Earlier on this year a smart VC, who multiple times has competed successfully against Google, told me that Google is generally willing to act as an “irrational economic player”. It’s willing to destroy value just so others can’t get at it, even if it means destroying value for itself.
Google just announced that it will offer navigation. Navigation has so far captured 70%+ of the $2 billion mapping market. The bad news for established navigation players like Tele Atlas and Navteq is that this will erode the value of navigation, just like the value of maps have been eroded. The good news for those players is that Google has now tipped its hand and shown that it’s willing to compete against the very ecosystem that it has been nurturing over the past couple of years.
Google’s strategy is to leverage maps, including navigation, to extend its current search franchise into local search. We’ve believed from day one of CloudMade that contextual search, using location data and some knowledge about the user (for example which app he/she is using) creates much more valuable CPM/CPC/CPAs. If you’re an advertiser or a merchant, you will pay more per impression if you know someone is a mountain biker and is near the bicycle mega store you own.
Google betting on one-size-fits all model
Google is betting on building a horizontal, local search franchise. I know that sounds like an oxymoron, but please read on, it gets clearer. Google is building a one-size-fits-all set of services around mapping, and will serve the masses with those. Think about Google Maps, Latitude or Earth. They are horizontal one-size-fits-all web apps with little or no segmentation. Everyone uses the same app.
Meanwhile, in the process of building out end-user applications rather than sticking to being a platform player, Google is causing considerable collateral damage. Its move into the territory normally occupied by mobile operators, OEMs and small, medium and large developers is turning the marketplace against itself. The honeymoon is over and the do-no-evil days have ended. Google has declared any monetizable pocket in tech a target, including the key franchises of Apple, Microsoft, the mobile operators and now also mobile application developers. The problem with Google’s approach is, the value is not in horizontal services, but in leveraging the democratizing effect of the app stores to use the 100,000+ vertical apps as a way to divide the market into tiny segments and let them flourish and gain traction.
CloudMade believes vertical apps will inherently nano-segment the market
Most of the impressions that will hit consumers, and most searches that are contextually and location oriented, will occur through vertical apps. Take the mountain biker example. Where will I be more likely to respond if I want to a) navigate my mountain bike though a new trail and b) click on that ad from the bicycle mega store? A horizontal Google app, or a well crafted, vertical app written by a focused developer who understands my special interests? We believe the latter. We believe the mountain biking savvy, app developing expert who knows the hidden trails in his/her community will be better at providing relevance to local mountain bikers.
As I speak to mobile operators and handset manufacturers about the CloudMade business, it’s clear that most of the players understand the value of their immense reach, and want to pick long-term partners with whom they can build a franchise in local search, local advertising and local geo services.
It’s clear that Google’s latest move has served as a lighting rod for clarity in the value chain. Over the past days I have spoken to people throughout the ecosystem. So far, the uncertainty about what Google was doing has actually caused a lot of mobile operators, handset manufacturers and app developers to take a wait-and-see attitude. Now, with Google showing their hand, and making it clear that it’s willing to compete directly with substantial parts of the eco system to get at the local search market, we’re seeing that the wait-and-see is over. With Google choosing to go it alone rather than cooperate with the ecosystem, the ground rules have been laid down, and the competitive landscape is clear.
So what are we doing at CloudMade?
We’re building out our traction in the vertical mobile application area, signing developers within key categories that we believe will drive massive volume. This is a real micro-segmented approach, aimed at driving traffic to thousands of narrow verticals. We’re already seeing this scale.
Local, relevant ads: We’re jumping the learning curve on Location Based Advertising and Sponsored POIs. We’ll shortly be offering developers, mobile operators and handset manufacturers a revenue share on Sponsored POIs and ads that they include in their apps. This is big news to developers, who so far have had to depend on the modest revenues from selling apps at $0.99, $1.99 etc.
Data Marketplace: We’re building a massive marketplace for geo data. In fact, in a few weeks we have our “opening day” at the CloudMade Data Marketplace, the Turkish Bazar from which developers can choose a variety of content that they can mash into their maps. This will result in more app diversity, deeper functionality and even completely new types of vertical apps.
Offering for operators and handset manufacturers: CloudMade has created a super interesting set of propositions for mobile operators and handset manufacturers. We’re doing rev share deals with them where we work side-by-side to build franchises in the local geo spatial arena. We offer them onboard maps (built into their devices) and offboard maps (loaded from our servers), we offer them navigation jointly with our navigation partners and we offer up relationships with our many vertical app developers. For Tier 1, 2 and 3 operators and handset manufacturers this is turning out to be very compelling. The big differentiator is that the operators get to decide what the services look like, they get to brand them and they get to make money from them. That is as opposed to the alternative, which is to take someone else’s services, accept that they are all branded by someone else, who also pockets the ad revenues from them.
