Long tail by Chris Anderson – win a copy of the book
TweetChris Anderson made a superb presentation about his concept of Long Tail at our customer conference 2 weeks ago. Before i talk about his presentation, i want to give you all a few pointers to pick up this concept. I strongly believe that all management professionals must understand the Long Tail because it has deep implications for everyone. If you already know about the concept you can skip the following section and go to the next section.
The Long Tail Concept
As a subscriber of the Wired magazine, i chanced upon the article that Chris Anderson wrote in Oct 2004 which talked about the Long tail concept for the first time. It didn’t hit me the first time i read it. Almost 6 months later, i came across Joe Kraus’s blog post about Long tail and his explanations and the charts he published made me really understand the concept. Make sure to read the powerpoint slides he has attached to his blog post as well. [In case you didn’t know Joe Kraus is the founder of Jotspot, now part of Google. Also note that some of stats he quotes about Amazon, Netflix etc were later revised making the long tail sales of Amazon/Netflix not as stark as these charts show.] To be sure, Long tail is just a new way of looking at what Pareto pointed out long ago about winner-take-all-power- law distributions. While Pareto focused on the head of the curve, Chris Anderson, has brilliantly pointed the whole world at the tail of the curve – hence the term, Long Tail.
Long Tail Presentation titled “The Economics of Abundance”
Chris Anderson gave a great presentation. Compared to Dan Tapscott’s earlier presentation, this was less entertaining. But Chris covered his ideas extremely well. I found a version of his presentation on Slideshare and i have embedded it below.
| View | Upload your own
The main idea he was presenting was that the whole internet revolution has given us the abundance (Amazon can
list as many books as there are available) as opposed to the scarcity we have faced (Barnes & Noble has only a limited
shelf space for books). He made the point that we need a whole new mindset to manage in the world of abundance. If you
look at slide #22 (next few slides give more contrasts) in the above presentation – in the scarcity dominated world, we need
to think about ROI whereas in the Abundance dominated world, we can figure out the ROI later because it is very cheap to experiment. I could personally identify with it. When we started our internal blogging, I had no idea about the ROI and when
i looked at the cost of running a blog server (we use the open source WordPress), it was so tiny that i realized ROI doesn’t matter.
So i went ahead and now it is a raging phenomenon within the company pulling in 3 million page views a month. If you want a
public example, look at how Guy Kawasaki started Truemors with just $12,000. With that type of costs, it is just much easier to experiment and figure out than write a
big business plan and run after VCs for investments.
The Long Tail Book Ever since I came across the concept, i have been following Chris Anderson’s long tail blog and bought the book as soon as it came out more than a year ago. The book definitely lived up to my expectations.
The Long Tail book has a lot more examples and they all serve to reinforce the idea better. The book explains the 3 forces that are driving the long tail
1. Democratize Production – how digital video cameras, blogs, pod casts etc have put the power of production in the hands
of the masses. He calls them “Producers”. 2. Democratize Distribution – How Amazon, Netflix, iTunes are able to create a vast distribution channel for niche content. He calls them “Aggregators”. 3. Connect Supply & Demand – how Google, blogs, recommendation lists etc are helping connect supply and demand. He calls them “Filters”. He devotes a chapter each of the 3
forces and explains them in detail with lots of examples.
In sum, Anderson shows how the culture has shifted from the water cooler talk – “did you watch the <popular sitcom> episode last night” forcing us all to be in sync with each other to a culture driven by individual needs, desires and tastes.
Win a copy of the Long tail book
Now to the contest. Just after Chris Anderson’s presentation i was talking to my colleague and he was debating me on how
the Long Tail doesn’t apply to software development outsourcing companies like the one i work for? I disagreed with that assertion, of course. So here is the contest – the best response to the question – how does the Long Tail affect software development outsourcing companies? wins a copy of the Long Tail book – I have a spare copy to give away.
Notes & References:
1. I wrote an article titled “Egocasting” on ITOTD. For some reason, unbeknownst to me, i failed to mention the Long Tail concept even though i encountered it before writing this article. I have recently amended the article. It will give you a good idea on how the power is shifting in the media world.
