[kictanet] TCO & Cloud

Francis Hook francis.hook at gmail.com
Tue Jan 31 12:37:39 EAT 2012


Hi Robert,
Thanks.  I think for every discussion there is need for some
dissenting points of view which help refine the discussion and also
allow multiple points of view.

The issue of clouds, whether public, private or hybrid, merits some
debate (and I seem to recall there was one such debate hosted here on
KICTANET but perhaps did not look at vertical specific scenarios or
infrastructure issues as such - its been a while - maybe it did).  In
some sectors it may be accepted as an elixir for infrastructure
management and cost issues, while in others it may result in over
exposure.

Rgds

F

On 30 January 2012 18:19, robert yawe <robertyawe at yahoo.co.uk> wrote:
> Hi,
>
> It is encouraging that you decided to post your thoughts on my post unlike
> many others if opted to call me directly to air their disquiet.
>
> I would like to belief that the systems our banks are using are centralised
> with all branches online to the data centre, until recently when a few
> providers implemented MPLS, the connectivity to those branches was point to
> point.
>
> With the point to point connectivity there was a minimum of 2 hops to get to
> the data centre, if those links delivered then imagine is the data centre is
> at, for example, Safaricom then you have an minimum hop of 1.
>
> The Kenstream & KenSat services when run by Telkom provided us with 95%
> uptime and that was without any redundant links.  The GSM providers need to
> keep their links up 99% of the time if only for the voice service and since
> the data is carried over the same fibre, microwave and satellite links.
>
> Taking that into consideration I believe we have no reason not to embrace
> cloud and shared services, yesterday.
>
> Regards
>
> Robert Yawe KAY System Technologies Ltd Phoenix House, 6th Floor P O Box
> 55806 Nairobi, 00200 Kenya Tel: +254722511225, +254202010696
> ________________________________
> From: Emmanuel Khisa <oloo.khisa at googlemail.com>
> To: robert yawe <robertyawe at yahoo.co.uk>
> Cc: oloo.khisa at gmail.com; KICTAnet ICT Policy Discussions
> <kictanet at lists.kictanet.or.ke>
> Sent: Friday, 27 January 2012, 17:45
>
> Subject: Re: [kictanet] TCO & Cloud
>
> Hi Bobby,
>
> I am impressed with your understanding of the financial sector ( Not many IT
> people in Financial sector have such a deep insight of IT Ops).
>
> I would like to differ though on the issue of cloud. One of the greatest
> issues raised against cloud and which is partly true is the connectivity
> (and reliability of the same). Let's say you head IT in a bank and have to
> manage the downtimes in places like Meru, Bungoma, Wajir etc, running off a
> Safaricom cloud may be abit of a longshot.
>
> Safaricom has done a great deal to bring us connectivity especially the last
> mile, I however think something is still amiss with the reliability of
> connection. Every executive committee will raise (and validly so) the issue
> of Connectivity and letting go of customer data to counter that cloud
> argument. That needs to be addressed and the only company in Kenya today
> that can take us into the cloud is Safaricom, I hope they listen and fortify
> this data segment like they did voice.
>
> On a lighter note, having used Temenos, I can take my money to the bank that
> finacle would beat it hands down. I don't care how many big financial orgs
> are using it, I have baby sat it and the truth of the matter is that it is
> like a Kluger (no offence to owners of this) very overrated and big in size
> but has the performance of a saloon car...
>
> Temenos are good at marketing but the real value is elsewhere.
>
> On Fri, Jan 27, 2012 at 1:09 PM, robert yawe <robertyawe at yahoo.co.uk> wrote:
>
> Francis,
>
> Most of the issues you mention on customer service are only applicable with
> a certain level of account holder who we would refer to as your corporates
> who make up less than 1% of the accounts.
>
> The other 99% are mainly salary account holders who receive a single inflow
> once a month and spend the rest of the time drawing down the amount.
>
> With the proliferation of ATMs, mobile and agency banking this kind of
> account holder is already being served through a cloud solution.
