TC-TS Marketing

Building on my earlier post, we’ve had some discussions internally on which market segments to target with the Thin Client-Thick Server (TC-TS) solution, and how best to do it. [Also see the TC-TS Economics post, so as to get an idea of the solution pricing.]

I have categorised this discussion into two parts: Generics, wherein I discuss the broad usage categories, and Specifics, wherein I address the first few organisation types that we need to address. Finally, there are some ideas on how we need to proceed.


  • Terminal Market: Wherever people are using “dumb” terminals, we can now move them to “intelligent” terminals, capable of doing much more since they have a GUI. Another way to think of it is to look at these areas where server-based computing is being used. This could mean replacing Unix terminals, Windows terminals or even Novell Netware terminals. Retail stores could put a terminal at different locations in the store. Banks could install it for automated banking or deploy it at the front-office. Call centres could use it as the desktop for their staff.
  • Shared Access Market: This involves targeting these segments where a computer needs to be used by different people at different points of time. Another way to look at it is to consider the solution to fulfill the computing needs for a floating population. Think of hotel rooms, airline lounges, cybercafes, schools and colleges. The same computer is shared by many. Since data is stored on the server (should the need arise), there is privacy for each user, unlike on a Thick Client hard disk.
  • New “X” Market: The X here means Users, Offices, Residential buildings, etc. Look for new buyers, companies setting up new branches/offices or hiring new people. These are users for whom the total cost of ownership can be lower by upto 67% over 3-years, so there is a clear economic driver. Also, new users may not have the baggage of “unlearning” Windows.
  • Low Usage / Few Apps Market: For those users who just need to have a basic set of applications (email, browsing, letter writing), the TC-TS solution can be more than good enough. Or, it could even be targeted at situations where there is the need only for text-based access. In these situations, cost becomes an important consideration.


  • Hotels: Hotel rooms equipped with a computer connected to the Internet, such that when one walks in, it is possible to be browsing in no more than 2 minutes. There is no need to boot-up one’s own computer (if one has it) if the need is for checking web-based mail and browsing. Would I pay USD 10 (Rs 500) per day extra for such a service? Sure! Today, just as a phone and cable connection is assumed in a hotel room, it now becomes possible to give not just an Internet connection but a computer in every room. There is a clear return on investment path for the hotel. Another interesting application can be to create an Intranet with the help of nearby shops to give guests an idea of what’s on in the neighbourhood, with the ability to buy and have the items added to the hotel bill. The challenge: networking (doing the cabling).
  • Engineering Colleges: Computer education is mandatory in the first year programme of most engineering colleges. This is where the Thin Client can be used to give exposure to Linux. Students can also set up TCs in their hostel (dorm) rooms.
  • MailServ Customers: We have 100+ corporate customers for our messaging solution. While our initial foray into talking to them was not too positive because of the need for supporting Windows applications, we cannot just ignore the corporate segment. Corporates, after all, account for 75% of computer purchases in India. We need to go to them with support for Windows apps. The approach should be to (a) see if we can use their older machines – this is where they would want to use Citrix today, (b) target non-users, thus helping them downstream computing in their organisation (c) look for users just needing the basic apps so they can be moved to a TC, and (d) solve the Microsoft licences problem with the TC-TS solution. A point to keep in mind: “delight” users, not “disappoint” them. Once we make a few breakthroughs with our MailServ customers, we can then target other corporates. For better chances of success, we need to see the intersection of the “generic” markets with the corporate customers that we target.


    I’ve been thinking of a multi-pronged approach going ahead:

  • Get the Windows applications working on Linux so we can talk to our corporate MailServ customers with confidence
  • Talk to Hotels and Engineering Colleges in Bombay to begin with
  • Create a Mailer identifying the benefits of the TC-TS solution, and send this out to a database of about 1,000 organisations, inviting them to our weekly seminars (see below)
  • Talk to intermediaries (assemblers, SIs) – they are prospective channel partners
  • Set-up weekly seminars (every Saturday morning in our office) so we can showcase the solution and talk about it in detail to interested organisations
  • Send the TC-TS CDs to a few organisations outside India, so we do not just get limited by geography, and begin work on “internationalising” it also

    In short, we need to think creatively. At the heart of our solution is the concept of a “computer on every desktop”. Think of those situations where a computer (rightly priced) could add value. So far, people in emerging markets like India have been used to thinking of computers as luxuries. We want to make it a necessity, a utility. This means thinking bottom-up of all the new markets in which computing can make a difference. One point to keep in mind is that we need 10+ users as a start to make this work in a single location.

    Continue reading

  • TC-TS Economics

    I’ve put together some thoughts on the Thin Client-Thick Server solution from a pricing and total cost of ownership point of view, as a precursor to identifying the early markets (will be posting thoughts on this shortly).

