Maybe Bing isn't trying to compete with Google after all
Earlier this year, Bing Chat was set to change the search game for good. Six months later, pundits are using statistics to write off Microsoft's big AI initiative. But look more closely and those numbers tell a different story.
"Bing Chat actions could be a game changer," we were told. "[W]e might be witnessing the start of a new era -- a renaissance, one might even call it 'a Bing-aissance,'" someone wrote, apparently without cringing. "Bing is king!" was an actual headline on an actual website that was not The Onion.
And then, six months later, the Wall Street Journal came along and let all the air out of that balloon with a story that called the new Bing with AI chatbot "cute, but not a game-changer." At ZDNET, Lance Whitney concluded that "the new capabilities of Bing have failed to jumpstart Microsoft's cut of the search engine market." As proof, he cited the same numbers that the WSJ used:
So, what do the numbers mean? As I have mentioned many times over the years, I think StatCounter does an admirable job at identifying big trends, such as the shift from desktop PCs to mobile devices and the relative position of web browsers.
But those StatCounter numbers do not tell us anything at all about who has what share in the search engine market. And don't just take my word for it. Go read this disclaimer on StatCounter's FAQ:
Is bing chat included in search engine share?
We have no way of measuring the number of queries performed in bing chat. However, we also don't measure the number of queries to regular search engines like bing or google either. Instead we track search engine referrals.
i.e. If you go to bing.com and do a search for anything and you click on a website result, we'll record that click as a search engine referral if that website had the statcounter code installed. It's the click to a website that we measure, not the actual search queries that were performed.
Well, that context changes everything, doesn't it? StatCounter is measuring the last click -- the one that took you away from the search engine -- and not counting any of the activity on the page itself. What the StatCounter data tells us is that among sites that use the StatCounter tracking service, the impact of referrals from Bing worldwide, on all platforms, hasn't changed dramatically during the past six months.
But even that statement needs some qualifying. Bing Chat only works in the Microsoft Edge browser or with the Bing app on mobile devices. That makes it a non-starter for people using the world's most popular browser, Google Chrome, or who use Safari, the default browser on Apple's mobile devices. With those roadblocks, the worldwide all-platforms numbers are unlikely to move much at all.
You can play around with some of StatCounter's chart-making tools to see different segments at work. For example, if you exclude mobile and tablet platforms, and look only at their desktop search statistics in North America since the Bing Chat rollout in February 2023, referrals from Bing are up 2.2%, while those from Google are down 2.8%.
I don't advise taking any of those numbers too seriously, though.
The WSJ also quoted SimilarWeb, which is in the business of selling big data reports to publishers, marketing agencies, investment bankers, consultants, and the like. On its FAQ page, "How We Measure the Digital World", the firm talks about generating more than 10,000 reports daily.
If you dig deep enough, eventually you get to the "Search Engines Market Share" page, which offers the same sort of customization tools as StatCounter. According to their numbers, Bing in June 2023 had a 3.23% share on all platforms worldwide. That number shoots up to 8.79% if you look only at desktop platforms worldwide -- and even more, to 12.55%, if you look only at traffic from desktop machines in the United States. Similarweb offers a line graph, but none of the data points are labeled, so you can't really tell whether there's been any movement during the past year.
There's so much noise in all those numbers that it's really hard to find the signal, which is why the chart on search engine market share contains this disclaimer:
This traffic and engagement data is based on aggregated and anonymized first-party analytics shared with Similarweb by millions of websites and apps and represents more than 500 billion pageviews. This data is for a subset of sites and as a result, is an estimation of the market share.
A third analytics firm quoted by the Journal is YipitData, which doesn't offer any information on how its numbers are collected. Three of the six most recent articles on the firm's blog are about bathroom renovation, and there's little else on the page to suggest they have any special insights into AI or search.
For Microsoft, a gain of even half a percentage point of worldwide usage for its search products would be huge. But without knowing the margin of error, it's hard to have much confidence in any of the numbers that the Journal is treating as gospel. And what's more, I think they are probably missing the real story.
Again, don't just take my word for it. In a confidential 2022 strategy document that was accidentally leaked in a court case, Microsoft CEO Satya Nadella discussed Microsoft's Search, Advertising, News, and Edge ("SANE") segment:
We capture ~2% of the $560 billion core [Total Addressable Market] in SANE. With $10 billion in revenue, we are a large digital advertising business, and our products are competitive. ...
Increasingly we are succeeding in closing the product experience gap with our large competitors as measured by objective user feedback. In some cases our product experiences are rated the best. However, our share does not reflect this as there are a range of structural impediments each of which is hard to overcome. However, we are seeing steady improvements over time.
The real goal, Nadella's memo continues, is "a step change in the scale of our advertising business." And one effective way to accomplish that is to "drive browser share within our install [sic] base This is a high impact lever for us given the size of our install base and even small gains from 'grinding the funnel' can result in large operating margin gains."
To put that suggestion another way: Maybe Microsoft isn't trying to compete head-on with Google search after all.
Microsoft's revenue from search and advertising would be a rounding error at Google, which pays twice that amount every year to Apple alone as a bounty for sending search traffic on iPhones, iPads, and Macs to Google. But it's a decent side business in Redmond, with plenty of room to grow, especially by "grinding the funnel" on a billion or more Windows PCs to pester those customers into switching to Edge. Annoying? Absolutely. But also a surefire source of incremental revenue.
Microsoft's goal with Bing and Edge isn't to take over from Google as the world's biggest search engine. The goal is to scale up an advertising business that is already paying its way as yet another segment that earns at least $10 billion in revenue for Microsoft every year. As a famous Senator might once have said: "A billion here, a billion there, pretty soon you're talking real money."