>
> Outages within most organisations are today much more frequent than those
> involving external providers, most ATMs are down not because of the
> connectivity but more for lack of cash. With services such as bank account
> to mobile money transfer and cash back the ATM is slowly becoming obsolete,
> at least in the Kenyan scenario.  Most ATM cards are now debit cards meaning
> that they are usable at various point of sale locations.
>
> As customers use less and less physical cash the easier it is to mine trends
> than before making it easier to extend credit to establishments since you
> have a better view of their cash flows.
>
> In the land of the blind the one eyed man is king until a two eyed man/woman
> arrives in the village, if we use the global definitions of SMB all
> businesses in Kenya, except mpesa, that we refer to a enterprises actually
> fall under this category and those we call small are actually micro.
>  Therefore from your take on the same then all businesses in Kenya should be
> embracing cloud solutions.
>
> MPesa is run as a cloud service no applications are installed locally right
> from Safaricom head office to the till in Gabatula, it has 34,000 plus
> branches, employs over 50,000 people, transacted over 600 billion last year,
> at peak times handles 200 transactions per second (Temenos can handle
> 3,000), has over 15 million accounts and made a profit last year of over 1
> billion.
>
> If MPesa was to be listed in the NSE it would be the most profitable
> organisation once you remove the burden of its parent.
>
> If this organisation can survive on a cloud solution even Kenya Breweries
> can move their systems onto the Safaricom Cloud (for lack of any other
> examples) especially since they are implementing a model similar to
> Safaricoms.
>
> Francis, lets accept it that we have no leg to stand on in as far as not
> using the cloud is concerned.
>
> Why have I not released my applications to the cloud? Because the only
> provider has been unable to provide me with a easy self service
> model. http://www.esds.co.in/cloud-servers-hosting.php
>
> Regards
>
> Robert Yawe KAY System Technologies Ltd Phoenix House, 6th Floor P O Box
> 55806 Nairobi, 00200 Kenya Tel: +254722511225, +254202010696
> ________________________________
> From: Francis Hook <francis.hook at gmail.com>
> To: robert yawe <robertyawe at yahoo.co.uk>
> Cc: KICTAnet ICT Policy Discussions <kictanet at lists.kictanet.or.ke>
> Sent: Friday, 27 January 2012, 11:28
>
> Subject: Re: [kictanet] TCO & Cloud
>
> Hi Robert
> I agree WRT TCO...perhaps what I overlooked to mention is ROI -
> increased visibility of accounts and account holders, Business
> intelligence (tracking bad debts, defaulters, speedy loan approvals
> for HNW and customers that pay their loans on time, etc), reduced
> queuing time thanks to having proper systems in place that enable a
> teller access requisite data needed to decide (ergo bringing a lot of
> decision making to front office = better customer service, customer
> retention, etc).
>
> By and large yes, the TCO argument should be weighed against the ROI
> (whether tangible is shs and cts, or intangible via better  customer
> service, increased staff efficiency, better GRC practices (esp this
> point as banks are increasingly under scrutiny), etc.  And yes, there
> is more gravitation towards OPEX from CAPEX (but at the core there
> might be some inherently irrefutable justification to take either path
> and this will vary from vertical to vertical)
>
> Back to the cloud issue - and  I'm now playing the devil's advocate
> :-> - thinking of instances that have been shared on this very forum,
> where the systems of an operator/provider/host etc go down (internet
> services, mobile money, voice services, govt websites etc) - from an
> institutional perspective, perhaps that is something to eschew?  If
> something goes wrong (at a connectivity, security level for example),
> and even with a back up site, then it more or less cripples the sector
> (and the economy).
>
> The cloud however, would be best positioned to SMBs who, owing to
> their size, needs and budgets, want to totally do away with CAPEX of
> this nature and instead have a pay as you go system (like you
> suggested where an operator hosts an app and charges end users on a
> pay as you go basis and then shares revenues with the app vendor).