2. Rashmi Sinha expresses her contrarian views on the Long tail concept. In my view, she has just explained how the internet accelerates and accentuates power law effects. In fact, i would argue that her Slideshare business is actually a long tail business.
3. Tim Wu of Slate slams the Long Tail with his article titled “The Wrong Tail”
4. The long tail post i did a while ago.
Very interesting post, Sukumar. Longtail also has the unique distinction of having been published entirely on the net before it was published as a book.
Your question is interesting. One of the first things that strikes me is that ‘Democratize Production’ is already entrenched in software industry, isn’t that what open source is all about.
I asked a friend of mine what Longtail was all about and he said this. People will not necessarily want to pay for everything. Lot of things which could be sold before is now available for free. The second assertion being small niche players will hold a large portion of the market share.
I have not read the book, it did not sound like my kind of book. Still not been able to complete, Dealing with Darwin.
Thanks Archana. As usual you are the first one to comment.
My question is about the effect of Long tail on the software outsourcing companies like Cognizant, Infosys, TCS etc. and not for software companies per se. I think open source definitely helps address the Long Tail of software products – if my needs are simple enough and i don’t have too much money to spend, i can use a open source software and not buy expensive software. By contrast, you go to a software development outsourcing company only when you have money to spend. So given that context, how do we apply the long tail concept to our industry.
I don’t think niche players will hold lots of market share. It is just that niche players have a chance to succeed in the Long Tail world whereas in the past they may not have had access to the distribution channels at all. For instance, if i make a video about my Pet Hippo – a major television channel may deny access, but i can get it published on Youtube and gain distribution. The reason i know about the Pet Hippo video is because Priya Raju found it on Youtube. So the maker of the Pet Hippo video found his/her consumer(s) using Youtube which might have been impossible in the Hits driven world. This is the essential point of the Long Tail phenomenon.
Hope that makes it clear?
Yes, I understand that Sukumar. When I said niche players would hold a large market share, I did not mean a single niche player. Lets say if we were to split the revenue generated by a particular industry into a pie chart. In the older economy, large portions of this pie chart will be held by big companies and there will be small portion referred to as others (too small to name them individually but aggregated they would contribute to a single digit revenue). In the Long Tain economy, the others will be much larger. Many small companies which carter to a focused niche segment would contribute to a much larger section of the pie. This makes sense because each small player is able to find his market in this economy.
Regarding your question about Long Tail and outsourcing companies – have you heard of Arzoo.com. It was started by Sabeer Bhatia and the purpose was to link freelance programmers to their market. From what I heard, lot of Russian freelance programmers used Arzoo.com. This appears like a Long Tail model. I heard Arzoo.com was a failure, no idea why.
Sukumar,
Interesting post and good question. Here is my take –
I think outsourcing services organizations need to start catering to the very small to small businesses and niche players and hopefully the revenue generated by such businesses could pay for the infrastructure needed to service such clients and still return a profit.
To enable this, outsourcing organizations should provide these services in a manner that is easily consumable and almost nil maintenance. In addition the framework should also be nimble enough that it could be easily adapted to specific requirements that a niche player could want. Of course the web is well suited for this and SAAS (software as a service) could provide the framework to make this a reality. However, the real trick is developing a reasonably all-encompassing framework that will provide service to and satisfy disparate requirements. Developing a unique service framework for every different business needs will not scale. Can such a framework service a mom and pop tax service provider, an insurance agent or a financial service provider?
I believe the infrastructure both in terms of software and hardware can be common for these varied businesses; however the application services needs to be customized quickly for each of them.
Now, do I think outsourcing companies will dedicate resources towards such a service model – probably not in the near future, primarily because the ROI is not clear on such investment
Ganesh
Arzoo is now a travel site catering to consumers of Indian origin. It was on our (mine + my client’s) watch list for the upcoming Indian Travel industry along with couple of others. None of them, unfortunately have been able to make a big impact in the space. Still in the long tail 🙂 Back to the point, Rent-a-coder and similar broker sites all will be part of the ‘long-tail of offshoring’.
