial, Azure, Dynamics 365, and other cloud properties, was $23.2 billion, $14.9 billion, and $9.5 billion in fiscal years 2018, 2017, and 2016, respectively.” The same 10-K says the new revenue standard left cloud offerings “substantially unchanged.”
Inference: That makes commercial cloud the best like-for-like series in the whole window. It grew about 57% in fiscal 2017 and 56% in fiscal 2018, and went from 10.4% of total (new-basis) revenue to 21.0%. Everything else combined (total revenue minus commercial cloud) went from $81.7 billion to $87.2 billion over the same two years, about 6.7%, and that “everything else” includes LinkedIn from December 2016.
Don’t mix run rate with revenue
Fact: Earlier 10-Ks didn’t report commercial cloud revenue. They reported a run rate. The fiscal 2016 10-K: “Commercial cloud annualized revenue run rate* exceeded $12.1 billion,” with the definition “calculated by multiplying revenue for the last month of the quarter by twelve.” The fiscal 2017 10-K reported a run rate that “exceeded $18.9 billion.”
Inference: A run rate is one month times twelve. For a fast-growing business, it will always sit above the full-year revenue: $12.1 billion against $9.5 billion of actual fiscal 2016 revenue, $18.9 billion against $14.9 billion for fiscal 2017. If you put a run rate from one filing next to a revenue number from another, you’ll invent growth that never happened, or hide growth that did. Pick one series and stay in it.
LinkedIn: the acquisition inside the growth
Fact: “On December 8, 2016, we completed our acquisition of LinkedIn Corporation for a total purchase price of $27.0 billion” (fiscal 2017 10-K). In the fiscal 2018 10-K, LinkedIn revenue is $5,259 million in fiscal 2018 and $2,271 million in fiscal 2017 (a partial year).
Inference: Fiscal 2018 revenue grew 14.3% on the new basis. Take LinkedIn out of both years and growth is about 11.5% ($94.3 billion to $105.1 billion). That’s still strong. But about a fifth of the headline growth was bought, not built.
Cash: The Series That Never Needed Restating
Revenue accounting moves numbers between years. Cash doesn’t care. Dollars in millions.
| Fiscal year | Net cash from operations | Additions to property and equipment | Operating cash minus capex (my arithmetic) | Capex as % of operating cash |
|---|---|---|---|---|
| 2013 | 28,833 | 4,257 | 24,576 | 15% |
| 2014 | 32,231 | 5,485 | 26,746 | 17% |
| 2015 | 29,080 | 5,944 | 23,136 | 20% |
| 2016 | 33,325 | 8,343 | 24,982 | 25% |
| 2017 | 39,507 | 8,129 | 31,378 | 21% |
| 2018 | 43,884 | 11,632 | 32,252 | 27% |
Fact: Fiscal 2013–2015 figures are from the fiscal 2015 10-K cash flow statement; fiscal 2016–2018 are from the fiscal 2018 10-K. The fiscal 2018 10-K’s comparison of fiscal 2017 with fiscal 2016 says “Cash from operations increased $6.2 billion to $39.5 billion during the fiscal year, mainly due to an increase in cash received from customers and an income tax refund for overpayment of estimated taxes, offset in part by an increase in cash paid to employees”.
Inference: Operating cash is the one series in this window that lines up with neither the scary revenue dip nor the scary profit dip. It dipped in fiscal 2015, the Nokia integration year, then rose every year: up about 51% from fiscal 2015 to fiscal 2018. Part of fiscal 2017’s gain was a tax refund, by the company’s own description, so don’t credit all of it to operations. At the same time capex roughly doubled, from $5.5 billion in fiscal 2014 to $11.6 billion in fiscal 2018. The 10-K lists “datacenter costs in support of our cloud-based services” among its most significant expenses. The cash went where the cloud revenue came from. Cash statements from different 10-Ks can carry small reclassifications, so treat the year-to-year comparisons as approximate.
