A great business idea…

Earlier this year, GoButler was founded by three ex-Rocket Internet executives. Backed by the venture capital firm General Catalyst Partners as well Hollywood star Ashton Kutcher, the firm aims to revolutionize the way people “get whatever they are looking for”. On Twitter, GoButler claims to be “your 24/7 concierge, your personal assistant… for FREE!” They promise to get you whatever you want as long as it is legal – you only need to send them a request via SMS. Likewise, Facebook recently beta launched “M.”, a virtual assistant that operates via Facebook’s Messenger. Seems like a cool thing! No need to have millions of different apps anymore, such as one to reserve dinner tables, one to book flights, and one to order cloths. GoButler is your one-stop shop for everything and aims to become THE new way of search.

…that, however, does not seem to work quite perfectly yet

I recently tested the service four times – so far, I’m not very impressed. I asked them to organize a daytrip to the Niagara Falls for me and friend while we were in Toronto a couple of weeks ago. Failed: Despite many messages back and forward explicitly stating the pick-up and drop-off location and times, GoButler was not able to organize a trip that met our stated expectations. We ended up organizing the trip ourselves. I send three more travel and shopping related search requests. Either they were not able to find what I was really looking for or the price was much higher than the one I found myself. These four test runs illustrated that what ought to be simplifying my life in fact made it more complicated. Even if GoButler manages to improve their search results over time, do they really have a chance of being successful in the long run? I see two major challenges: (1) Scalability and (2) profitability.

Is this business scalable?

Currently, all SMS requests are handled by dozens of humans sitting in an office in New York City. Scalability of the business seems rather challenging. I’d argue, either a significant amount of automatization and artificial intelligence is required to handle the search requests, or a tremendous amount of people need to be hired. The former, technology-wise doesn’t seem perfectly feasible nowadays and the latter seems extremely expensive.

Can this business be profitable?

GoButler promises not charging customers any on-top fees for their concierge service. Instead, the firm aims to monetize indirectly through an affiliate model: Once they have established a large enough user base, they aim to approach partners such as delivery restaurants, doctors, airlines, and online shops to negotiate revenue share models for every generated order. I see multiple challenges related to this approach: 

First, it seems questionable whether a revenue share approach will result in the best experience for the customer. The pizza delivery company that pays the highest revenue share to GoButler may not necessarily deliver the most delicious Pizza, too. Yet, GoButler may be inclined to place orders through this very delivery service due to financial incentives.

Second, it is unclear whether GoButler’s customer leads indeed create any additional revenue for their affiliates. Would customers have placed the orders anyway? If there is no proof of incremental revenues, partners are likely to disregard the concierge service.

Third, related to scalability, it seems rather challenging to easily build up partnerships with so many goods and services providers in order to really make money off every type of order placed through GoButler.

Fourth, even if the three aforementioned concerns turn out not to be valid, I’d question whether the unit economics of this business can really work out profitably. Let’s run the numbers: In my test inquiries, the average processing time per request was approximately 10 minutes. Hence, I’d assume one employee can handle six customer requests per hour. Let’s conservatively assume further that he/she manages to convert each of the leads into an actual deal worth on average USD 15 (average price of a food order). Lastly, let’s assume GoButler’s affiliates are willing to share 10% of their revenues. This would result in a total revenue for GoButler of USD 9 per hour. Pretty slim – is this even enough to cover the employee’s salary? Not to mention, the chance that every search requests indeed gets converted seems rather unlikely – just remember that none of my four test runs were even close to be converted.

Time to shift gears

One thing seems to be clear: GoButler will need to burn a lot of the investors’ money to test the viability of its business model. With GoButler’s current monetization approach, it will require significant growth of the customer base before one can even assess whether or not GoButler will indeed be able to establish the necessary partnerships to generate revenues and eventually profits. Following Clayton Christensen’s advice – a management guru from Harvard Business School – who says one should be “patient for growth but impatient for profit”, I would recommend to shift gears a little bit: Instead of offering the service to any and every one free of charge, GoButler may target high value customers who see a clear benefit in this concierge service and are thus willing to pay a premium for it. Alexander Goerlach – visiting Scholar at Harvard who recently visited the NYC office of GoButler in his role as the U.S. representative of the German Federal Association of German Startups – says: “There are so many extremely busy people out there who need to get things done – GoButler seems like the perfect service for them. If they get on board, I can well imagine that Google will consider consolidation options with GoButler.”