Let me think up other points and write something to win my copy of the book. (will be my 3rd copy, have given away my first 2 to clients)
Thanks Ganesh. SAAS is definitely a potential solution as you pointed out. I guess Salesforce.com, Amazon and eBay storefronts are an example of this phenomenon. But the question remains, if i
want to do custom work, what do i do? As Archana and Sibu have pointed out, Rent-A-Coder, Topcoder etc may be able to service these needs.
Sibu, thanks. Look forward to your other points.
Sukumar,
I hope you will also give consideration to the first commenter while deciding on the prize winner 😀
Sukumar,
This was the question – “how does the Long Tail affect software development outsourcing companies”. I was thinking more in the context of established outsourcing companies and particularly the big ones such as Cognizant. What do they have to do to get the ‘small’ dollars from the ‘little’ customer which when summed up will be significant.
I was trying to address this from a different angle, rather than looking at niche areas that a single service provider can address. I was looking to see if Cognizant can address the various niche areas/customers and still profit from it. The “long tail” comes from a single provider servicing millions of such small customers – hopefully. Let me know if I misunderstood.
Do you think companies like yours will ever consider this?
Ganesh
Thanks Archana. Of course, commenting first counts 🙂
Ganesh, thanks. Your understanding of Long tail is correct. Now I’m beginning to understand your point. If there’s a strong business case, I’m sure Cognizant will consider it. But the problem I see is, how do we create a generic enough framework that can meet the custom software needs of small businesses that could be quite varied?I’m wondering mainly about the technical feasibility of such a framework.
Maybe we could create a framework for particular horizontal or vertical markets? I think this is what Salesforce.com has done by focusing on CRM as a horizontal market?
It is possible that I haven’t understood your point still? Maybe you can give some specific examples if that’s not too much trouble.
Ganesh has hit the nail – How does a large company like Cognizant take advantage of the Longtail? The challenge is definitely in perfecting the business model – in Chris’ terms, cutting the cost of distribution dramatically. Currently, we can serve large Fortune 100 customers with a dedicated customer team because we can afford to do so from a P&L standpoint. When it comes to small&medium customers – having dedicated delivery, sales & management teams may not be feasible because of financial reasons.
Sharing the resources becomes one obvious thought.
Delivery: Sharing delivery teams is not that tough – they are anyway based primarily offshore and the sharing model doesn’t have to be perfect. I’ve already seen this in action, primarily in Infrastructure Services (not for Longtail reasons though).
Sales: Sharing the sales team is also not that difficult, Creating the right incentive model for the Sales force to go after the long tail will be the tough part. So will have to rely on some lower cost sales channels.
Management: This is the expensive resources who has to be shared judiciously. Enabling Junior Management Resources to service the long tail customers is one way to do it, but eventually Customer Management is an activity that is best done by Senior Resources and figuring out a way to balance this will be key. One way is to do a geographic division of labor.
Making the ‘custom development’ and management of the same as automatic as possible is another thought. Web based project management tools exist today that can be leveraged effectively. Rolling out the right tools, Making sure the engineering & tool disciplines are followed and adaptation of the tools will be key challenges to overcome.
Automatic Servicing should also include simpler legal and contract issues – One significant challenge to overcome. Even Financial Operations should be made simpler – maybe as extension of the same toolset for project management. One click a month by the customer and he should be able to see the invoice, and pay by CC just like an e-commerce site.
So far, I’ve written about addressing the pure long tail customers – how to service the not so large and small customers and still make a profit. To me, longtail also presents some incremental opportunities within the existing large customers itself. For one, longtail as a phenomenon is impacting your current customers business, so helping them address the challenge presents unique opportunities. (have to state that i am fortunate to have couple of customers in this category)
-Sibu
Thanks Sibu. I guess combined with the Generic Framework that Ganesh talked about and low-cost sales and distribution model we can make an attempt to service SMBs.
