Disruption is a funny thing


Its not just the core construct that gets disrupted. If we take SaaS as our example. I’ve blogged before about who else besides SaaS ISV’s will disrupt on-premises ISV’s. But in an interesting way the auxiliary services are also disrupted. I’ll deliberately use this example from IDC because of its relevance and irony. IDC state that integrators are going to be affected by SaaS, that is disrupted.


Here's where the irony kicks in. Ben Kepes blogged about the absurdity of this piece of news by IDC. And what has struck me is that the research houses themselves arc being disrupted.

I remember dealing with IDC here about 2 yrs ago. At that time IDC didn’t even have a taxonomy that included SaaS. They were calling it hosted application management or HAM. If you’ve ever worked with a research house you will know that if something doesn’t fit the taxonomy it creates a major panic. I think that world is gone, I think we are entering an era whereby online citizen journalism is going to usurp research or even the news. The analysis and insights provided by the likes of Bob at Smoothspan , Will Price and all the others out there are providing more timely, more accurate information than these research houses. I personally rely more on the blogging community than the annual, out of date documents put out by these places now. The only thing lacking (and I mean no disrespect) is the credibility that these houses have, it is very hard to know if numbers put up by Bob or Will are accurate. Unless these companies embrace web 2.0 and revise their methodologies they are in for a rough ride in the future. The speed of change is too dramatic, the pulse of the market is out here on the net not in spreadsheet.

 

Continue reading

More on Customisation

Like I said in my last post, customisation is a fascinating subject when it comes to software. A couple of regular readers have both held forth on the subject (I hope more will follow otherwise this blog suffers from the same issues as Techmeme!! I know, ironic link). Ben Kepes appreciates the concept, recognises the value for non-competitive applications being consistent but holds to the view of customer centricity being infinity valuable. Ben does make the statement that.

doesn’t it start us down a slippery slope to mass-homogenization where there really is nothing different between one enterprise and another”

 Bob at Smoothspan thinks that this isn’t a mutually exclusive scenario. Some things can be commoditised and some things can't. Bob also states that taking things vanilla breaks the rule of IT adding value by not introducing change.

 
Here’s my two cents on it. According to
Michael Porter, there are basically 3 ways companies compete:

  1. They can build a differentiated “best” product in the market.
  2. They can serve a niche better than the “best” product leader.
  3. They can be the low cost provider.

If what I say is true, that a business can take code vanilla (BTW I actually agree with Bob this isn’t a universal truth) they will accrue savings. They will save on code, customisation and implementation. They can choose the low cost provider (by now you’ll see where this is going) they can in fact BE a low cost provider or use the EBITDA savings to invest in other things. These other things could be differentiation of their own product or building a niche channel.

Ben’s version of mash-ups plays to either a (1) differentiated best product but that would be quite expensive to do on a mass scale so it’s more likely to be (2) a niche play. Where’s Bob's example is more likely to be (1) mass differentiation. Simple customisation on a mass scale (sounds a lot like salesforce.com).

So where does this leave us on the debate? I think for consumers and producers of SaaS alike the answer is know your business and your business model, in particular know thy customer. When you do, you can pick any of the above plays to suite your scenario

 

 

Customise the code or change your business?

 

There have been quite a few blogs ( Smoothspan, SaaSWeek and SaaSblogs comments ) over the weekend about just this topic and to me its fascinating topic. It seems to me that it shapes SaaS’s future. Let me explain.

Its my belief  (not experience) that one of the major drawbacks of on-premise software, especially in the ERP  / CRM market, is that the time to install massive. When I was dealing with SAP in the UK it was close to 2 years for many customers. And the reason for that is that ERP deals were more an exercise in business process modelling than software. That is, the customers would take the software and then force it to be tailored to their specific business after undertaking a large and expensive consulting exercise. This of course is closely followed by an ever costlier customisation and implementation exercise. The end result is that the customer usually got something fairly closely resembling what they wanted (sic) and effectively had THEIR version of the code, which of course THEY had to support and upgrade and basically…they could never leave because of the massive investments (both in money and customisation).

