Newer posts are loading.
You are at the newest post.
Click here to check if anything new just came in.

November 07 2012

Four short links: 8 November 2012

  1. Closely — new startup by Perry Evans (founder of MapQuest), giving businesses a simple app to track competitors’ online deals and social media activity. Seems a genius move to me: so many businesses flounder online, “I don’t know what to do!”, so giving them a birds-eye view of their competition turns the problem into “do better than them!”.
  2. The FT in Play (Reuters) — very interesting point in this analysis of the Financial Times being up for sale: [Traditional] journalism doesn’t have economies of scale. The bigger that journalistic organizations become, the less efficient they get. (via Bernard Hickey)
  3. Big Data Behind Obama’s Win (Time) — huge analytics operation, very secretive, providing insights and updates on everything.
  4. How to Predict the FutureThis is the story of a spreadsheet I’ve been keeping for almost twenty years. Thesis: hardware trends more useful for predicting advances than software trends. (via Kenton Kivestu)

May 03 2011

Snap to the graph, not the grid

About 10 years ago, "geo" and "local" were two very distinct product verticals in most Internet companies. This was in part because what we now think of as local arose out of a Yellow Pages market, while the mapping use case arose from more enterprise, maps-and-navigation origins.

At the time, most Internet businesses were keen to perpetuate extant geo-local user experiences, rather than seeing them as crude, if profitable, interim steps to something better. Nowhere was this more evident than at my former employer, Yahoo: as late as 2006, investment cases were still being pitched for either maps or local; the bulk of Yahoo's geo engineering team resided in an entirely different country; and the mobile group had its own building with few, if any, connections to the rest of the company.

Today, entrepreneurs and execs know that local, geo, and mobile are not distinct product groups and technology stacks, but rather essential components of a unified toolset that better connects people with the world around them.

Nothing better reflects the diversity of this geo toolkit than the products on display at the recent Where 2.0 conference. The three-day event embraced the full spectrum of all things spatial: coding, licensing, marketing, daily deals, 3D visualization, and a bit of mapping thrown in for the traditionalists in the audience.

In this context I gave a short presentation on "Big Data, Big Local." I've embedded the slides below. Although I anticipate receiving an award for including Borat, Homer Simpson, Bender, Winston Churchill, porn shops, kittens, surfing, and unicorns coherently in a single presentation on geo — the deck does not stand well on its own. Go figure.

I wanted therefore to provide a brief overview here, and perhaps tackle it more fully in subsequent posts. This is not intended to be a detailed argument, simply an accompanying narrative to the deck.

The general tenor of the discussion focuses upon how we have traditionally employed coordinates as the final word in how we position things and people in the physical world. The problem with this approach is that a coordinate pair lacks context — they are regular and orderly, which is great for machines, but are entirely ambiguous when used to represent more conceptual places like states, cities, stores and neighborhoods. That's not so good for people. In fact, a single coordinate pair can be used to represent all these places, and their different associations, concurrently.

Social location has moved us toward a new form of positioning: by business or points of interest (POI). These geo-referenced entities exist across the world in an irregular and poorly typed graph, but they are rich with context. People exist and events happen at places, not coordinates — social location applications "snap" us to these nodes, and their context becomes ours.

We are stretching the traditional use of business listings beyond their limits. Looking back, however, you can see how they've been continuously extended to fit new models of bringing consumers and businesses closer together. You'll soon see listings data act as hooks in the mobile payments space, and they will become increasingly pivotal in creating rich topological graphs among people, brands, and physical places.

We can do great things with business listings and POI as geo-commercial beacons, but they come with their own problems, two of which I cover in the deck. The first is a more pedestrian example of name canonicalization: stores in the same retail / restaurant chain are represented today in too many forms. The evidence presented by variants on "Subway Sandwiches" in slide 12 (below) would suggest that current data providers expend comparatively little effort normalizing this content. The problem may exist because this information is currently intended to serve an extant, limited use case — make the name discoverable using free text search and present to users — rather than drive new, data-driven markets.

Poor / Absent Canonicalization

The second, more important issue is that we are faced on the Internet with a multitude of electronic incarnations of one physical entity. While intended to enhance discoverability across properties, the current situation effects a near-complete inability to understand how users engage electronically with a physical place outside a single caisson such as Facebook, Twitter, or Yelp. This need is increasingly important, because URLs are now often employed as surrogates for a place — they can be resolved to machine-readable attributes and in fact act exactly like nodes in a graph of linked data.

