Microsoft has announced how much users will have to pay for Copilot, the generative algorithm now included in Office that it announced with a video a few months ago showing how to create texts, presentations, spreadsheet formulas, etc using machine learning assistants. But the $30 the company has added to the price of the Office license has angered many.
Thirty dollars per user per month is a major investment for many companies, meaning many will not make them available to all their workers. So why has Microsoft, a company that tends to think in the long term and in the advantages of mass adoption of its tools, set such a high price?
The answer is very simple, but for many, surprising. When OpenAi launched Dall-E in January 2021 and ChatGPT at the end of November 2022, they were free. ChatGPT soon set a record for the number of unique users it immediately attracted. Being able to have a conversation with a massive language model was obviously very attractive, and made headlines around the world.
What was ChatGPT doing? Simply applying the Silicon Valley philosophy: supported by funding rounds, it leveraged its growth to gain unprecedented brand recognition and company valuation, while accumulating debts with Microsoft through its use of Azure. But the company knew that if it achieved a pioneer’s positioning it could pay those bills, and that the mass use of its algorithms allowed it to continue developing them with users’ data by taking advantage of the absence of legislation in this regard.
For ChatGPT, and later for Microsoft, which leveraged its early investment in the form of Azure credits to take a stake in the company, offering Bing searches supplemented by ChatGPT was no big deal: the search engine was very small, with a market share of about 3%. Bing usage would skyrocket and start to generate more costs was a situation every company in the world would like to be in.
It was a different story for its competitor, Google. If it included a generative algorithm in its products, it had a potential credibility problem: if Bing hallucinated and provided a bizarre answer, that’s fine, it’s an experiment by a small competitor. But if Google’s search engine did it, the reputational and trust cost was potentially much higher. In addition, Google’s business model, based on clicks on sponsored results, could suffer severely: Bing generating less business is not terrible for Microsoft’s bottom line, but every percentage drop in the number of ad clicks on Google is reflected in the company’s revenue… and users, when instead of getting ten links to click on, were given a perfectly worded paragraph with the answer to what they are looking for, tended to stop there.
Last but not least is the question of cost. For a search engine, running a generative algorithm is appreciably higher than a simple search on a highly optimized database. Again, Microsoft could assume this in the case of Bing, given its small market share, but for Google, with its 8.5 billion daily searches, was a huge drain on resources.
But if Microsoft now wants to include OpenAi’s product, Copilot, in Office, we are no longer talking about a small minority search engine with a small market share like Bing. Instead, we are talking about one of the company’s flagship products, used by some 345 million licensed users, with more than 958,000 daily active users. The question of all these people happily making use of generative algorithm queries every day is no longer a trivial matter: given the cost of these queries, it is potentially a very important part of the company’s costs. Therefore, a price must be put on it, and that price must be sufficient to anticipate high usage. Now, it is Microsoft’s turn to deal with the problems that Google had at the time, and which conditioned its slow and disappointing response. The multiplier effect of a bad cost analysis applied to the huge Office user base is very important, and the calculation must be very careful so that the offer is sustainable, increases the attractiveness of the product, but does not become a significant drain on resources. The result? $30 per user per month.
The problem is that OpenAI’s launch (which subsequently priced its license and provided only a limited version for free), means many people expect generative algorithms gratis, without even having to provide any kind of bank details. And that makes a license fee of thirty dollars per user per month an unpleasant surprise. For Microsoft, moreover, the corporate market is absolutely fundamental: one slip-up, and another competitor with a smaller market share that is willing to offer generative algorithms could consider throwing its hat in the ring.
In short, generative algorithms cost a lot of money to run, although costs may fall as companies learn to apply economies of scale. We’ll have to see how the market responds, how companies react, and whether businesses are prepared to pay serious money to use generative algorithms.
—
This post was previously published on Enrique Dans’ blog.
***
Join The Good Men Project as a Premium Member today.
All Premium Members get to view The Good Men Project with NO ADS.
A $50 annual membership gives you an all access pass. You can be a part of every call, group, class and community.
A $25 annual membership gives you access to one class, one Social Interest group and our online communities.
A $12 annual membership gives you access to our Friday calls with the publisher, our online community.
Register New Account
Need more info? A complete list of benefits is here.
—
Photo credit: iStock