> This would make sense for smaller banks (or SACCOs and MFIs).  Also
> other SMBs that need some CRM/Payroll/HR, etc functionality but do not
> want to buy the whole bus since they have no where to park it or no
> one to drive it.
>
>
>
> On 27 January 2012 11:03, robert yawe <robertyawe at yahoo.co.uk> wrote:
>> Hi Francis,
>>
>> I will disagree with you in the issue of security as we all know that its
>> not about where the data sits but how secure it is.
>>
>> Many a times a go to an ATM just to find a mini-statement sticking out of
>> the printer slot which tells you that the issue is not really account
>> visibility but how difficult it is to manipulate.
>>
>> Still looking at the security issue, who is likely to offer better
>> security,
>> some small fly by night operation like the one that developed the web
>> sites
>> for the recently hacked government web sites or a muli-billion telco?
>>
>> For micro institutions, less than 250 employees, like we have in Kenya
>> getting on the cloud does not require those text book approaches as most
>> of
>> the so called core applications are not mission critical which is why some
>> of our large banks can continue to operate even with this so called core
>> banking application offline.  Many of the organisations do not even have
>> an
>> online backup site.
>>
>> I consulted for a bank some years back who switched their core banking
>> application over a weekend with almost no disruption to their services.
>>  Once the data conversion process had been perfected all that was needed
>> was
>> 6 hours to upload onto the new system.
>>
>> There is a very clear reason why I had TCO (Total Cot of Ownership) in the
>> subject of my post, many organisations do the calculation you just did to
>> justify capital expenditure (CAPEX) and forget that there is an operating
>> expense (OPEX) such as insurance, maintenance, power
>> production/consumption
>> and none core staff to that CAPEX.
>>
>> If those funds where lent out to the members/customers at 12% pa the
>> organisation would receive and additional US$ 6.24 Million per annum while
>> saving on certain OpEx  such as 12% on comprehensive maintenance (6 M), 4%
>> on insurance (2.1M), office space (1,000 sq. feet @ 80/- = 0.8 M), none
>> core
>> IT staff (4 M) and power (4 M - KPLC & generator) which adds up to about
>> Kes. 17 million per annum.
>>
>> With the same provided as a service at a charge of 5/- per transaction
>> with
>> 200,000 accounts of which 30% are active each with 2 transactions per
>> month
>> adds up to 7.2 Million per year which is covered by the interest on the
>> money that has been used to acquire Temenos.
>>
>> The Kes. 17 million saving from the OPEX will bring in an additional 2
>> million in interest.
>>
>> All this just by using my form 2 accounting knowledge, imagine what I
>> could
>> come up with if I applied my advanced finance skills and was actually
>> being
>> paid to do it?
>>
>> Regards
>>
>> Robert Yawe KAY System Technologies Ltd Phoenix House, 6th Floor P O Box
>> 55806 Nairobi, 00200 Kenya Tel: +254722511225, +254202010696
>> ________________________________
>> From: Francis Hook <francis.hook at gmail.com>
>> To: robert yawe <robertyawe at yahoo.co.uk>
>> Cc: KICTAnet ICT Policy Discussions <kictanet at lists.kictanet.or.ke>
>> Sent: Friday, 27 January 2012, 9:59
>> Subject: Re: [kictanet] TCO & Cloud
>>
>> Hi Robert,
>> Approaches to the cloud have to be rather cautious...especially in the
>> financial sector and other sectors that handle personal info.    Some
>> organisations will opt to have a hybrid crowd where they have their
>> private cloud (with Temenos, HR systems, etc - sensitive stuff) and
>> perhaps farm out to a public cloud apps like CRM, Sales processing,
>> customer feedback  interfaces, etc
>>
>> On the issue of K-rep - if they have 200,000 accts and the system cost
>> Kshs 52 million - this works out to Kshs 260 per account holder -
>> something that can be somehow "loaded" onto ledger fees over say three
>> months without too much furore - but in the long term the customer
>> benefits from more innovation in products, greater responsiveness to
>> their needs by the bank, timely and accurate info, etc etc.    For the
>> banker's sacco...hmmm....I think then we have to ask how small is
>> "small" - at that level it works out to abt 2,600 - looks like a big
>> number but I suppose the account holders are all bankers :-) with good
>> deposits (vis a vis Krep's who might largely be low/middle income).