This is a problem that requires solving, and we are tackling it at Factual. I include some topical commentaries from Chris Dixon and Albert Wenger as artillery support in slide 18:

The need for common namespaces

So the problem is that we are collectively going to need this normalization, very soon. I conclude noting that Factual's duty does not end with the creation of tab-delineated files. Instead, we are looking to create data-centric platforms that instill order on an increasingly chaotic Internet, and ensure that the result is meeting the future business needs of our contemporaries in the local space.


Sponsored post
you are awesome!
Reposted bysirthomasbolton sirthomasbolton

March 22 2011

Four short links: 22 March 2011

  1. EveryBlock Redesigned -- EB has been defined for a while now as "that site that makes my city's statistics useful and relevant". Now they're getting more into the user-reporting: As valuable as automated updates of crime, media mentions, and other EveryBlock news are, contributions from your fellow neighbors are significantly more meaningful and useful. While we're not removing our existing aggregation of public records and other neighborhood information (more on this in a bit), we've come to realize that human participation is essential, not only as a layer on top but as the bedrock of the site. They have a new mission: our goal is to help you make your block a better place. That's a bold goal, and quite a big change from where they were at. Will they manage any aspect of journalism, or will this become a GroupOn-ad-filled geo-portal for MSNBC? Looking forward to finding out.
  2. Typography in 8 Bits: System Fonts -- nifty rundown of fonts from the microcomputer days. I still go a bit weak-kneed at the sight of the C64 fonts. Which aspect of the system you're building will be remembered with weak knees in (gulp) thirty years' time? (via Joshua Schachter)
  3. Twitter in the Christchurch Earthquake -- analysis of the tweets around the quake, including words and retweets. (via Richard Wood)
  4. ChumbyCV -- computer vision framework for Chumby. CV on low-power ubiquitous hardware makes devices smarter and be higher-level sensors of activity and objects. (via BERG London)

January 29 2011

TERRA 602: Cotton-Top

The Cotton-Top Tamarin is endemic to Colombia’s Caribbean Coast and is critically endangered. ; Thanks to “Proyecto Titi’s” conservation efforts, the Cotton-Top and the people from the region have an opportunity to succeed.
TERRA 602: Cotton-Top

The Cotton-Top Tamarin is endemic to Colombia’s Caribbean Coast and is critically endangered. ; Thanks to “Proyecto Titi’s” conservation efforts, the Cotton-Top and the people from the region have an opportunity to succeed.

September 02 2010

Toward a local syzygy: aligning deals, check-ins and places

Three significant trends in the local sector -- deals, check-ins, and place pages -- are on a bender and headed for an exciting convergence. When they meet we will see one of three things: a train wreck of incompatibility, an awkward confluence, or a very powerful alignment. I'm hoping for the latter, a sort of local syzygy, because a well-conceived orchestration of these trends will benefit the consumer and it has real potential to take us entirely out of the Yellow Pages era and into exciting, unexplored territory.

This is a two-part post: here I look in more detail at check-ins, deals, and place products (including, briefly, the adventurously named Facebook Places) with an eye to what might follow. In a following post I will discuss how we may more actively ease their convergence with linked data and some basic adherence to extant standards, specifically how these efforts will affect the local consumer.

Place pages and check-ins

The check-in is hardly the apogee of the local consumer experience but it works, and this is what is most important about any product. However successful it is now, the check-in will remain an interim solution for identifying long-term customer/business affinities and physical point of presence. So what's next?

I've written about check-ins previously: since then, Facebook has thrown its hat into the ring with their own place/check-in product, offering little feature distinction outside the problematic ability to check-in your friends on their behalf . Thanks to Facebook, the "Ferris Bueller Problem" -- in which a friend checks you into (say) the Von Steuben Day Parade when you are officially at home, ill -- may soon find its way into mainstream parlance. Expect a rise in just-for-larks in absentia check-ins to the local "gentlemen's club" and similar places of sophomoric amusement.

