Local cityguide Citysearch has launched a major overhaul of its site and user experience. I discuss the desktop portion of the site on my blog Screenwerk. There are more social features (in part via the capacity to log in using your Facebook ID) and the content is now searchable by neighborhood.
Citysearch mobile (which had been powered by Earthcomber and may still be [haven't confirmed]) is taking the medium much more seriously (press release). It will use device detection to optimize the experience for different mobile handsets and platforms and leverage location using a couple of methodologies. The company will also be releasing an iPhone app soon.
But the most significant aspect of the new Citysearch mobile experience, from my point of view, is the ability to write reviews from a mobile device in a highly simplified way. As we've written before more and more local user reviews will be coming through mobile devices and Citysearch is seeking to take advantage of that emerging trend.
Google Voice Search is apparently now finally available to download from the iTunes Store. However there's nothing different about the iTunes app that I can tell from the associated copy.
But here's the Google mobile page that offers a link directly to the iTunes store to download the updated app.
The "first impressions" are are now coming in quickly: some people love it, some are giving it a more qualified endorsement. So far no one is highly critical of it. Here's the complete blog roundup on Techmeme.
What people should remember is that this is the beginning of something . . . We'll have our own take in the near future.
We're still waiting for the Google app to update for voice search. Theories abound about why it hasn't yet been released. You can see the demo video on my personal blog Screenwerk. Here's what I wrote about it over the weekend after a conversation with Google's Mike Cohen and Gummi Hafsteinsson:
The first thing that is both intriguing and very different is that there are no buttons to push (although you can if you want apparently) to initiate voice search. Once the app is open on the iPhone you hold it up to the side of your head as though you were going to talk on the phone and simply speak the query. Search results then appear as they would if you had manually entered a query.
Google says it has learned tremendously from its experience with Goog411 but its desktop search query data is also contributing knowledge to the effort. These and other technical factors, beyond the scope of my expertise, will make the system more accurate that what has been possible in the past, said Google’s Cohen.
Beyond its reported accuracy, the usability of the system is striking. Most voice control on mobile handsets requires that buttons be pushed. There’s also often a “walkie-talkie” style experience, with the phone held out in front of the user to speak the query or command into the phone.
By removing the need to push a button and simply mimicking the experience and handset position of talking on the phone Google’s voice search may prove to be quite a bit more “natural” and intuitive. Another benefit for Google is that by having the phone’s receiver closer to the mouth of the person speaking the system gets a better, cleaner input.
After talking to Google I became hopeful that what Google is introducing will be a leap forward. If it indeed is we should see increased query volumes and longer query strings — and increased search monetization from mobile for Google. But all this will become more clear once the app launches -- hopefully today.
Chipmaker Qualcomm is incorporating Skyhook's various location awareness technologies into its new chipsets (gpsOne Platform). According to the press release this morning:
Skyhook's WPS is a software-only system that produces accurate location information by detecting Wi-Fi access points and comparing them against a known database of geo-located points. By combining Qualcomm's market-leading gpsOne solution – the most widely deployed A-GPS solution available – with Skyhook's location technologies, Qualcomm will enable future mobile device manufacturers, mobile operators, third-party service providers and application developers to utilize a single, integrated hybrid positioning solution.
I spoke to Skyhook CEO Ted Morgan about the announcement. He told me that the company's technology works best when it's integrated into the chipset, as with this annoucement. And the deal doesn't simply pertain to mobile handsets; it will include phones but also extend to laptops, cameras and potentially other mobile devices.
Morgan reminded me that "every CDMA phone has a Qualcomm chip in it." He also said that, combined with Skyhook's other relationships (e.g., Broadcomm), the deal gives Skyhook penetration into "north of 90% GPS supplier market."
We're entering a world where location is going to be available, at the chip-level, in the browser or via other mechanisms, to advertisers and publishers at almost every turn whether in mobile or on the traditional Internet. It may take a few years to realize this ambition but there can be no question of its inevitability.
Indeed, very soon users may start seeing pop-ups like this when they visit selected sites online:
Google is introducing a new app for the iPhone that will allow users to perform voice-based Internet searches. I'm trying to talk to Google about it to get a little more information than is available in the piece in the NY Times. Basically this appears to be like what Yahoo did with Vlingo for oneSearch earlier this year.
The first voice apps for the iPhone -- despite the fact that lots of companies have them in development or in the Apple approval pipeline -- were introduced by Dial Directions: Say Who and Say When. We wrote about both previously.
In the realm of "free DA" we have:
Then there's mobile ChaCha, which is a full Web, voice search engine (with humans in the background).