The Attribution Problem: Who Built the Cloud?
This is where the filings stop helping a turnaround story and start pushing back.
Fact: The fiscal 2014 10-K’s executive officer section: “Mr. Nadella was appointed Chief Executive Officer in February 2014. He served as Executive Vice President, Cloud and Enterprise since July 2013. From 2011 to 2013, Mr. Nadella served as President, Server and Tools.”
Company assertion: Microsoft’s Feb. 4, 2014 press release announcing the appointment says that “Since joining the company in 1992, Nadella has spearheaded major strategy and technical shifts across the company’s portfolio of products and services, most notably the company’s move to the cloud and the development of one of the largest cloud infrastructures in the world supporting Bing, Xbox, Office and other services.”
Fact: The fiscal 2014 10-K, filed after the appointment but covering a fiscal year that began seven months before it, already lists “commercial cloud-based services such as Microsoft Dynamics CRM Online, Microsoft Azure, and Office 365.”
Inference: The cloud build predates the CEO title. The company’s own announcement credits the incoming CEO with that build in his previous jobs. So the filings support “the cloud business grew fast in fiscal 2016–2018.” They do not support “the CEO change caused the cloud business.” The same person running the same strategy with a bigger title is continuity, not a turnaround, at least on this one line. What the CEO change may have changed (capital priority, the decision to stop pouring money into phones, how fast the cloud was funded) is a question about decisions inside the company. Annual reports don’t answer it.
Inference: The same logic runs in reverse for the Nokia charges. The deal was announced in September 2013 and signed before the CEO change. Blaming the $7.5 billion write-down on the new CEO is as unsupported as crediting him with all of the cloud. What the filings do put on the new team’s watch is the decision to write the asset down within about 14 months and restructure the business, rather than carry it at the purchase value.
Disconfirming Evidence
- The cloud predates the CEO. Commercial cloud services were already listed in the fiscal 2014 10-K, and the CEO ran Cloud and Enterprise from July 2013 and Server and Tools from 2011. Any “CEO saved the company with the cloud” story has to explain why it isn’t simply continuity.
- The Nokia deal was inherited. Announced Sept. 3, 2013, five months before the CEO announcement. Its write-down is not evidence of the new CEO’s judgment either way, except for the speed of the write-down and exit.
- Profit didn’t recover in years one to three. Before the identified items, operating income was about $27.8 billion in fiscal 2014 and about $29.3 billion in fiscal 2017. Most of the visible improvement over 2014–2017 is the absence of charges, not new earnings. The real gain is concentrated in fiscal 2018.
- Some growth was bought. LinkedIn ($27.0 billion) supplied about a fifth of fiscal 2018’s revenue growth.
- No peers. I didn’t build a matched peer set, so nothing here says whether Microsoft did better or worse than the market for enterprise software and cloud services in those years.
Alternative Explanation
Editorial: A credible alternative reading: Microsoft didn’t have a turnaround in fiscal 2014–2018 at all, in the sense of a business that was failing and then recovered. It had a profitable, cash-generating core that never stopped producing (operating cash never fell below about $28.8 billion in any year in the table), an acquisition that failed and was written off, a change in how one product was billed that temporarily hid revenue, and a cloud business that kept compounding along a path set before the CEO change. On that reading, the leadership change mattered mostly through capital allocation, cutting losses on phones and funding datacenters. Those are real decisions. But they’re a smaller and different claim than “turnaround.”