More interestingly, casual requests for a similar product from LinkedIn, and the introduction of a third-party check-in offering for Twitter demonstrate that geo and social products are becoming more integrated in the mind of the consumer, and corporate product strategies: Greg Sterling remarked on this trend recently in's new eat, play, live marketing campaign which attempts to transform the brand from its staid origins into "a lifestyle guide that also happens to feature contractors and plumbers." He is spot-on: some lines of business will not fit within the check-in model, but they nonetheless must be accommodated in any successful business-to-consumer product.

I have no desire to see another check-in clone arrive anytime soon. Jeff Holden, the founder of Whrrl, recently noted that the check-in will shortly be a commodity, and Foursquare's Dennis Crowley believes it already is. If you are an entrepreneur or developer thinking of building a new check-in service, please don't. Instead, consider some of the more exciting challenges that provide real consumer benefit:

  • Divorce the business from user authentication and social network: Social/local apps are entirely insular: you must be a Foursquare user using a Foursquare app (or the Foursquare API) to check into a Foursquare venue. I can see why this is, but it need not be. For example, consider an alternative platform where users supply instead an authenticated OpenID and an hcard. Authenticated user-to-venue relationships could be exposed across multiple social networks under the user's control, and the access control list could even accommodate attribute criteria such as place type or business category. This is not simple aggregation or shimming, but a lower-level, cross-platform integration.
  • Auto check-in: Manual check-ins are so 2009. Instead, think toward the endpoint where a geo-commercial footprint will be created automatically as an artifact of our day-to-day activities, should we wish it. The first steps in this direction will be toward improving the consumer's retail experience. Shopkick is effectively an auto check-in service. Other efforts toward this end will include credit-card geo traces, perhaps combined with a "local favorites" list used to prescribe which businesses may be checked into automatically. There are critical privacy implications here, so this footprint must be controlled by users, and employed by small and medium businesses (SMBs) to engage their customers on their own terms -- offering deals and events rather than advertising. As sharing your location with others becomes more codified, it will place demands on forthcoming products and start-ups that expose this under very granular, controlled conditions. This includes timestamps, geofences, disparate social networks, and leveraged venue classification.
  • Micro beacons: Expect over the next two years a bloom of micro-positioning products: sensor-and-platform combinations that assist with indoor navigation (malls), determining when a user is genuinely "in store," and guiding the consumer directly to a product on-shelf. Shopkick's iPhone product uses the device's microphone (simple, clever) to do this, but "Bluetooth beaconing" and QR codes are other low-energy, low-hassle options. We shall see both passive and active implementations to assist with the auto check-in and location disambiguation, depending on the particular use case.

Advances along these tracks should obviate the check-in as we know it today. This is a good thing -- check-ins are something to get over, an intermediate solution to tolerate until we have something that works better. Foursquare certainly knows this. The excitement -- for Foursquare's business and users -- lies wholly in their ability to deliver utility, novelty, and serendipity beyond the check-in.

Get it here: the deal

Groupon's $134 million series C funding in April and its estimated $1.34 billion valuation woke investors and entrepreneurs to the monetary value of group buying, and local deals more generally. There are now hundreds of variants and multiple aggregators, while "Groupon clone scripts" can be purchased (caveat emptor) from any number of freelance developer sites.

Groupon and its ilk tend to get bundled under the "group buying" or less-apt "social couponing" monikers, but -- in regards to Groupon certainly -- there's little that's actively social about the products. The less charitable may argue that Groupon's success is due as much to the severity of discounts on offer. However, Groupon and others have raised awareness that advertising is no longer the only solution. Specifically, SMBs are slowly gaining access to tools to engage their customers on mutually favorable terms. Examples:

  • Just-in-time delivery: Businesses increasingly have the ability to deliver discounts, incentives, and upsells to consumers in a highly relevant, just-in-time manner. Think of services that provide attractive discounts when tables are open at a restaurant, when spare inventory is available, or the business simply requires immediate revenue to help with cashflow. These new approaches secure short-term, incremental revenue for the business and offer great value to the user. Win-win.
  • Topical content: Most of these products now deliver deals based on city and zip code with little accommodation for either personal preferences or current location. Expect "relevance" to become the new watchword in deal and coupon delivery. Relevance will be obtained via platforms that exploit an accessible selection of local business favorites: Facebook is in a position to deliver, as is Hunch local, or my former shop, LikeList.
  • Coupon organizers: Digital deals are big, and will become only more common as delivery channels are improved. Jim Moran, co-founder of daily deal aggregator Yipit, records more than 100 companies offering deals in the U.S. alone, most recently Yelp. With these numbers, digital deals threaten to become as unmanageable as their hardcopy equivalents. In the same way that local favorites will assist in filtering the signal from the noise in the forthcoming local clamor, I would expect that intelligent, digital coupon wallets will be employed to remind or introduce users to deals that are relevant to their immediate location and circumstance. One coupon vendor called Savings Sidekick (apparently not related to the Shopkick iphone app) is doing something similar by using the tip feature in Foursquare to remind users of nearby deals from their coupon books. This particular implementation is as interesting as it could be annoying.

Part 1 wrap

Local is huge and only getting bigger. As a litmus, Borrell's recent ad forecast notes that "local online advertising should grow by almost 18% [...] to $16.1 billion, in 2011." Money follows money: we can expect further me-too products around deals, check-ins, and place products, but there is huge scope for investment into products that contribute genuine value to the consumer experience and enhance SMBs' ability to connect to their customers.

Much of this will take place at the data and platform levels. In my next post I'll take a look at how linked data might help cross-platform integration, and join deals, check-ins and place pages to the benefit of the consumer.


May 12 2010

Why check-ins and like buttons will change the local landscape

Notable advancements in the geo sector -- GIS, GPS, slippy maps -- punctuate an otherwise steady equilibrium; progress in the geo world is subtle, and tends to sneak up on us without our at first knowing its significance. I think we've just passed one of these stealth milestones whose importance will be realized only after the amplitude of the ripple effect shakes the local landscape. I'm talking about the humble check-in and the like button.

On first blush the idea of a "check in" is unworthy of note -- broadcasting publicly that you are at a particular venue is neither technologically sophisticated nor particularly novel. However, the driving impulse here is what's interesting: the check-in is much more about who you are than where you are. The passion driving the uptake of Foursquare, Gowalla, and their ilk is sociological rather than geographical.

People check-in to venues for the same reasons they so proudly display swooshes on shirts, Coach emblems on handbags, and glowing apples (or penguins for the more enlightened) on computing devices: these emblems communicate a facet of one's identity to the world in a facile, shared language. Checking into "Zeitgeist" in San Francisco tells the world you have roguish good looks (surely), plus a penchant for beer, outdoor smoking, and tamales. Checking into "Costco" tells the world that you own a car, and have an apartment or house in which to store large bulk items; checking into "SFO Long Term Parking" tells the world that you check-in too much, and should maybe ease-off a bit.

What logo or symbol can say so much, with such brevity, about an individual? The venue makes the man.

The magic here is the expression of affinity with a business, which is where the idea of the "like" has similar power in the local space (here I use "like" as any systemic convenience to express affinity, not exclusively Facebook's). In many ways liking a local business has significant advantages over a check-in: you do not have to be on-site, you like only once in perpetuity, and your affinity for the business is not measured by the number of times you visit. Check-ins, however, measure on-site presence, a hugely important metric for any business that values feet-through-the-door.

Thus the check-in holds the most potential for point-of-sale, in-the-moment marketing ("I see you are next door; come try a free latte at our place instead"), while Likes may offer greater potential for geo-relevant demographic targeting ("Here are deals for restaurants, similar to the ones you like, in areas where you are commercially active"). Likes and check-ins are really two sides of the same coin, but unfortunately share the same curse: there is no way to realize this user-to-business affinity and the value it represents uniformly across platforms.

Where previously this lack demanded only a frustrating duplication of effort, the problem has become more serious: the advent of likes and check-ins drive a semantic necessity to open and aggregate -- the problem is no longer one of mere inconvenience.