We’re focusing on making maps look the way the merchants and developers want them to look. Through CloudMade’s Style Editor anyone, even non-technical people, can produce advanced, custom maps that reflect the brand and identity of their company or their customers company. We’ve found that many, especially those with no yellow in their corporate identity, prefer this to a one-size-fits-all yellow map.
Navigation: We’re continuing to ramp up our work with key players in the turn-by-turn navigation field to increase the scope for navigation solutions. We are building assets to help drive the verticalization of the field, so we see custom navigation for different verticals. Again, CloudMade doesn’t believe in one-size-fits-all. We think different types of navigation will require different types of apps. For example, if I’m hiking in the mountains I don’t care about roads. I want to navigate hiking trails, be directed to places where I can fill my water bottle, and I want to know where the vista points are.
Last, but not least, the most important component of CloudMade’ approach.
Tools, tools, tools: CloudMade is focusing a lot of effort on the tools side. We believe that the best way we can serve the community of mappers and developers is to put all the control in their hands. We’ve demonstrated this through our efforts already, for example with our iPhone libraries. We’re similarly putting the control of Location Based Ads and Sponsored POIs into the hands of developers. Put yourself in the shoes of the vertical app developer. You know better than anyone else how your constituents want to be advertised to, what ads they want (and don’t want) to see and how to place those apps in your app.
Similarly, we’re readying a suite of tools that makes mapping easier, faster and more powerful.
CloudMade’s web based Mapzen tool, and Mapzen POI Collector for iPhone will be available shortly. This suite of tools enables us to further serve the 180,000 person large OpenStreetMap community that is building the most detailed, finely textured and accurate map of the world. The community was founded by the founders of CloudMade, and most members of the CloudMade team are active members of the OpenStreetMap community. The map we’re building in the community is stunning in it’s detail. It’s essentially the Wikipedia of maps. After all, who knows better how to map a community than people who live in the community.
The Mapzen suite enables us to bridge the needs of app developers, who serve as a proxy for consumers in a given vertical segment, and the mappers, who are members of local communities or specialists in a certain type of mapping (e.g. mountain biking trails). Bridging the consumers needs for maps and the mappers’ desire to create maps that truly reflect their local community will result in mobile and web applications that both feature better maps and be more attractive to users.
So what’s the bottom line?
Over the next twelve months, we will see hundreds of thousands of vertical apps use maps and location services to better serve consumers with data about where they are, where they are going, how they get there and what is surrounding them at their location. We will see mappers collect map data that is highly relevant to local, narrow communities. We will see owners of diverse datasets make their data available though the Data Marketplace. In turn, we will see app developers jump on the opportunity and leverage the map data and Data Marketplace datasets into highly targeted, compelling and enchanting apps. Many will choose to monetize the apps through a combination of app store revenues as well as carefully selected Location Based Advertising and Sponsored POIs.
If you are with a larger company with a need for a mapping provider that does not compete with you email me at firstname.lastname@example.org.
Talk about coincidence. Just as I was about to share a couple of numbers from our geocoding traffic report, someone sent me a link to an interesting read by Nick Lansley of Tesco. In a nutshell, he, as well as dozens (hundreds?) of other iPhone developers have suddenly found themselves cut off from Google geocoding.
Back to our latest traffic report. I thought it was interesting enough to share some of it with everyone. While it is expected that all of our products show usage growth over time, the usage of geocoding has simply skyrocketed lately. According to our traffic report, the number of geocoding requests has increased more than ten-fold over the last four weeks, without any sign that the trend is changing. I am also quite impressed by adoption of version 2 of our geocoding API. Given that most requests come from mobile applications that have necessarily longer update cycles, the new API’s 30% share of traffic in only three weeks is really a large number. The most popular request type?
In the meantime, we’ve been working on better support for EU-style addressing, which will be available real soon. Also in the works, fast and accurate US addressing which will be brought online by the end of next month. More to come…
While no one was watching, we deployed the latest version of our geocoding engine. The key objective for this release has been performance and stability, and I am sure it is immediately noticeable. Benchmarking based on actual user queries have shown that it is up to 24 times faster than the previous implementation for certain classes of requests. Even for simple, one-word queries, it is up to three times faster.
Along with the new engine, we are introducing new Geocoding API V2. In an effort to simplify usage, we have reduced the number of methods from eleven to only one, without sacrificing a single bit of its flexibility. The old API? It, too, uses the new engine to deliver faster results, but we do encourage everyone to switch to the new API as soon as you have a chance to.
In the meantime, we are already working on the next release. The main focus will be relevancy and better parsing of freeform queries. Stay tuned!