I personally believe that the extremely busy people amongst us – who not have a real personal assistant – may rely on a concierge service such as GoButler for most of their searches one day. But nonetheless, I find it hard to imagine that one day search engines will be disregarded completely. Let’s stay tuned and see how GoButler’s and Facebook’s virtual assistants roll out over time.

By: Noshad Irshad


read more

Yandex.ru is the Russian company providing search services to users. It’s currently the 4th most popular search engine in the world. In spring 2011 Yandex raised $1.3 billion in an initial public offering on NASDAQ. It was the biggest U.S. IPO for a dotcom since Google Inc. went public in 2004. And still is ranked #9 in the list of largest IPOs launched by Internet companies in the world.

Yandex.ru was founded in 1997 in Russia the same year as Google.com domain was registered. Yandex’s search engine still has 62 percent of the Russian market, compared with 27 percent for Google, according to researcher LiveInternet.ru. So over the last 15 years Google is still an underdog on the Russian market. Why is that and will it ever change?

I think there are several reasons for Yandex to significantly dominate the Russian search market.

First of all, it’s the first mover advantage.

In the late 80s the first search technologies were developed by Yandex co-founders  Arkady Volozh and Ilya Segalovich, well before Google co-founders even met. At that time, they developed a handful of search programs, including The International Classifier of Patents and Search through the Bible, which took into account the Russian-language morphology. That technology was named Yandex in 1993. And in September 1997 the search engine yandex.ru was launched.

Google first made Russian-language search available in 2001 and opened its first sales office there only in 2005, giving Yandex and Rambler (the second largest search service in Russia) a long head start.

And unlike its struggle to catch up in China, another international market, with its local search engine Baidu (#3 search engine in the world), Russian market battle was not influenced by political reasons. In China Google was allowed in only after agreeing to self-censor its search results and still is subject to many restrictions. Russia, in contrast, allows Google and local search companies to operate without political interference. According to NewYork Times article, for instance, a search on www.google.ru for polonium — the radioactive substance used to kill the former Russian spy Alexander V. Litvinenko in Britain recently — returned 11.1 million results.

The second reason for Yandex’s dominance is a better understanding of language and local market context.

Despite the fact that Sergey Brin, one of the founders of Google, was born in Moscow in 1973, and the first words out of his mouth were Russian, Google’s algorithm is far from perfect to perform search tasks in Russian language.

Yandex’s search algorithm is rooted in the highly inflected and very peculiar Russian language, in which nouns may be one of three genders and be declined in up to six cases. So words (nouns, adjectives, verbs) can take on some 20 different endings to indicate their relationship to one another, and that makes search extremely difficult. Largely because its technology failed to take into account the complexities of Russian grammar Google’s market share for much of the decade hovered at around 6%. Then, in 2006, Google hired a few dozen Russian engineers to address the language issues, and its market share jumped instantly. However even nowadays Google fetches the exact word combination you enter into the search bar, leaving out the slightly different forms that mean similar things while Yandex has found a way to catch them all.

A lot of articles name language issues the main reason why Google falls so far behind Yandex, however I think that it’s a largely overstated problem. In the end of the day, Google has arguably the best engineers in the world and plenty of experience of adopting the search algorithm to other languages. For example, the hardest language in the world, according to one of the rankings, is Polish and the complexities of that language are similar to Russian (it has 7 cases and 7 genders). However if we look at the market shares of various search engines, Google has more than 92% share. So it’s not only the language hardships that restrict Google’s growth on the Russian market.

Ms. Gofman at Rambler Media said local companies’ understanding of the Russian market extended beyond the language. It’s the overall business context understanding and adjustment that matters. For example, Yandex developed ways to receive payment for Internet ads using traditional accounts at brick-and-mortar banks, for example, which helps compensate for the fact that relatively few small businesses have credit cards or online payment capabilities. They even develop their own payment system called Yandex.Money, which works similar to Paypal.