    First, a quick recap on the Strengths and Weaknesses of our TC-TS Solution.


    – reduces cost of hardware needed on desktop / can use old PCs
    – reduces cost of software by leveraging open source
    – Windows-like GUI reduces learning time
    – provides base set of apps (email, browser, unified IM client, word processor, spreadsheet, presentation, PDF reader)
    – can read/write most MS-Office documents
    – can co-exist in a Windows world; this is not an either-or solution (either Windows or Linux TCs)
    – centralised administration can provide IT staff easier and greater control on desktops
    – can control the client desktop entirely from the server
    – centralised storage of data makes backups easier


    – need a minimum of 10 users (ideally, 25) to amortise cost of Thick Server
    – LAN speed of 100 Mbps needed due to high data transfer
    – Limited support for Windows applications
    – Non real-time situations (since the TC can occasionally crash – we are working to fix this)
    – No multimedia support on desktop (this is actually possible – not a technical issue, just one of TC cost)
    – needs an admin with some basic knowledge of Linux – and an open mind (because we are not Windows!)


    An old PC working as a TC is available for USD 150 (Rs 7.5K). These would be typically low-end Pentiums of mid-to-late 1990s-vintage. Newer stripped-down PCs based on Via or AMD processors can be available for USD 200 (Rs 10K), with an additional USD 100 (Rs 5K) for a new keyboard-mouse-monitor.

    The TS investment would be about USD 1,200 (Rs 60K) for a dual-CPU PIII 1 Ghz server, with software RAID on 2 80 GB disks and 1 GB RAM. Such a server can support a minimum of 30 TCs.

    Our software should be priced at no more than USD 40 (Rs 2K) per user per annum, or USD 4 (Rs 200) per month. The monthly/annual subscription fee ensures regular updates of the software applications.

    So, if we assume that companies / institutions get their own TCs, we can think of a rental model on the TS and software. Assuming 10% p.a. interest on the server pricing and our need to recover the cost in 2 years, this implies a monthly cost of USD 60 (Rs 3K).

    So, if we consider the 3 options of 10, 20 and 30 users in an enterprise, the TS and software rental costs on a per month basis would be as follows:
    – 10 users: USD 60 + 10*4 = USD 100, or USD 10 (Rs 500) ppm (per person-month)
    – 20 users: USD 60 + 20*4 = USD 140, or USD 7 (Rs 350) ppm
    – 30 users: USD 60 + 30*4 = USD 180, or USD 6 (Rs 300) ppm

    This does not include support or training costs.

    Thus, an organisation can now provide hardware and software (a full computing solution) for a one-time cost of USD 150 (Rs 7.5K) for the TC and an annual cost of no more than USD 120 (Rs 6K).

    As a comparison, a Thick Windows desktop would cost USD 600 (Rs 30K) with an additional USD 500 (Rs 25K) for software [Windows XP, MS Office and Anti-Virus].

    3-year TCO

    Taken over a 3-year period, the Total Cost of Ownership, assuming a 10% hardware AMC, is as follows:
    – TC-TS Solutions: USD 150 + 15 + 15 + 3*120 = USD 540 (Rs 27K)
    – Thick Windows Desktop: USD 600 + 60 + 60 + 500 = USD 1220 (Rs 60K)

    The TC-TS option could be even lower if one assumes 30 users (the above assumes 10 users). For 30 users, the 3-year TCO is USD 150+15+15+(3*72) = USD 396 (Rs 20K).

    Thus, the TC-TS solution can bring down total cost of ownership over a 3-year period by 55-67%.

    TC-TS R&D

    Some of the areas that need more work on the technology side are:

  • Optimising Memory Management, so that when the apps like OpenOffice and Evolution run on different Thin Clients, how can we ensure that there is only one instance on the Thick Server. Currently, each apps is loaded into memory every time it is executed on the TC.
  • Reducing LAN Bandwidth needed from 100 Mbps first to 10 Mbps (so it can run on WiFi) and then to 1 Mbps (for cable modems). There is a lot of data transfer happening between TCs and the TS. In a corporate on a 100 Mbps LAN, it doesn’t matter. But in the other situations where we want to target homes or look at a wireless LAN, the bandwidth used is going to make a difference.
  • A closer look at VNC (virtual network computer) software and seeing how we can use it better. Today, we use it to get a Windows desktop on the TC.
  • Architecture and Server Sizing, to support 100+ users. Today, we assume there is a single TS which can support 35-40 users. What if there were more users? How can multiple TSs work together? How do we create a scalable architecture on the server side? We may need to use some Linux clustering ideas.
  • Windows applications support on Linux, either through open source solutions like Wine or commercial solutions like CrossoverOffice, Win4Lin, and others. We need to take common applications like Foxpro, Visual Basic, Oracle, MS-Office and Tally (in India), and make sure we get them running on Linux. We cannot go into enterprises without support for Windows applications.
  • WiFi and Power-line modems testing, to meet the challenge of cabling. It may not always be possible to lay out new Ethernet cables for the TCs. In this situation, we should see (assuming cost were not a constraint) how we can use some of the newer technology to connect the TCs with the TS.
  • Using Mozilla as a platform to build applications, with one idea being to integrate a micro web server into the browser to provide real-time two-way updates. Much of this would be useful as we build out our digital dashboard.
  • Support for other languages, so we can take it internationally. Here, we need to especially see how to support Unicode / double-byte.

