Archive for the ‘google’ tag
There was an interesting New Year’s Day post on TechCrunch about something we’ve all noticed: for certain keyword phrases, Google is entirely spam. A search for a high-value keyword like “online degrees”, for instance, turns up little more than affiliate directories run by spammers with a solid grasp of SEO. So many people are gaming Google that it has lost much of its value.
But — contrary to Wadhwa’s implication — this isn’t a special failure on the part of Google’s engineers. Rather, it’s a fundamental characteristic of dominant technologies. Through market dominance, a technology can become the sole target of those who wish to exploit: an easy ROI for scammers, marketers and anyone else out to make a buck. Rather than building and optimizing for multiple competitive technologies, system gamers must only target one.
This is of particular concern for monopoly technologies, or borderline monopolies. Microsoft ran into the same problems fifteen years ago and continues to suffer the fallout. Hackers and virus creators knew that they only had to optimize for one operating system — Windows — and could target a massive share of the market. They ignored Unix and Mac OSes, giving those systems a reputation of relative security and safety against viruses and hackers. But have no doubt that if, say, Mac OS gained sufficient market share and corporate adoption, malware creators would see a new opportunity and begin writing viruses and malware for Macs. Suddenly, finding exploits in OS X would become orders of magnitude more important than it is today.
And thus Wadhwa’s conculsion (“We need a new Google”), makes no sense. We don’t need a new Google, an overwhelming search monopoly. We need a diversity of competitive search engines. Blekko’s engineers are no better than Google’s. And even if they were better, creating a search engine that is immune to gaming is fundamentally impossible, with increasing difficulty as the search engine’s market share increases. Blekko is simply not spammed because it’s not worth the spammers’ time to figure it out.
Display advertising is a great counter-example of a market with diverse technologies, protocols and big players. While display isn’t totally immune to gaming — click arb and ads that launch malware, for instance — it doesn’t fundamentally challenge the value of the technology as overzealous SEO does to search.
When a technology is in a constant arms race with competitors, users win. When it is a black box inside a giant monopoly, the internet’s underbelly rolls up its sleeves and gets to work.
There is a certain watershed moment in the evolution of most consumer web entrepreneurs. In this moment, entrepreneurs recognize that they do not understand the way most people think about and interact with the internet. They recognize that if you are going to market to “Normals“, you have to leave your intuition of user behavior on the web behind.
People who are successful at marketing and serving content to Normals recognize that most internet users can’t distinguish between platform and app, content and ad, “good” content and “bad” content. Normals need a different level of clarity than savvier users, and if something isn’t immediately apparent, they’ll leave. Normals type URLs into Google, Microsoft Word or their email search bar. Each Normal is unique, and there are many different categories of Normals, but they all interact with the web in different ways than us (“early adopters”, “techies”, “bleeding edge”, etc), ways that may or may not be predictable to even the most seasoned designers and entrepreneurs.
Take journalism, for instance.
Old-line journalists regularly equate newspapers with journalism and blogs with something lesser and dirtier. In truth, there are a significant number of blogs that do far better journalism than newspapers have ever done. But there are many, many more that are simply content factories that benefit from a deep understanding of Google’s ranking algorithm that smaller (and journalistically superior) players don’t have. Demand Media, for instance, built a massive business by understanding that Normals have difficulty distinguishing quality, well-researched content from content that was generated by an Indian freelancer paid half a penny per word.
So what does any of this have to do with food? Well, the core problem with American food production in the early 20th century was informational: normal people didn’t have sufficient knowledge of their food’s origins and quality and were thus unable to distinguish between “good” food (e.g., meat processed in a clean environment) and “bad” food (e.g., meat processed in filthy sweatshop-slaughterhouses). In an economic environment in which buyers cannot distinguish between high quality and low quality products, all producers will move to solely creating low-quality products. High-quality producers will simply be priced out of the market.
Today’s content production industry is in the midst of that phase shift. As the vast majority of consumers cannot distinguish between good and bad content, mass-produced low-quality content is slowly pushing high-quality, journalistic online content deeper and deeper into niches.
Created in 1906, the FDA set out to solve this informational asymmetry in the food production industry by introducing basic hygiene standards coupled with labeling, inspections and reporting. This is analogous to what Google claims to be doing today for content — decreasing our informational asymmetries around content by calculating the “authority” of various content sources and offering content to users accordingly.
But I don’t think Google is doing a particularly good job. Google’s inability to effectively determine and communicate the value of content is, after all, why Demand Media and its kin can exist as businesses. It is why searches for high-value keywords (“online degrees“, for instance) return a bunch of affiliate honeypots and garbage content size wholly focused on acquiring users via SEO. And if you’ve ever ranked well for a particular search term, you know well that those are low-value users as opposed to users acquired via other channels. Web search has become a dramatically inferior way to discover anything online.
But unlike the FDA, Google is not a government agency. It’s a private company with private competitors. Competitors that may — no, will — eventually unseat Google as the king of search.
So who is going to write The Jungle of content?
Almost everything is digital. It started with simple stuff, like calculators, watches and measurements, and now the digital revolution totally owns cable television, video, music and photography. But there’s still one huge pile of valuable information that has yet to be captured and digitized: our conversations.
The conversations we have with other humans are one of — if not the — most valuable piles of information in the world. Far more valuable than video content (how much is taped versus just spoken?) or written content (how much information is actually written down versus simply spoken?) This information is not just valuable in the aggregate, but it is specifically valuable to the individual. If I were to have an easily searchable log of all my conversations, my productivity would increase by at least a third. Followups would be easier. Business opportunities wouldn’t drop by the wayside. I would pay good money for this.
The technology (lapel microphone, a mechanism to transmit streaming audio to a hard drive or the web, speech-to-text, search) is all out there in some form. Sure, some conversations — such as those in bars or on airplanes — would be lost. And I’d have to spend 5 to 10 minutes every evening tagging specific conversations with my contacts so I know who said what. (And perhaps with an evolution of technology, the app would learn to associate certain voices with certain contacts.) But the value proposition to the end user is huge. This is going to happen; it’s simply a matter of when and who does it.
There are a laundry list of potential uses. Some of them, privacy concerns notwithstanding:
– Early identification of “trending topics”, with particular relevance in finance
– Vastly improved real-time ad targeting. Possibly bigger than web search.
– A real-time gauge of public opinion and beliefs, with particular relevance to politics and brands
– A great data set to test theories of human interaction and sociology “in the wild”
That said, I’m not totally sure how this should be priced. On one hand, I am willing to pay upwards of $50 a month for a service like this — assuming I own the conversations and they won’t be used to serve me ads or packaged and re-sold to hedge funds. But I’m not sure this is the best revenue model; the business may be much more profitable by giving the voice-capturing service away for free and leveraging the aggregate data. What if Google had charged users a monthly fee in 1999?
Regardless of how it’s done, this is going to happen. It simply isn’t sustainable that the largest and most valuable medium of information in the world isn’t being captured.