Sensitivity: Does the Conclusion Change With the Filing You Pick?
| If you use… | FY2016 looks like | FY2017 looks like | Does “turnaround” survive? |
|---|---|---|---|
| Old-basis revenue only (FY2017 10-K table) | −8.8% revenue | +5.4%, still below FY2015 | No. Looks like a decline in the new CEO’s second year. |
| New-basis revenue spliced onto old FY2015 | −2.6% | +5.9% | Barely, and the comparison is invalid across the break. |
| Old basis plus the disclosed Windows 10 deferral and FX | Small growth (about +$2.1B in constant currency, my estimate) | About $96.7B before the deferral | Yes, as modest growth, not a rebound. |
| Operating income before company-identified items | About $27.2–27.9B | About $29.3B | Flat through FY2017; real gain in FY2018. |
| Commercial cloud revenue | $9.5B | $14.9B (+57%) | Strong growth, but continuity with pre-2014 strategy. |
| Operating cash flow | $33.3B | $39.5B | Steady improvement after a FY2015 dip; FY2017 includes a tax refund. |
Inference: The verdict flips between “decline” and “growth” depending only on which filing’s revenue you pick. It doesn’t flip on the cash or cloud series. That’s the practical rule: when an accounting change splits the story, anchor on the series the change didn’t touch.
Why No Grade
Stagnation Assassins’ CEO rubric grades decisions within a defined window. It requires a peer set chosen before looking at outcomes, separate durability panels, and at least 85% evidence coverage, with operating economics, cash, and governance all evidenced. This piece has none of the peer set, no customer-outcome evidence beyond revenue, and no post-succession panel (the CEO in question was still in office at the end of the window). So the rubric says withhold, and the grade is withheld. This is a comparability test, not a verdict on Satya Nadella or Steve Ballmer, and it makes no judgment about either person.
The Lesson for Your Own Board Deck
Editorial: In 10-Minute Transformation, Chapter 2, Todd Hagopian makes a point that applies directly here: when a one-time charge, an acquisition-accounting effect, or a mix anomaly distorts a single year, strip it out if you can identify it and judge the business on the multiyear trend. Microsoft’s fiscal 2014–2018 filings are close to a laboratory demonstration of why. In five years you get a $10.0 billion acquisition charge, two years of Windows 10 deferrals totaling $13.3 billion, a $13.7 billion tax charge, a $27.0 billion acquisition, and a full retrospective restatement. Any one of those, read in isolation, produces a false headline.
Three checks before you call anything a turnaround, yours or anyone’s:
- Find the break. If the accounting changed, find the year it changed and never draw one line across it.
- Read the footnotes on the scary years. In this case, every “bad year” headline had a footnote that sized the distortion as it happened.
- Anchor on what the change didn’t touch. Cash from operations and a consistently defined product series beat any headline revenue line when the rules move.
For the same discipline applied to other companies on this site, see five evidence tests before you applaud a turnaround, a graded look at Gerstner’s first three years at IBM, and how Toyota’s fiscal 2010 recall costs break down in its filings.
FAQ
Did Microsoft’s revenue really fall in fiscal 2016?
On the basis reported at the time, yes: from $93,580 million to $85,320 million. Microsoft’s own fiscal 2016 10-K said that was “primarily due to” a $6.6 billion Windows 10 revenue deferral and about $3.8 billion of unfavorable currency. Restated under the new revenue standard in the fiscal 2018 10-K, fiscal 2016 revenue is $91,154 million.
Why are there two different revenue numbers for the same year?
Microsoft early-adopted a new revenue recognition standard on July 1, 2017, using the full retrospective method. That moved Windows 10 revenue to the time of billing and delivery instead of spreading it over the device’s life, and restated fiscal 2016 and 2017. The 10-K says the change had no effect on cash flows.
How much did Microsoft write off on Nokia?
In fiscal 2015, $7.5 billion of goodwill and asset impairments related to Phone Hardware ($5.1 billion goodwill and $2.2 billion intangibles, per the 10-K text), plus $2.5 billion of integration and restructuring expenses, for $10.0 billion in total. The business had been bought for a total purchase price of $9.5 billion, including $1.5 billion of cash acquired.
Is commercial cloud “annualized run rate” the same as revenue?
No. Microsoft defined the run rate as the last month of the quarter’s revenue multiplied by twelve. For fiscal 2016 the run rate exceeded $12.1 billion, while actual commercial cloud revenue, reported later, was $9.5 billion.