This last example also shows that there is a third reason why Yandex is dominating the Russian market and is unlikely to lose its leadership position any time soon. It’s the company’s innovation and integration of various services offered for users.

While Yandex was the first search engine on the market, it kept innovating and adding new services, sometimes ahead of Google.

Well before contextual ads were started with AdSense (which Google started using in 2001) in 1998 Yandex already launched contextual advertisement on its search engine. In 2001 it launched theYandex.Direct advertisement network, which allows automatic placement of ads.

Yandex launched many products which are comparable or replicable of Google’s products: Yandex drive, market, pictures, browser etc.  Many of the services have unique features that don’t exist in Google. It also diversified into areas that can create synergetic effects for its current users. Payment system discussed above is one of the examples. Another example is the introduction of a player of free legal music in Yandex search results Yandex.Music launched  in 2009.


But most importantly, Yandex launched a couple of arguably unique products that have huge network effects for users. One of the best examples of such product is Yandex maps. This service provides real-time assessment of traffic of roads in Russia based on tracking mobile devices of people who use Yandex maps to navigate the roads. In June 2008 Yandex acquired SMI Link, a Russian road traffic monitoring agency, to merge it with Yandex.Maps services. This service is currently one of the most popular for Yandex and the more users use it, the more accurate traffic information becomes across the country. Currently, no other company can provide more accurate information about the situations on the roads, and there are also some very cool features, like opportunity for users to manually report issues. For example, if there is a traffic jam, users currently in it can write news updates on what exactly is happening, and  if there is a car accident, which lane it is on, so that the cars can move to the opposite lanes. Due to its superior mapping services, Yandex became the mapping partner for Apple in the region, substituting Google as a partner with development iOS 6 and the iPhone 5. On top of Yandex.Maps they also build a separate application for taxi services orders, where any taxi company can register for the service and compete for the pool of GPS-enabled taxi requests from users using Yandex.Maps.

Continued innovation of Yandex engineers and integration of various services within the Yandex umbrella ensures that Google will have hard time fighting established network effects and high multi-homing costs for Yandex users.









read more

Chinese version of Microsoft Adcenter vs. Google – Who wins this time?

It’s interesting to see the exact same battle between Microsoft Adcenter and Google happening in China, but with a different result, at least for now.

Baidu (Chinese goggle) has long long been established as the dominant search engine in China, especially after Google exited China in 2010. It currently has 62% market share by 2013 Oct, based on volume of unique visitors. Qihoo 360 (Chinese Microsoft Adcenter in the sense of competitive position in this battle, not business model) with its major business in security software has 21% of market share in search engine by 2013 Oct. However, if you look at the market share back in 2012 July, you will find the data look like 81.5% for Baidu vs. 0% for Qihoo 360. Yes, within a little more than one year of the launch of its search engine, Qihoo 360 has amazingly taken 20% of market share, mostly from Baidu! Totally different story from what Microsoft ended up with! So what’s the brilliant mobilization strategy adopted by Qihoo 360?

As I mentioned, Qihoo 360’s major business was in security software provider. It disrupted the industry by providing free personal security software to consumers while all the major players back then were adopting the ” free trial and pay” model. So it quickly acquired customers and occupied more than half of the market share. Then by leveraging its large consumer base of its security software, it first started bundling the software with its own internet browser to compete with Microsoft’s IE browser in 2008 and after its browser became the No.2 in the market, it changed the default search engine of its browser from Baidu to its own search engine in 2012. After that, we saw the amazing data above that it quickly took over 20% market share from the once-unbeatable Baidu.

Despite all the charges against Qihoo’s “dirty” tactics, it indeed succeeded, twice, in beating the monopolies in two sectors. Now the question is: Is Qihoo 360 going to win at the end? My view is relatively pessimistic.

Customers that Qihoo won over by far are mainly non sophistic customers. They either don’t use search engine as often as the rest of the population or they do very basic researches. According to a few searching result comparisons I conducted, Qihoo’s algorithm is far less accurate than Baidu’s regarding relatively complicate searching need. So sophiscated users will easily find out which one is better.

