I was working at Microsoft/Tellme at the time Bing411 was launched and would like to state a few facts:
- When launched, Bing411 was essentially a re-branding of Tellme's 1-800-555-TELL service.
- 1-800-555-TELL has been running since 2000 or something like that. I attended the 10th anniversary party, which I believe was in 2010.
- Microsoft didn't need to launch a Microsoft-branded 411 to collect utterances, for they had acquired Tellme and had many utterances coming in under their umbrella of services
- Tellme's bread and butter was speech recognition on the telephone and they are still in that space today, so Bing411, 1-800-555-TELL are still relevant for them.
I can also say that a lot of the speculation in the article about Microsoft's "reaction" is just wrong. But, I won't go into specifics because they aren't relevant.
I enjoyed the author's original point about making sure you are playing the same game. Unfortunately, the example used is factually inaccurate.
Don't all those points fit into the story in that blog entry? No one is saying that Microsoft needed voice data. From e40's comment: "I think the point of the article is that they don't need to mine utterences with Bing 411, and that they merely did the service to compete with Google, without knowing why Google made the service in the first place." Isn't the speculation about Microsoft's reaction the heart of the issue? If Microsoft did buy Tellme's service so that they could compete with Google's that would explain how Microsoft wasn't playing the same game Google was.
Edit: I do agree that the blog post is all speculation. I'm just saying that I don't think those points prove that it is wrong.
From that article: "This is actually a product that Google’s been testing in various formats for some time. Steve Poland (a regular contributor here) is pointing me to some posts (and here) by Greg Sterling from last year that discuss this. The earliest reports on this are from October 2006, and the service may be from an acquisition of 1-877-520-FIND. More information here." Again though.. Pure speculation.. It's not implausible because big companies do keep tabs on their competition. Also not sure if Microsoft's history with their product goes back even further.
If I were the product manager of a search engine and our chief competitor had an offering (411) that we didn't (at least branded in our name) and the cost to deploy said offering was cheap (because we already had one very similar due to a recent acquisition), wouldn't it be worth the relatively trivial investment (from a large company's perspective) to deploy said offering? Even if it were for reasons of "me too," so what? I believe most brand managers would reach the same conclusion, especially if the costs were low.
Furthermore, does the track record not show that Microsoft re-brands (assimilates) its acquisitions eventually? How do you (or the author of the linked post) know that the deployment of a Microsoft-branded 411 service was not part of this process?
Finally, everyone in the speech recognition industry knows that it requires a large sample set (utterances) to refine a speech recognition engine. If you are building one from scratch (Google), you will do anything and everything in your power to collect utterances. To think that Microsoft and everyone else in the industry did not recognize what Google was doing with GOOG-411 is preposterous.
Given the article that someone cited in their reply to me on this comments page that said that Google has been launching it's product in many forms since October 2006, it still seems possible that they made the decision to buy that company because of Google's actions. As far as everyone knowing what Google was doing, I don't think that is true. Didn't a lot of people think that they were really just trying to get into the 411 business? I mean before they actually launched the product in their labs thing and stated their intention.
The tone of this article bothers me. There are some smart people working at Microsoft. The level of discussion is higher than "Hey, google made a thing, we need the same thing."
Like the article points out, Microsoft has no shortage of utterances due to their acquisition of Tellme. Why, then, do they need to mine utterances with Bing 411? (Assuming they don't--I'm sure they do.) Bing 411 is a genuinely useful service that adds value to the Bing brand. It needs no additional reason to exist.
All this aside, the article is factually incorrect. It implies that localeze was contracted to provide Bing 411. In fact, localeze provides listings (according to the press release the article links), and are presumably one of many listing providers used. Microsoft/Tellme provide the actual Bing 411 service. I don't know for sure, but Tellme's (now Microsoft's) 1-800-555-TELL number almost definitely predates GOOG-411.
> Why, then, do they need to mine utterances with Bing 411?
I think the point of the article is that they don't need to mine utterences with Bing 411, and that they merely did the service to compete with Google, without knowing why Google made the service in the first place.
I think Microsoft has very smart people but they don't have much control and the people at the top are more marketing and less genius and innovation. There is no shortage of examples of Microsoft doing just what you said, do it poorly and it flops. Microsoft's big successes where mostly do to masterful marketing, and less because the product was truly innovative. All that in mind, Photosynth still blows my mind.
