As Microsoft has added more consumer subscribers to the rolls of Office 365, the amount of revenue per subscriber has fallen, according to estimates based on the limited information the company releases.
The downward trend in revenue-per-subscriber hints at a significant shift to the less expensive rent-not-own plans available to consumers, such as Office 365 Personal, a $69.99 annual deal that provides one Office license for a personal computer and one license for a tablet.
In the fourth quarter of 2014 -- the most recently reported -- Microsoft said it had 5.7 million more paid subscribers than the same period the year before, and that revenue had increased year-over-year by $97 million.
On average, then, each of the 5.7 million additional subscriptions generated $17.01 during the quarter. But because Microsoft recognizes that revenue over the period of a subscription -- in other words, in four equal allotments in the course of a year -- those subscribers were each worth an average of $68.04 ($17.01 X 4) annually.
That was the lowest amount over the last four quarters, according to the figures Microsoft has provided in its filings with the U.S. Securities and Exchange Commission (SEC).
For example, in the third quarter, the average per-subscriber revenue was $19.02 (or $76.08 for a year), while the second and first were $25 ($100 annually) and $22.27 ($89.09), respectively.
(Jan Dawson, chief analyst with Jackdaw Research, provided revenue estimates for the June quarter, as Microsoft closed its fiscal year 2014 at the end of that quarter, and so used its SEC filing to spell out the year's numbers rather than those for the three-month stretch.)
During the year, Microsoft's average per-subscriber revenue from its consumer Office 365 plans dropped 24%.
Admittedly, Computerworld's calculations are sketchy: They only account for the year-over-year growth in both subscribers and revenue, and so don't provide a complete picture of the average revenue for all subscribers. Also in play are Microsoft's number rounding and the fact that consumers are subscribing throughout each three-month stretch, leading to potential artificial fluctuations if, say, more customers sign up near the start or end of a quarter.
The latter may have been the cause of the second quarter's uptick in revenue per subscriber. Office 365 Home, which debuted in late January 2013, only began collecting significant numbers of subscribers in the June quarter. If most subscribed early in that quarter, then renewed a year later, in Q2 2014, the revenue would have skewed higher since more than the usual number would have paid throughout most of the period.
Office 365 Home gives buyers the right to install the Office suite on up to five Windows PCs or Macs, and on as many as five tablets. It lists for $99.99 annually or $9.99 monthly.
Still, the general trend was clear: Average per-subscriber revenue declined in the second half of 2014.
That's no surprise.
Microsoft launched Office 365 Personal in April. By the numbers, many consumers opted for the cheaper plan, as the quarter-over-quarter declines in the third and fourth quarters were 24% and 11%, respectively, implying a mix that increasingly skewed toward Personal.
The precipitous drop in the third quarter may also have been driven by the March launch of Office for iPad. Throughout Q3, and into part of Q4, the least expensive way for iPad owners to get editing functionality for those apps was Office 365 Personal. (Microsoft freed more features in Office for iPad in November.)
Other possible explanations for the decline in per-subscriber revenue include increased sales of Office 365 University -- the for-college-students-only plan that costs just $79.99 for four years -- which Microsoft includes in the "consumer" group; customers signing up for Office 365 on the monthly plan but not sticking with the subscription for a full year; and more aggressive discounting by Microsoft. All would impact revenue.
On the latter, Microsoft does discount Office 365; it doesn't collect the list amount for each Office 365 subscription. Amazon.com, for instance, sells a "key card" -- the retail offer, which includes a registration code customers use after they've downloaded the applications -- for $56 (Personal) and $70 (Home), discounts of 20% and 30% from the all-online deals.
Interestingly, Q4's average per-subscriber revenue of $68.04 was actually lower than the list price of the cheapest plan, the $69.99 Office 365 Personal, a strong indication of discounting, a preponderance of one-license subscriptions, large sales of Office 365 University, or a combination of the three.
It may also be a hint of a high Office 365 Home "churn rate" -- the number of customers who drop a subscription -- since those paying $99.99 annually would probably pay the same for their second, third and following years.
How much lower can Microsoft's per-subscriber revenue fall? Good question.
The slowing of the decline in Q4 -- at 11%, less than half the sequential drop of 24% in Q3 -- may mean the bottom is within sight. Failing a major price cut or extreme discounting, the largest downturns should be behind Microsoft.
The company has taken steps to, if not slow the decline in per-subscriber revenue, then to retain those customers so that it can collect more over the lifetime of a subscription. In June, it upped the OneDrive storage space for Office 365 subscribers to one terabyte (1TB), then three months later removed all restrictions, offering unlimited storage.
Whether moves like unlimited OneDrive can boost retention, whether Microsoft can even make money from consumer subscriptions, is still unknown. Most analysts are convinced that Microsoft has a hard row to hoe when the competition, including Google Docs and Apple's iWork, is free. Of the major productivity application providers, only Microsoft charges consumers for its wares.
"Microsoft is feeling pressure from the bottom end of the productivity market," said Wes Miller of Directions on Microsoft in an interview last year.
This story, "What's a consumer Office 365 subscription worth to Microsoft?" was originally published by Computerworld.