Competing with traffic is one thing but monetizing the traffic is a totally different story. Monetization requires not only traffic, but also strong sales team, excellent customer services, and well-established bidding system that really helps customer track result and improving effectiveness. On those fronts, Baidu enjoys disproportionately advantage to its relative user base over Qihoo. Although Qihoo is the second in user volume but it only has less than 5% of the market in terms of value. There is a still a long way to go for Qihoo to enhance its monetization capability. Without strong revenue stream, it has less in return to invest in improving its search engine further.

Interestingly, one of the big barriers that Google enjoyed, the large number of affiliate networks doesn’t play out for Baidu. Baidu’s affiliate networks only contributed a very small amount of its revenue and it seems not dedicated to expand the affiliate network.

read more

I would like to introduce a new product I came across while reading the news a few weeks ago. According to its website:

AdTrap is a small, zero configuration device that removes ads from your Internet connection before they reach any of your home devices. Video, music streams and mobile apps and websites, all ad free. AdTrap works with any browser and all of your connected WiFi devices. You can also connect AdTrap directly between your computer and the wall/switch (work, hotel, conference room). AdTrap also works with all-in-one modem/routers provided by your Internet provider.

This product is currently priced at $139, and to date, is the only available hardware product that blocks ads before ad content actually reaches the PC or any other WiFi connected device. It thus offers a hassle-free and standardized way to block ads out of the internet browsing experience. Users, though, still have the option to “whitelist” certain websites and to view their ads.

The founders raised around $200,000 through a Kickstarter crowd sourcing campaign.

While a tech-savvy friend told me that banner advertisers could find their way around the capabilities of this product by making ads that have picture file formats, it is quite evident that such ad blocking hardware would have an impact on a wide range of lucrative ad platforms, such as YouTube’s video ads or hyperlinked display ads on practically all websites.

Also, even though AdTrap might not become very ubiquitous as a standalone piece of hardware, if manufacturers of WiFi routers find it appropriate to acquire AdTrap or its technology, the outcome could be an integrated router that incorporates ad blocking technology at a minimal markup over current router prices. If ad blocking hardware embedded within a wireless internet router becomes the norm, search engines and other advertisers might find it very difficult to win the fight against routing equipment manufacturers.

While having an ad free internet browsing experience is favored by many, opponents of such an outcome argue that it could bring about the end of free internet content, when users would have to pay for all types of previously ad-supported services.

Faced with a potentially dangerous disruption, internet giants that rely on ads as a major revenue stream should seek to devote their vast engineering resources to finding ways around this ad blocking hardware’s capabilities. Alternatively, they should acquire this company and its (hopefully) patented technology at a generous price tag to remove it from the market before it makes headlines again.

The following YouTube video illustrates AdTrap’s functionality: http://www.youtube.com/watch?v=MonxXYLV44U


read more

Where will you search for your flights 5 years from now?

 If you are a mainstream internet user searching for your next Christmas flight tickets, odds are that you will probably go to a travel search engine or meta-search engine. A site such as Kayak.com saw 899 million user queries in 2011 [1] and demonstrates that they are not marginal players anymore. You will immediately have all prices for more than 500 airlines and many thousands of destinations. You will finally be redirected to the appropriate merchant and Kayak.com will be paid a lead generation fee by the merchant.

Vertical search is almost as old as e-commerce. Price comparison websites such as Expedia (founded in 1996) or Priceline (1997) in travel quickly emerged to make consumer life easier by syndicating prices from multiple merchant websites. These aggregators remained known only by early adopters as e-commerce was just burgeoning. As e-commerce grew rapidly, entry point for new users was usually a search on a generic search engine that directed users to these more specialized vertical search engines.

For a decade, this dynamic has fueled search engines’ tremendous success and profitability. Search engines’ business model is built on referring traffic looking to buy goods and services to commercial websites through ads placed on search result pages. Verticals with strong online dynamics and much dollar spending such as travel, electronics or financial products have been generic search engines’ bread and butter as they triggered strong online ad spending to attract prospect customers.

Today, generic search engines still account for a significant share of travel search engines’ traffic today (although as verticals mature, they get more organic traffic). Priceline for example spent almost a billion dollars or more than 20% of its revenues in online advertising in 2011, most of which being in search ad [2]. However, this nice and profitable situation started to be challenged in the second half of the 2000 decade for several reasons.