In terms of mobile voice clients, there's Microsoft's Live Search, Tellme, Google Maps for BlackBerry and, as mentioned, Vlingo for Yahoo! oneSearch. The proof, however, will come in using the new Google iPhone voice-search capability, which is based on the same recognizers as Goog-411.
There's a lot more to say about all this, including about some research we conducted on voice search and the consumer appetitie for voice, that is part of a forthcoming report.
For now, I've offered some additional thoughts in my post at Search Engine Land as well.
Google has introduced a new look for its search results pages for the iPhone and iPod Touch. Rather than simply reflect the desktop results as they appear online, they're more iPhone-friendly, often with telephone numbers and a directions link (where appropriate). The thumbnail to the right is an example of an individual listing page for a San Francisco Museum.
The new formatting affects images and other categories of results, but it's most pronounced in the case of local results. A "click" on the associated phone number triggers a call (potentially laying the groundwork for a PPCall ad model over time).
I did a search, for example, for "New York Hotels" and got a bunch of listings, with two ads at the top of the page. They were two out of the three ads that appeared on the desktop -- fewer ads but more likely to be clicked on given the new formatting. In several other searches I performed the ads shown at the top of the page were fewer in number than on the corresponding desktop result. In several cases there was only a single ad where there were two or three on the desktop (I wasn't able to discern a pattern, though I'm sure there is one). In certain cases the ad shown was either the top ad on the desktop or the ad in the second position.
This new formatting is currently only available in the US and it doesn't appear from all search "entry points" on the iPhone. If you, for example, search Google from another web page, you'll get search results that are identical to those on the desktop. You have to go to Google.com on the iPhone and conduct a search before you can see the new results.
TheFind announced its iPhone App this morning, as the challenging holiday shopping season begins. The App has been out for a few weeks. It taps GPS/location awareness on the iPhone and allows people to simply search for a particular item, brand or product category (e.g., children's furniture) and see a display of "nearby stores" or "Web items." TheFind crawls for data but also is getting a feed from NearbyNow (and maybe Krillion).
Other mobile-specific apps that provide local inventory data include Slifter and ShopSavvy (video demo), which features a barcode scanner. ShopLocal also has lots of data on retailers and sales that will be disseminated in mobile. ChaCha, I believe, is also working on this capability for those who don't have smartphones (e.g., "in San Diego, where can I buy a Wii?)
I've encountered skepticism recently when speaking to people about the coming popularity of mobile shopping apps and mobile sites that allow for in-store price comparisons and nearby inventory checks. But these activities and services will be very popular with smartphone users -- maybe more broadly if they're integrated into voice search or SMS services.
Their benefits are self-evident.
Amdocs has signed a definitive agreement to acquire Changing Worlds Ltd. (ChangingWorlds), a privately-held provider of software that supports personalization of mobile portals, for $60 million. The deal is set to close at year's end. ChangingWorlds and Amdocs already have a number of mutual accounts among the ranks of wireless carriers around the world, including Sprint, Vodafone and Telefonica/O2.
The plan is for ChangingWorlds' software to be integrated into Amdocs' Customer Experience System (CES). The software speeds the search process by capturing user behavior and usage patterns without requiring diret input from users. Mobile subscribers will be the initial beneficiaries but, like all the major search and content management rivals, Amdocs sees the acquisition as a component in a "three-screen" strategy that spans PCs and TVs as well as mobile phones and devices.
We are also pleasantly surprised by the number of reviews we have received from the phone compared to our website and consumers seem to find it very convenient to write a brief review while at the restaurant.
The following chart shows the distribution of their application's downloads by geography, which corresponds to the T-Mobile 3G rollout:
And here's the distribution and volume of search activity ("unique locations where users performed searches for local restaurants. These do not include searches by cuisine, keyword or location separately"):
I finally got a chance to hold and play with a G1.
I want to caution that it was not enough time to render a final judgment. Many things were nice about the device, but it was not as intuitive and elegant as the iPhone. In addition its greatest purported differentiator -- the presence of the physical keyboard -- is also a weakness in some situations. There appears to be no virtual keyboard (or that I couldn't find). And so when you conduct a general mobile search, you have to flip the device into landscape mode and enter the query.
It would just be faster if a virtual keyboard came up and you could do it that way. For SMS or longer text entries the keyboard works well.
Finally, I want to return to the BooRah comment about mobile reviews quoted in the passage above. The conventional wisdom is that people won't necessarily enter reviews when they're on mobile devices because of the awkwardness of the form factor. However I dispute this, especially if the reviews process is simplified (see GoodRec).
As I mentioned in the post immediately below, there's going to be a flood of local or location-specific content that will come online from people in physical places using mobile to capture and upload that information. Reviews is just one category.