Did Satya Nadella create Microsoft’s cloud business?
The filings show he ran Server and Tools from 2011 and Cloud and Enterprise from July 2013, before becoming CEO in February 2014, and Microsoft’s announcement credited him with leading the move to the cloud in those roles. So the cloud build predates his time as CEO. The filings can’t separate how much of the later growth came from the CEO change itself.
Limits
This is a reconstruction from public filings and two company press releases, read from outside the company. The fiscal 2014–2015 rows are on the pre-2018 revenue basis and fiscal 2016–2018 rows on the new basis where noted. The constant-currency residual for fiscal 2016 is my arithmetic on two company estimates. Add-backs use only items the company itself identified in table footnotes; other items may exist that I haven’t adjusted for. Cash flow figures come from two different 10-Ks and may reflect small reclassifications. There’s no peer comparison, no customer-outcome data, and no attempt to measure culture or anything else the filings don’t report. Stock price isn’t used. No one is graded, and nothing here makes claims about anyone’s motives. This is educational analysis, not investment advice.
Run the Test on Your Own Business
- Book Todd to speak: bring this comparability check to your board or leadership team. toddhagopian.com/speaking
- Read the books: Stagnation Assassin, plus 10-Minute Transformation (Koehler Books, release date February 23, 2027). toddhagopian.com/book/
- Free tool: find out whether your business is an ATM or a money pit with the ATM Test Scorecard.
- Want to read 10-Minute Transformation before launch? Join the launch team.
Sources
- Microsoft Corporation, Form 10-K for fiscal year ended June 30, 2014 (Item 1 cloud services description; executive officers, Nadella appointment and prior roles; Note 9 NDS purchase price $9.5 billion; five-year table). SEC EDGAR
- Microsoft Corporation, Form 10-K for fiscal year ended June 30, 2015 (five-year table and footnote (a), $10.0 billion Phone charges; goodwill impairment $5.1 billion and intangibles $2.2 billion; Phone Hardware Integration Plan, ~19,000 positions; Phone Hardware Restructuring Plan, up to 7,800 positions; cash flow statement FY2013–FY2015). SEC EDGAR
- Microsoft Corporation, Form 10-K for fiscal year ended June 30, 2016 (Windows 10 net revenue deferral $6.6 billion; revenue decrease $8.3 billion and ~$3.8 billion FX; $1.1 billion phone impairment and restructuring; commercial cloud annualized run rate and definition). SEC EDGAR
- Microsoft Corporation, Form 10-K for fiscal year ended June 30, 2017 (five-year table, old basis; Windows 10 net revenue deferral $6.7 billion; $306 million severance; run rate above $18.9 billion; LinkedIn acquisition $27.0 billion; ). SEC EDGAR
- Microsoft Corporation, Form 10-K for fiscal year ended June 30, 2018 (five-year table restated for the new revenue standard; Note 1 adoption, full retrospective method, Impacts to Previously Reported Results, no cash flow impact; $13.7 billion Tax Cuts and Jobs Act charge; commercial cloud revenue $23.2B/$14.9B/$9.5B; LinkedIn revenue; cash flow statement FY2016–FY2018 and FY2017 cash commentary). SEC EDGAR
- Microsoft, “Microsoft to acquire Nokia’s devices & services business, license Nokia’s patents and mapping services,” press release, Sept. 3, 2013 (announcement date; company source). Microsoft News
- Microsoft, “Microsoft Board names Satya Nadella as CEO,” press release, Feb. 4, 2014 (appointment date; company description of prior roles). Microsoft News
- Todd Hagopian, 10-Minute Transformation, Chapter 2 (strip identifiable one-time items and anchor on the multiyear trend). Koehler Books, release date February 23, 2027. Paraphrased, not quoted.
Calculations are mine, from the filed figures cited. Educational analysis only.