You are right, there are definitely great opportunities for our customers to service their own long tails and that may present an opportunity for companies like Cognizant. I am glad that you are already seeing such opportunities. I have seen some as well.
I must admit it took sometime for me getting back to the blogs that I follow. Well Work of course is a necessary evil for my laziness in going through posts. Sukumar, as usual you have the ability to throw up something totally new. I have been looking for the presentation on “The Economics of Abundance” by Chris Anderson. Great to find it here. At least I’m happy to view it. 🙂
Thanks for your kind words Samiran. Thanks for stopping by. Now that you found the presentation, maybe you can participate in the contest?
Sukumar,
I agree with you that building a robust and flexible framework is the main challenge from a technical standpoint. I think the framework could be built/layered starting from the lowest layer –
* Database
* Object to Relational Mapper (ORM) – supports type inheritance, multi-value attributes, containment relationship etc.
* Generic Business Object (GOB) layer
* Business object layer specific to a particular vertical
The database and ORM provides the foundation for the framework and the GOB provides the mechanism for quick customization to a specific vertical.
I have not thought too deeply about sales, marketing, business development and strategy etc. – how do you make a business case for such a technology, what is the potential revenue, how much will the ‘small’ customers be willing to pay, why would they go with a big company like Cognizant rather than a niche provider etc. The last one will probably be the biggest blocker for a small customer – how much can they trust the big gorilla to provide what they want at a price point that is acceptable, but still treat them like a big customer!!
Ganesh
Hi Sukumar! I chanced on your wikinomics post just now. I hope I won’t break the rhythm of your blog by commenting on it here, since I cannot post a comment on that post anymore. I had actually blogged about Wikinomics back in Feb ’07 and had added a link to an Harvard Business Online Ideacast on the Wikinomics concept. Perhaps you could add the Ideacast link to your original post.
The idea of the long tail is an interesting one. Certainly presents a juicy technical challenge!
//Shuchi
Shuchi,
What a surprise. thanks for stopping by. I will edit my wikinomics post and provide the links you suggested.
Thanks
Thanks for the clarification Ganesh. As you say, we can create a generic platform and keep adding subsystems for specific vertical/horizontal markets. I agree, a Cognizant may not easily get interested in such a thing.
I am still not satisfied with the ideas we got so far. There has got to be some more, that can easily appeal to a Cognizant ?
Shuchi,
I edited my post to include a link to the ideacast as well as your blog post on Wikinomics. I also opened the comments on my wikinomics post.
Hi Sukumar,
There was a program on NDTV this morning about Long Tail Phenomenon in the publishing industry. It talked about how content is now being sourced from many small time writers on the net. It was titled ‘The Davids tie up with Goliaths’
Mind blowing Sukumar. Matter of fact, I though all through my treadmill session about the applicability of Long Tail model to software outsourcing industry. My observations…
Major Categories of Software
1) Prepackaged Commercial Software
Typical Build or Buy decision usually favors buying off the shelf. MS Office etc
Scope for outsourcing is 100% including support we pay for.
2) Packaged Custom Software
Customize minimally and maintain with outside support. Like Peoplesoft, ERP.
Scope for outsourcing is 60-80% including support we pay for. There could be some in-house support
3) Custom Software
Ground up development that goes through life cycle management. Industry spends a major chunk of their budgets on this. Unless all this is put in one custom IT bucket (excluding the cost of licenses of 1), 2) ,hardware and infrastructure, it will be a good slice in companies that are in dynamic business environment.
Scope for outsourcing is 40-60% including support we pay for. There could be some in-house support here as well. Long tail here may be there but quite challenging to exploit.
4) Next Gen Software with high level of service reuse
With Web 2.0 family of services and API one could develop a business idea into a company with minimal investment and maximum returns. Here idea is the king. Implementing an idea using Google APIs, eBay API, Ma href=”http://www.amazon.com/AWS-home-page-Money/b/ref=sc_fe_l_1_3435361_1?ie=UTF8&node=3435361&no=3435361&me=A36L942TSJ2AJA”>Amazon Services and the famous Facebook API etc building a web 2.0 company is weaving of services and mashing of data into something which has intrinsic value to the customer. I could quote an example.