>> But I agree for a single branch it might be like buying a Leyland bus
>> to take my three kids to school...
>>
>> My two cents (...with 12% interest)
>>
>>
>> On 27 January 2012 09:18, robert yawe <robertyawe at yahoo.co.uk> wrote:
>>> Hi Listers,
>>>
>>> I came across a news item about how KREP Bank is to "invest" US 600,000/-
>>> (Kes 52,000,000 aka 52 Million) for a new banking software.
>>>
>>> The system, Temenos T24, is a monster of an application also being under
>>> used by Kenya Commercial Bank in addition the Kenya Bankers SACCO barely
>>> uses it in a single site.
>>>
>>> We keep bragging about how we are a technology savvy nation and
>>> developing
>>> a
>>> knowledge economy yet we are the most under-utilisers of technology.
>>>
>>> Which is a clear explanation why Vodafone cannot understand our obsession
>>> with wanting the MPesa servers installed in Kenya yet I am sure we
>>> currently
>>> utilise less than 1% of the capacity of the servers in Germany and 5% for
>>> 2
>>> days in a month.
>>>
>>> Temenos can handle over 3,000 transactions per second and handle over 25
>>> million accounts across 2,000 branches
>>>
>>>
>>>  http://sqlcat.com/sqlcat/b/technicalnotes/archive/2011/04/26/lessons-learned-from-benchmarking-a-tier-1-core-banking-isv-solution-temenos-t24.aspx.
>>>
>>> Temenos can handle the needs of all the banks in Kenya (approximately 6
>>> million unique accounts) and still not break a sweat while consuming less
>>> energy and resources than a mobile base station.
>>>
>>> Please let no one raise the issue of lost IT jobs because last time I
>>> checked a server was not a charcoal "jiko" that needs fanning.
>>>
>>> KREP has less than 200,000 accounts & 35
>>> branches http://www.k-repbank.com/branches.html, with Temenos they can
>>> run
>>> a
>>> transaction for each one of their clients every 58 seconds so what does
>>> the
>>> application do the rest of the time?  Kenya Bankers SACCO is an even more
>>> telling story with 1 branch and less than 20,000 members also uses
>>> Temenos
>>> that is internally installed.
>>>
>>> If Equity was to use the same software it will take 21 minutes to run a
>>> transaction for each of its 4.3 million clients so why do we complain
>>> when
>>> MPesa takes 5 minutes (over 14 million accounts) to handle a transaction
>>> on
>>> the 15th and 30th of the month?
>>>
>>> I am sure many of you in the banking sector are wreathing in anger
>>> because
>>> I
>>> am refusing to acknowledge that the data needs to be kept secure that is
>>> why
>>> you must look the servers in your individual server rooms.  So why doesnt
>>> the same arise when you open up your applications to KenSwitch for shared
>>> ATMs, to the mobile providers for mobile banking and to the ISPs for
>>> online
>>> banking.a
>>>
>>> Here is my solution, Safaricom (I am using them as an example as they are
>>> less likely to sue me) buys off all the licenses of Temenos that all
>>> those
>>> small banks, DTMFIs, MFIs and SACCOs have bought and then install Temenos
>>> in
>>> their 2 billion cloud.
>>>
>>> Buy all those Sun and HP servers that have been running those core
>>> banking
>>> systems and donate them to universities around the country, they can
>>> recover
>>> the costs by writing them off next year as obsolete therefore passing on
>>> the
>>> cost to the taxman.
>>>
>>> Then they can offer a complete package to the banks for using the cloud
>>> services without a nimbus and stratus pricing structure.  Safaricom,
>>> being
>>> experts at per second billing, can offer a similar solution to the
>>> financial
>>> institutions for use of the system.  To the financial institutions the
>>> use
>>> of the cloud will be treated as an expense therefore reducing their tax
>>> obligations.