Yep, not saying it was a secret... Tim O'Reilly had it nailed back then... but the story I have to remind myself of is that too many times we go off face value rather than digging in and listening to people like Tim... This happens all over the place and it was worth capturing the story that replays in my head each time I see a new revision of it.
well, Marissa Mayer of Google explicitly stated in a public announcement in October 2007 that it was going to be used for speech recognition algorithms. And many, many news blogs published that.
Your article makes it seem like the fact that Google was using 411 for voice analysis was somewhat 'unknown' until 2009, whereas I would say that the knowledge was actually quite mainstream and, frankly, hard to ignore.
I would argue that assuming that the Bing team completely failed to see all the news articles and press releases (from Google, no less) about this for 2 years, is perhaps a little presumptuous? Maybe, as another poster mentioned, they are still getting value out of the 411 service?
I do agree with your point, and there are many many times I've seen companies do this, but Bing 411 vs Google 411 might not be the right example since Google's intentions were very clear, right from the start.
Disclaimer: I work on MS Office (completely unrelated to Bing & the 411 service, but still not what you would call an unbiased source)
In any case, it was really well known what Google's goals were. In fact it was a pretty standard feedback technique used by all automated phone voice reoognition systems.
For Bing though this is also a branding opportunity. Google has relatively little to gain in branding with their 411 service, but for Bing it still has great upside potential.
While I cannot speculate on Microsoft purpose of a 411 service, I think it is pretty obvious why google dumped theirs - they now have access to an enormous source of human speech through google voice.
Yeah. The voicemail transcription feature truly is like an open invitation to provide them with real-world data across a wide range of accents, phones, and people.
More so, it's not a very good business to be in. The same functionality at a much, much higher profit margin is in smart phones, and that's where you want to put your effort because it has such a higher rate of return.
I'll play devil's advocate here. Are we sure Microsoft is not doing the same data mining on Bing411 as Google did on Goog411, and haven't dumped it because it is either a valuable service in their eyes or they are not done gleaning what google gleaned from their head start?
I somehow doubt this is the case but I'm not so sure we can assume that it is not.
Microsoft doesn't play the same sort of long game that google does. Microsoft plays a traditional long game: market penetration, brand loyalty, development stack, "synergy" (real or imagined), etc. Google plays a much more subtle game. They'll do something like GOOG411 just to get better voice recognition algorithms. They'll also do things like develop a state of the art web browser just to jostle the browser development community out of its laziness and propel web standards and browser performance forward. That sort of thing is just not in Microsoft's DNA to do or even conceive of doing.
It still seems really odd to me that Google 86'd all explicit phone number directory -related stuff. The "phonebook:" operator is no more either. Yes, I know traditional web search and Maps are pretty good most of the time, but I come across many cases where they are not. Especially when it's a residential phone number (that is otherwise publicly listed) or a less web savvy local business.
Phone directories may, in fact, be very antiquated but they're still a thing, the data is there and often useful. Why not index it? Why not let me expressly look-up a phone number anymore?
Interesting read. But the best part for me was learning about a new free 411 service. At least once a week, I start to call Goog411 before sadly remembering the service was discontinued.
Is it really an interesting read? (this isn't a criticism of bimbly, more at poor quality blog posts in general (a particular bee in my bonnet at the moment))
The blog post basically consisted of a made up example to make a point that the blog author felt was valid. Now I happen to agree with the basic point the author was trying to make, but I think it's a reflection on the really poor quality of so many blog posts where people have some idea they want to pitch as insightful and then they fail to do any actual research and/or wrap it in a made up story. The 'parable' is pretty worthless when people like Indygreg above state that the motivations that the author guessed at was completely wrong.
Even if we accept the very dubious premise that Bing411 was motivated entirely by a misunderstanding of Google's aims for Goog411, so what? A good cautionary tale needs to demonstrate the dire consequences of whatever you're cautioning against, but you haven't actually showed anything bad happening to Bing or Microsoft as a result of their alleged confusion. Really not the best vehicle for your moral.
- When launched, Bing411 was essentially a re-branding of Tellme's 1-800-555-TELL service.
- 1-800-555-TELL has been running since 2000 or something like that. I attended the 10th anniversary party, which I believe was in 2010.
- Microsoft didn't need to launch a Microsoft-branded 411 to collect utterances, for they had acquired Tellme and had many utterances coming in under their umbrella of services
- Tellme's bread and butter was speech recognition on the telephone and they are still in that space today, so Bing411, 1-800-555-TELL are still relevant for them.
I can also say that a lot of the speculation in the article about Microsoft's "reaction" is just wrong. But, I won't go into specifics because they aren't relevant.
I enjoyed the author's original point about making sure you are playing the same game. Unfortunately, the example used is factually inaccurate.