ABI Research, which appears to issue a press release a week, put out a press release today that said:
Mobile location-based social networking is expected to become a key driver for the uptake of location-based services as it provides a unifying framework for a large set of applications such as friend finders, local search and geo-tagging. While many LBS applications will include features allowing the sharing of real-time experiences via fixed social networking sites such as Facebook and MySpace, fully-fledged mobile location-based social networking sites will also gain momentum with more than 82 million subscriptions expected by 2013.
While the company doesn't define (at least in the release) what it means by "subscriptions," this forecast has elements of truth but largely misses the mark in its emphasis.
The embedded location awareness in mobile handsets will inform and permeate a wide range of mobile applications and services. However most "stand-alone" mobile-social networks will fail. Existing desktop social networks (e.g., Facebook, MySpace, Bebo, etc.) will leverage location awareness in mobile and will generally crowd out upstart mobile only networks. There will be a few exceptions and success stories.
Conversely, most mobile apps will contain social elements to varying degrees (e.g., send to a friend, etc.). The thing not suggested in the release is the way in which location-aware handsets will impact destkop applications and local content available online. Indeed, this is arguably where the greatest impact of "LBS" will be: the tagging, reviewing and uploading of content, images and information obtained via mobile devices that can then be displayed online.
Mobile and the Internet more generally must be seen as distinct yet symbiotic. They are not two parallel universes but will increasingly interpentrate and overlap. Those publishers and applications that exist only on one side or the other will be at a tremendous disadvantage.
As far as "subscriptions" go, the overwhelming majority users aren't going to pay ongoing fees, to mobile social networks, unless it's a one-time-only payment for the app itself. Facebook and MySpace mobile are free and any new network, unless it's supremely niche/vertical/specialized, that seeks to charge users for access is DOA (as they say in the vernacular).
To promote its upcoming conference, The Kelsey Group put out findings from a new consumer survey on mobile usage. The survey was based on an online survey of 512 US mobile phone owners this October. The information is directionally accurate and the data are interesting, but I would argue with particular data points.
Here are some of the publicly released findings:
The Princeton-based firm also reported, "the percentage of mobile users who access the Internet from their mobile devices increased from 32.4 percent in 2007 to 38.9 percent in 2008, an annual growth rate of 20 percent."
We previously found (as did TMP Directional) using larger sample sizes that 16% had conducted local searches using mobile devices. This figure appears generally in agreement with the Kelsey data.
However, in my view, the mobile Internet access numbers and social networking numbers are too high to be representative of the mobile population as a whole. It's probably because of an overrepresentation of smartphones in the sample.
On this point the Kelsey data show that about 19% of survey respondents have a smartphone. It's closer to 13% (or so) for the general mobile user population. This discrepancy probably accounts for why some of the numbers in the survey are high.
In a previous online study with about 800 respondents we found that 29% had accessed the mobile Internet, and these LMS proprietary numbers were higher than Nielsen and comScore's numbers. (More important than the question of whether people had ever accessed the mobile Internet is the question of frequency and engagement.) I would argue in general that the trajectory of growth Kelsey shows from 2007 to 2008 is correct, in terms of mobile Internet access, but the figure (38.9%) is too high in terms of the general mobile user population in the US.
We all have seen that smarphones dramatically affect mobile Internet access and mobile search activity. From the recent TMP data:
In addition the 9.6% social networking number in the Kelsey findings is also directionally correct but too high regarding the mobile user population more generally. We found, as did comScore, that the number is closer to 6%.
Clearly mobile Internet access is gaining rapidly and future gains are tied largely to smartphone adoption. Perhaps the most interesting data point released in the Kelsey findings is that just over 49% (49.2%) "plan to purchase an advanced mobile device" (presumably a smartphone) "in the next two years."
Despite these intentions, price will determine how many of these aspirants actually purchase smartphones (see this post). But, as we've argued many times in the past, are where the market is heading.
The Kelsey Group based much of its mobile revenues forecast on the expectation of growth for ad-supported free DA. Logically this made lots of sense and I also was quite bullish on this segment at one time. But it simply hasn't materialized as a major driver of call volumes (or revenues accordingly) at this point. And as more people adopt smartphones -- and go direct to the Intermet from mobile -- the ad revenues contribution we can expect from the "Free DA" segment is propotionately smaller.
Skyhook Wireless CEO Ted Morgan sent me the image below of a CNN monitor that appeared last night on a CNN newscast. It shows CNN reporters covering the US presidential election in the various states. Morgan said they're using iPhones to track their people on the ground. Skyhook provides some of the location awareness technology in the device.