>>>
>>> What about the staff you say, the only ones who are likely to loose are
>>> service providers like myself who have actually become over glorified
>>> cleaners more because we charge more than really the service we render.
>>>
>>> We have all been throwing profanities at the Walter of the Lazy African
>>> article fame yet we continue to prove him right every single day, yes,
>>> you,
>>> him, her and I.
>>>
>>> Have a lazy day.
>>>
>>> Regards
>>>
>>> Robert Yawe KAY System Technologies Ltd Phoenix House, 6th Floor P O Box
>>> 55806 Nairobi, 00200 Kenya Tel: +254722511225, +254202010696
>>>
>>> _______________________________________________
>>> kictanet mailing list
>>> kictanet at lists.kictanet.or.ke
>>> http://lists.kictanet.or.ke/mailman/listinfo/kictanet
>>>
>>> Unsubscribe or change your options at
>>>
>>>
>>> http://lists.kictanet.or.ke/mailman/options/kictanet/francis.hook%40gmail.com
>>>
>>> The Kenya ICT Action Network (KICTANet) is a multi-stakeholder platform
>>> for
>>> people and institutions interested and involved in ICT policy and
>>> regulation. The network aims to act as a catalyst for reform in the ICT
>>> sector in support of the national aim of ICT enabled growth and
>>> development.
>>>
>>> KICTANetiquette : Adhere to the same standards of acceptable behaviors
>>> online that you follow in real life: respect people's times and
>>> bandwidth,
>>> share knowledge, don't flame or abuse or personalize, respect privacy, do
>>> not spam, do not market your wares or qualifications.
>>
>>
>>
>> --
>> Francis Hook
>> +254 733 504561
>>
>>
>
>
>
> --
> Francis Hook
> +254 733 504561
>
>
>
> _______________________________________________
> kictanet mailing list
> kictanet at lists.kictanet.or.ke
> http://lists.kictanet.or.ke/mailman/listinfo/kictanet
>
> Unsubscribe or change your options at
> http://lists.kictanet.or.ke/mailman/options/kictanet/oloo.khisa%40gmail.com
>
>
> The Kenya ICT Action Network (KICTANet) is a multi-stakeholder platform for
> people and institutions interested and involved in ICT policy and
> regulation. The network aims to act as a catalyst for reform in the ICT
> sector in support of the national aim of ICT enabled growth and development.
>
> KICTANetiquette : Adhere to the same standards of acceptable behaviors
> online that you follow in real life: respect people's times and bandwidth,
> share knowledge, don't flame or abuse or personalize, respect privacy, do
> not spam, do not market your wares or qualifications.
>
>
>
>
> --
> "One hour per day of study will put you at the top of your field within
> three years. Within five years you'll be a national authority. In seven
> years, you can be one of the best people in the world at what you do." Earl
> Nightingale1921-1989
>
> Oloo Khisa
> P.O. Box 24324-00100
> Nairobi
> 0721321086/0734321086
> http://ke.linkedin.com/in/olookhisa
>
>
>
>
> _______________________________________________
> kictanet mailing list
> kictanet at lists.kictanet.or.ke
> http://lists.kictanet.or.ke/mailman/listinfo/kictanet
>
> Unsubscribe or change your options at
> http://lists.kictanet.or.ke/mailman/options/kictanet/francis.hook%40gmail.com
>
> The Kenya ICT Action Network (KICTANet) is a multi-stakeholder platform for
> people and institutions interested and involved in ICT policy and
> regulation. The network aims to act as a catalyst for reform in the ICT
> sector in support of the national aim of ICT enabled growth and development.
>
> KICTANetiquette : Adhere to the same standards of acceptable behaviors
> online that you follow in real life: respect people's times and bandwidth,
> share knowledge, don't flame or abuse or personalize, respect privacy, do
> not spam, do not market your wares or qualifications.



-- 
Francis Hook
+254 733 504561




More information about the KICTANet mailing list