One of my friends, owns a Deli at Tampa. Of course there are 100s of Deli’s everywhere. But his family since 3 generations owns a hot sauce recipe which according to him is a huge differentiator. He wants to expand nationally. For him building an e-commerce site with the help of Amazon/ eBay is very easy. Once he test markets the product, he can really decide on how to take his company forward. This entry barrier is reduced to a great extent these days. Still he may want to maintain the unique identity of his company. There will be 100s of such opportunities we as large service providers will not even look into.
We can have subsidiaries like what Toyota does. Toyota makes Lexus. Buyers of Toyota want to buy a usable car with high reliability and mileage. But users of Lexus want luxury and experience. Bottom-line will be very high in Lexus. But the number of buyers will be very small. On the other hand we know that Toyota is the largest car maker in the world (even surpassed GM twice this year and will beat GM throughout next year). Indian outsourcing providers can also create subsidiaries where the onsite/offshore ratio will be 5:95 using cheaper technologies like email, Webex and other chat clients, develop from tier 3 cities, train teams which can implement these eCommerce APIs, then send the product to specialized teams who will customize/ skin them to the client’s requirements, testing teams which will then do final QC and deliver. This could a factory model and has potential to tap the long tail mass market.
Your thoughts,.
Vamsi,
I was thinking along the same lines, but partnering instead of subsidiaries. In the architecture that I had proposed with the database layer, Object to Relational Mapping layer etc., I was wondering how we could scale building the business objects layer specific to a niche market/vertical. The answer – partnering. This is more of a franchise model, where Cognizant provides the infrastructure to to build your company for a niche market quickly. The service provider can either build the additional layers themselves or leverage others.
In terms of revenue, a model needs to be devised for revenue/profit sharing between the infrastructure provider and the service provider.
Ganesh
Interesting Post Sukumar. I am assuming that the long tail in this case is the Small to Medium businesses (SMBs) that large outsourcing companies like Cognizant ignore and could potentially tap to grow further. I am also assuming SMBs are companies that have annual revenues of anywhere between 5 Million to 50 Million dollars. These SMBs currently get serviced by small players (mom & pop software shops) or use off the shelve products & tweak these products by having a very small in house IT team. These companies are a step above using services like rent-a-coder or top coder.
Chasing the SMB market is definitely not in line with the business model of the large outsourcers. The current business model of the large outsourcers that rely on high touch customer relationships definitely does not support servicing these numerous small “David’s”. The whole business premise is based on finding the next Goliath. For example, in Cognizant we pride in having a small customer base and deepening our relationship further with these customers. Given these factors, we need to explore the reasons that could make the SMBs attractive to Cognizant & other large outsourcers. Couples of thoughts come to my mind on what could lead the large outsourcers to consider the SMB market.
1. The head of the curve is getting saturated. The large outsourcers need to look at other market segments to deliver the 30 – 40% growth that the market is accustomed to. We have seen moves by the large companies in looking at complementary services like IT IS, KPO, Market Analytics etc to sustain the double digit growth.
2.There is huge revenue potential in the long tail. If the companies can get the model right, there is potential for exponential growth.
3.Instead of venturing out into complementary service areas, these companies can focus on their core strengths of developing application software to service the SMBs.
If large outsourcers are interested in the SMBs, then the next big question is the service model that could be used to service the SMBs. Some thought on a potential service model are given below.
1.Borrowing on the “Economics of Abundance” theme, companies like Cognizant could tap into their large unutilized talent pool of programmers. Given the current offshore utilization rates of around 60%, we can safely assume that at any point in time, there is a talent pool of at least 1000-3000 programmers that are not fully allocated to any project. These numbers also include the entry level trainees.
2.Create SBAs (Small Business Accounts) that can matrix the unutilized talent pool to deliver custom solutions to the SMBs on an as needed basis.