    Continue reading

  • Real-Time Enterprise: Bajarin

    From eWeek:

    Where real-time operations were once the province of nuclear power plants and mission-critical manufacturing processes, the real-time concept has been sneaking into the vocabulary of the “next big thing” technology futurists. The most recent explanation I’ve heard was during a presentation by Creative Strategies President Tim Bajarin at a CRM expo in New York last week.

    In Bajarin’s view, a combination of standards-based integrated applications and wireless devices will let corporate “corridor warriors” always know the financial health of their companies, suppliers and customers.

    Continue reading

    TECH TALK: Rethinking the Desktop: Game Over?

    Amongst all the technology innovation of the past decade, the one area that has been relatively untouched is the screen which we see everyday when we start our computers the desktop. The collection of files, directories, folders and icons has changed little in the past decade. Yes, we have gone through multiple versions of Windows (from 3.1 to 95 to 98 to 2000 to XP). But the basic underlying concepts have remained the same. Desktop innovation has been noticeable by its absence.

    I can think of two reasons for this. Firstly, much of the focus for innovation has been on the Web, though the Web browser seems to be also now frozen in time. The only interesting idea that I have across has been Mozillas tabbed browsing, which allows one to open up new windows as tabs making clicking across windows a wee bit easier. Secondly, Microsoft has won the desktop game, and so there seems to be little interest by other companies to see what can be done. Distributing a new product will probably be well nigh impossible since the Windows desktop operating system comes pre-installed on most computers.

    If we look at the two desktop OS alternatives, Apple has of late been trying to come up with innovative ideas for its Mac platform with its OS X, while Linux has had two desktops in the form of KDE and Gnome. While OS X has genuinely tried to make things easier for users, its limited platform availability makes it a non-started. The Linux desktops have simply tried to copy the look-and-feel of Microsofts Windows. The underlying feeling behind all this is that the desktop game is over.

    The desktop game is over only as far as the current set of computer users go. For todays 500 million users, the Microsoft Windows-powered desktop has become more of a gateway to connect to the Internet. The email client (Outlook or Outlook Express) and web browser (Internet Explorer) have effectively become the applications there they spend much of their time, complemented by MS-Office in the workplace and Instant Messaging in homes. And as history has shown, it is very difficult to change entrenched habits.

    But the desktop game is not yet begun if we consider the next 500 million users. This is an audience that has only limited access to computers, if at all. They are in the emerging markets of the world, waiting for a dramatic reduction in the price of computing to adopt the technology. They have no love for one OS over the other; they have no baggage of the past.

    So, can we take many of the discrete innovations that have been happening and apply them to give a different and better experience for these new users? They will be, as a group, less savvy than the current set of users. They will also be children of the Internet making little distinction between the LAN and the WAN. Their first taste of computing will come in a connected world.

    Tomorrow: The Decade That Was

    Still-hot Tech Sectors

    Forrester highlights tech hotspots (via – security patching, search and UWB:

    Forrester predicts that startups specializing in security patching, that is, filling security holes in corporate enterprise systems, have a bright future. In one memo, Rutstein notes that exploitation of a known security hole causes nearly all security incidents. Yet large enterprises are consistently vulnerable owing to the complexity of finding, assessing and testing those patches.

    In another memo, Forrester software analyst Joshua Walker points to a surge of corporate interest in search technology. Buyers include multinationals that can’t find critical intellectual property on their intranets and online sellers with buried merchandise. Walker contends that a single type of search engine meets all large enterprise needs so that there’s room for several winners. He also notes that the rewards can be large, citing two startups in the sector, Verity Inc. and Endeca Technologies Inc., which have average-size annual contracts of $385,000 and $450,000, respectively.

    Forrester also projects opportunities for VC investment in a wireless subsector: ultra wideband, also known as UWB or digital pulse wireless. That is a technology for transmitting large amounts of digital data over a wide spectrum of low-power radio frequency bands for a short distance. It can carry signals through doors and other obstacles that tend to reflect signals at more limited bandwidths and at higher power. Rutstein notes that while the military has used UWB for many years, new innovations are bringing it into the commercial sector.

    Continue reading