Earlier this week the BlackBerry Partners Fund announced its initial investments:
Previously the fund announced the winners of the BlackBerry Developer Challenge:
A healthy software applications ecosystem is critical now to long-term success in mobile. That's why the BlackBerry Partners Fund was formed and why Google funded the Android Developer Challenge. It's also why Microsoft is strongly considering creating its own Windows Mobile Apps store.
Here are the latest data from Skyhook Wireless, which is tracking location-based apps on the iPhone (and Android).
Week of October 3:
As of October 10:
It appears that in the week that separates the two charts there are about 75 more location-aware apps for the iPhone.
What's also interesting to note about Android apps, is that ShopSavvy is the most downloaded app in the first 24 hours. This is partly because the app was an Android Developer Challenge winner and has received a lot of publicity but also because it has inherent appeal.
Local and mobile search provider Poynt (Multiplied Media) won the prize for its "personal productivity and lifestyle" category and was one of three overall winners in the BlackBerry Developer Challenge.
Each grand prize winner gets $150K and becomes part of the BlackBerry Fund Partners JumpStart program. The other two "grand prize" winners were music related applications: Strands Social Player and Nobex Radio Companion.
MapQuest, which has been quite busy of late rolling out desktop site improvements and a new Local product, has just launched a new mobile site optimized for the iPhone. It features maps, directions, business search and gas price information.
Overall it offers a very strong user experience, nicely rendered. The only drawback is that because it's Safari-based and not an application it can't take advantage of the phone's built-in location awareness. Consequently most locations need to be manually entered.
Voice and local mobile search provider CallGenie announced that it's expanding its services beyond its core directory and DA markets. From the press release out this morning:
Call Genie is now bringing its groundbreaking mobile local search and advertising products to newspaper, television and triple-play publishers and is attracting both local and national brand advertisers interested in capitalizing on this newly created inventory.
In a very challenging economy with a limited number of buyers in the core YP and DA segments, local and mobile search vendors need to diversify.
I was in the Apple iTunes Apps Store and discovered that the top featured apps on both the "What's Hot" and "Staff Favorites" lists are local apps -- Urbanspoon and GoodRec. (I wrote previously about GoodRec here.)
In addition TheFind's shopping application (with local inventory data) "Where to Shop" is now available:
The application allows users to search by keyword, item or brand and find the desired item(s) online or in nearby stores (it uses the iPhone's location awareness). The latter offers a conventional list view and map view. Some of TheFind's local inventory data are from NearbyNow. Slifter also has a local-mobile shopping app. There are also two among the initial wave of Android apps.
A piece in AdWeek captures the topline figures on an eMarketer "location based services" forecast for mobile:
eMarketer forecasts that 486 million consumers worldwide will be using location-based services by 2012, up from 18.9 million in 2007.
Of course the size of the potential market all depends on how you define "location based services." There's the geotargeted advertising angle and then there's the conumers demand/use angle. eMarketer is talking about the latter.
Basically everyone who has a mobile phone and conducts a search is going to be searching for local/offline information from time to time. That doesn't take into account applications and favorites that are location-specific or location aware.
We found most recently in our own survey work that 29% of mobile phone users accessed the Internet and 16% searched. That 16% figure is identical to the recent TMP-comScore survey. Of course smartphone users access the mobile Internet and search much more than the average (over 50%).
Local is one of the top content categories across all mobile consumers surveys (Opus, comScore, Nielsen). In practice it's a subset of all query volume on mobile handsets (see "What Do 20 Million Queries Tell Us about Mobile Search?")
The number of LBS/local content users on mobile devices is thus a function of overall mobile Intenet access and, in particular, smartphone growth. Today (per CTIA) we have approximately 260 million mobile users in the US. The range of mobile Internet access is between 15% and 30%, let's say. That represents (in the US alone) between about 40 and almost 80 million people -- today.
It should thus be very easy to hit the eMarketer numbers in four years, on a global basis.
AdAge reports on mobile search and some of the recent TMP Directional Marketing-comScore survey findings. The survey of roughly 3,000 online consumers went into a broad range of areas. Mobile was a piece of the larger set of data.
We'll be rounding up this and other third party data in an advisory that will publish this week. For now here's a slide reflecting data from the TMP survey pertaining to mobile:
According to the data above, the majority of smartphone users (including PDAs) conduct mobile searches and local searches on their mobile phones. Putting aside DA and ad-supported DA, it thus becomes a fairly straightforward formula to predict the growth of search on mobile handsets -- it's tied to smartphone adoption. And on that point, NDP said that three of the top five selling phones in the US are now smartphones.