3.The services can be offered by creating a “software services exchange” of sorts where the SMBs can post the custom requirements. The SBAs will evaluate the customer requirements & create the project team to deliver the customer requirements by utilizing the unallocated talent pool that is available.
4.Instead of the traditional high touch approach that we rely on for large customers, the SBAs can rely on the low touch, product service approach to service these customers (could be in line with what Sibu or Vamsi proposed).
5.The SBAs could also look at potential incubation opportunities for small businesses that have an exciting idea that they need help in taking forward.
The advantage for the large outsourcers is that this model opens up a potential channel for revenue and also has the added benefit of getting its entry level / less experienced programmers trained on real cases without exposing them to the larger strategic clients. Since the relationship is low touch, the resources assigned to a project could easily be rotated out if there is a need in a larger strategic engagement. Questions do come up on Brand dilution, potential market plan for SMBs, why would SMBs go with big gorillas as Ganesh alluded to in his post.
Thanks Archana, i don’t think i have seen that program. Vamsi, Ganesh and Joe, thanks for the detailed comments, i will respond when i return from my vacation.
Great post and comments, all.
I chanced on this post a few days back and have been thinking on how it might apply to a company like ours. Vamsi, Ganesh, Sibu and Joe have got it pretty much covered on the sell side.
More on the Sell side:
Companies like ours with a lower cost model and the global delivery model probably started exploiting the Long Tail of the F-1000 and the F-2000 for IT Outsourcing, as these end user companies probably could not afford the IBM, EDS, etc. of the time to provide them integrated solutions.
Today, as TCS, INFY, etc. have graduated up the value curve, and are dropping smaller non-growing customers, other service providers like iGate, MindTree are servicing those abandoned by the larger compaines, but have their own significant branding and growth challenges.
My take is that given the relationship-intensive model of companies like ours, there will be a limit to how small a customer companies in our model can service. Unless the model changes, the Long Tail will be difficult to service cost profitably.
However, what if we look at the Supply side:
In the Global War for Talent, and here I surmise, maybe our physical reach goes into 20% of the Supply pools from where we get 80% of our Talent. Whether at the ELT stage or the lateral hire stage, we probably have strategies, programs that ensure that the numbers we require are being met.
By thinking about Chris Anderson’s 98% rule, what if the company brand reaches into portions of the market that we are today not able to physically cater to? Each and every one of use had heard that if you respond to an Open Position on a company’s web site, you will never get a response, or if you do, it is a vanilla boilerplate email.
Are there screening programs and methods that a company could adopt, that would actually reach out to the balance 80% at all levels? Are there cost effective methods to do this efficiently? If so – would it not be possible for a company to meet its Talent numbers, but probably at a 3% or 5% or maybe even 10% better Talent quality? What difference would that make to the quality of future leadership of the company?
Today, we probably reach into a certain number of named universities, and go to a certain number of cities to recruit. This method has a clear cost, timeline and dedicated resources. The Long Tail here are those people in universities we don’t visit, whole cities we don’t go to, resumes that go unresponded to, that our brand today permeates.
If that were the Long Tail of our Supply side, what if we could build channels for those people to come to us? Maybe we could have dynamic online tests (to avoid plagiarism) that people could take, that the company could then screen or be in touch with for the future.
Now, I am getting into details, but I wanted to suggest the Supply side for companies in our model as where the Long Tail might be relevant. Your comments and responses welcome.
– Shridhar
More on the Long Tail on the Supply Side – for consumer goods companies. You’ve probably heard of what P&G is doing for product innovation. They concluded that in-house innovation facilities (number of scientists, problems being solved) were not going to be sufficient to meet marketing goals, and decided to post on the Internet – what kind of products with attributes they were looking for. Detergents that could do this, toothpaste that could do that, etc.:-)
Scientists, chemical companies, and individuals from around the world have contributed to that site, and with the legal framework in place, P&G expects that half of all new products would have some component of ‘not made here’ in them in the near future. I don’t have the statistics, and have not heard or read of P&G use the Long Tail of Production, but I would like hear your thoughts on this, cheers.