First, aggregators of aggregators or meta-search engines such as Kayak started to appear. Most early vertical search engines moved down the value chain to increase their revenue, profits and trying to propose a more differentiated service. Travel players started to propose services that went beyond just aggregating price information and are now handling the full customer relationship and executing the transaction (acting more as travel agencies or retailers). Whereas these traditional vertical players are being complementary to generic search engines, meta-search engines’ only purpose in life is collecting information from third-parties, organize it and giving back to customers the most relevant information related to their search. This looks really similar generic search engines’ mission (remember Google motto:  “Google’s mission is to organize the world’s information and make it universally accessible and useful”). They are substitutes to generic search engines: it is unlikely that a user would search a generic search engine first to be redirected to a travel meta-search engine to be finally redirected to an e-commerce site (either online travel agency or airline website). After the first use, a user would most likely skip the generic search engine step and go directly to the meta-search engine which offers him exhaustive travel information. Vertical meta-search engines are therefore a direct threat to generic search engines business model.

Second, increase in e-commerce volume allows vertical search engines and especially meta-search engines to reach a critical size and build strong brands offline. Indeed, while e-commerce was marginal, aggregators liked to use search ad due as a variable cost (revenue is directly linked to search ad spending modulo the conversion rate). As e-commerce becomes mainstream and users represent a larger share of the population, aggregators now redirect their traffic acquisition money from online ad to offline ad such as TV to reach a larger number of users at a lower cost per user.  The loss for generic search engine is triple:

  •  A shift of ad budgets from search ad to other forms of ad decreases keyword bidding competition and therefore average cost-per-click (CPC)
  • Increase in brand awareness for aggregators increases direct navigation to their website (thus bypassing generic search engines) or navigational search (when a user searches the name of a site or of a brand) which typically has much lower CPC than related keywords
  • Higher brand awareness means higher click-through-rates (CTR) on ads of these brands. To simplify, ad appearance in major generic search engines depends on a score which is a blend of CPC and CTR (a search engine will make as much money with an ad with a CTR of 10% and a CPC of $1 and an ad with a CTR of 1% and a CPC of $10). New entrants without that brand awareness have to pay much higher CPCs than established aggregators which make them less likely to be successful. Competition is therefore reduced, which drives down average CPC

 The third effect is related to the last point. Established aggregators have reached a level of sophistication in their ads and on traffic conversion that new entrants have hard time to match, again reducing competition and average PPC.

In this situation, how generic search engines will respond to that change will greatly impact the profitability and even the survival of this growing pool of aggregators and vertical meta-search engines.

We can’t imagine generic search engines not reacting. If they stay the course, they will see their search ad revenue plateauing as users will all go to Kayak.com to look for flights, Hotel.com for accommodation, to Amazon to buy goods, to MoneySupermarket.com in the UK to buy financial services and will go to Google or Bing only to look for football game scores, the date of a historical event or the nearest social security office which drive no or little revenue.

Generic search engines probably have no ambition to serve directly customers. This is not where they have experience or a competitive advantage. But it is most likely that they will want to include some sort of vertical meta-search in their search results. Google has engaged on this path already: it bought and integrated ITA software (that provides travel information to other sites like Kayak) in 2011, launched Google Travel [3] and strongly pushes Google Shopping. There is no obvious reason that could prevent them from succeeding: there are little to no network effects, barriers to entry are low for an Internet behemoth like Google who has experience in developing fast search algorithm and demonstrated capabilities to develop user-friendly products, Google is well positioned to take advantage of mobile usage shift with its strong Android platform and the trend towards more customization is something that generic search engines have been working on for years.

The game is just starting and there are also advantages in not being Google. Amazon might be out of reach now for goods, but I think the companies like Kayak, which just went public last summer, or the numerous start-ups in this space such as Hipmunk or SuperFly might have a hard time in the coming years.

[1] http://techcrunch.com/2012/07/20/kayak-and-palo-alto-ipos-looking-better-than-facebooks-whew/

[2] http://files.shareholder.com/downloads/PCLN/2122866626x0x561766/CB103EAD-2799-49FD-991C-F192C5F8223A/PCLN_2011_Annual_Report.pdf

[3] http://www.google.com/flights/


read more