Gary Gale calls this problem "Geo Babel"; and the local landscape has all the hallmarks of the eponymous tower. Without a common means, our likes and check-ins will continue to be bound to the platform on which they were created, hamstringing their potential. Elbowing my way into Tim O'Reilly's metaphor, I would propose that this is a critical, missing component of the Internet Operating System. It seems outrageous that we find ourselves in 2010 without the means to refer to a business in a unique and unambiguous manner. Developers are left holding the bag for a number of reasons; here are my big three:

  1. Focus on listings data as end rather than means: Local search as we know it today is the parthenogenous child of the Yellow Pages industry. Many local search sites, and the data vendors they rely on, remain grounded in YP-era thinking, where the value was found in owning the listing data, making them discoverable in alphabetical order, and advertising against these listings. Local search for ages focused on being an electronic version of the Yellow Pages. Few organizations have looked above the horizon and considered carefully what value could be realized if listings were viewed as a means to connect users to businesses, rather than only advertise against their search.
  2. Attempts at distinction with common data: With peripheral exceptions, every local site out there has the same data. Boil these listings down to their hcards and you're looking at dupes across the board. This is of course known by all the big players, which is why they focus on differentiation with ancillary data such as ratings and reviews, or making their listings the most inclusive or the most current. This strategy drives competitors to attempt the "one ring" approach, where each tries to climb above the other in an effort to become the premier source of online local listings. It is a bugbear for developers because efforts to differentiate a commodity are always frustrating and rarely carry the sector forward.
  3. Over-fascination with pins on maps: The advent of GPS-enabled devices and Wi-Fi positioning gave us the ability to show people's location on the map; this has proved hugely distracting while the sector over the last four years has attempted to make buddy-finding interesting. We've since learned, I think, that people do not navigate by Long/Lat -- Foursquare and Gowalla succeed in part because they use business listings or venues as mechanisms of personal expression within a shared social landscape. Social location is best referenced by place not space; the coords tell me nothing.

This lacuna can be resolved one of two ways -- either one will do the trick but I am hoping for both:

First, we can work together toward an open database of places. Erick Schonfeld's post on TechCrunch a few weeks back is the most recent clarion. This is not an insignificant undertaking, but is certainly the preferred outcome. However, even with OpenStreetMap at our back, realizing this goal will take an eon in Internet time unless we see an unanticipated outbreak of altruism on the part of a data supplier. Of course, the 'new' players in the game -- those whose businesses use listings as a conveyance of value understand this need and are best positioned to deliver against it.

The second alternative -- one that is much more realistic in the short term -- is an open and accessible Local Listings Crosswalk, essentially an API that takes your URI and translates it to my URI so we can ensure we are speaking about the same business (I am less concerned about geographic place-names because such services already exist); Placecast's Match API went live as I was writing this post, so I hope that theirs will be the first step in getting us to where we want to be. The outcome,of course, will be achieved when we have a number of these services crosswalking via callbacks and hcard/og-aware crawlers, perpetuating listings data and ensuring they are up-to-date and uniformly accessible.

One or both of these services will allow us to put the bother of business listings management behind us, so we can get on with what's really exciting about local: connecting consumers with businesses they love, and providing genuine value to both.


February 06 2010

Feedback and analysis: the missing ingredients in local's recipe

There's plenty of enthusiasm for local / hyperlocal projects, but the sweepstakes has yet to be won. PaperG CEO Victor Wong digs in to some of the missed opportunities in a guest column.

I found this excerpt intriguing:

How useful would it be to know when local used-car dealerships have a large increase in inventory (and thus are probably more willing to sell at a lower price)? Other data like new-car listings could show what the local population is buying by examining what is posted and taken down by the dealers. Publishers can even create new content by encouraging users to input data about what sorts of deals and treatment they got, which would be useful for other local buyers and could be turned into a local car-buying guide.

O'Reilly Where 2010 ConferenceWong has a stake in the local game -- PaperG focuses on local advertising -- but that doesn't diminish the point he alludes to in the excerpt: feedback and analysis are the missing parameters in the local equation.

So many of these local efforts rely on traditional information delivery through news articles or databases. That material has use, no doubt. Yet few projects take the extra step and put that data into context. They don't explain why the information is important. They don't connect the dots.

A lot of this reminds me of web analytics. It's easy to grant access to traffic data, and the access itself has a low level of value. But the insight that guides decisions comes from deeper analysis. You need to know why a particular keyword or topic is resonating.

Older posts are this way If this message doesn't go away, click anywhere on the page to continue loading posts.
Could not load more posts
Maybe Soup is currently being updated? I'll try again automatically in a few seconds...
Just a second, loading more posts...
You've reached the end.
No Soup for you

Don't be the product, buy the product!

YES, I want to SOUP ●UP for ...