So we move to SaaS, in this instance the customisation can happen but all of the customisation tends to happen on the front end and the complexity is dealt with by the provider. They have to build a system that allows for customistation per customer while still operating on a single instance of the code. But by doing customisation, businesses do delay the time to benefit in a SaaS world. For example, we’re 9 months into a Salesforce.com rollout here!

So my question is, why don’t business’s just take the code vanilla? If you did a TCO on an on-prem software installation and compared the cost of customisation and lifetime cost of ownership with the cost of changing the way you do business to suite the code what would you find? If you took the SaaS vanilla wouldn’t that speed up your time to benefit? Wouldn’t it make the SaaS providers job easier and therefore the code service should be cheaper???

I know every business believes their unique, its a universal mantra, At a recent spat of focus groups I must have heard this 20 times …“my business is different”, “that might work for them but we do things different”. I think its wrong. And believe me, at the end of one very insightful group, the businesses there got that too….

So my question is….If you took you’re on-prem code vanilla, wouldn’t that attack one of the core value propositions of SaaS – speed to market and easy upgrades? I could finance all the bits and give you a monthly rental. I’m being provocative but am genuinely interested in the answer… please if you’re a lurker give me you’re comments.

In a SaaS world, Its ALL about the platform

SaaSweek just posted with provocative title, is PaaS a passing Fad?  The answer is hell no.

 

If you buy into the rudimentary definition of SaaS that Ben Kepes, Smoothspan and I got to recently, then the whole SaaS ecosystem will be dependant on platforms. And logically, it doesn’t make economic sense for every SaaS vendor to create their own.  (I know this gets into shaky ground because I still don’t believe there is a good understanding or even consensus of what a platform is.)

If every SaaS vendor were to build their own platform, then literally there is a cost transference from the on premise business to the SaaS providers and the only people who will win out are the hardware and storage vendors. That just doesn’t make economic sense. There is already a bunch of data (here and here) showing that its harder to get to profit for a SaaS business. Why would you want to lump the expensive costs of setting up your own datacentre, operational framework, billing, backup, site diversity etc when you can just buy that?

Would a SaaS company build another “internet” to connect to its customers? Again, no. SaaS companies use the internet as its network because of sound economic reasons (name another globally ubiquitous network, with understood standards – loose as they are – that’s cheap?).  The infrastructure, diversity cost are provided by other platform providers (ISP’s) and they can do this through economies of scale.Platform providers, those doings PaaS, offer the potential for the same basic economic advantages.

SaaSweek raises a couple of valid points about why PaaS isn’t globally accepted.I’d counter those with the following comments.

  • Apart from Force.com, none are approaching scale so the benefits aren’t really there. But they will be.
  • Expertise, its really hard to build a good SaaS application, let alone build a good datacentre, operational framework, back up and restore environment, internet connectivity.
  • Speed to market. Given that SaaS is sold on a subscription basis, it is subject to the rule of 78. the sooner you hit the market the soon you’re revenue starts compounding. Also the faster you develop the less costs and that’s a good thing too.
  • PaaS brings its own community and the associated advertising, social networking and open source development mantra. This helps you with speed of development, resources and of course customers.
  • Funding, what is to stop a PaaS provider supporting a great idea on its platform in a VC capacity?
  • PaaS is still in its infancy, the models, rules and mores of how to be a PaaS provider aren’t 100% formalised. Smoothspan talks about being Switzerland and questions whether Salesforce.com can pulls this off. I’d agree that you do need the neutrality. I’d also say watch this space. It’s my prediction that Salesforce.com will become wholly focused on PaaS, and be more than willing to let the CRM business fight its own fights. Why? Well CRM only solves part of a business’ needs. Force.com is solving the rest.  And this is what platforms are all about, more services from same infrastructure.

I’ve said before Telco’s do SaaS, they also do platforms really really well. Don’t believe me, well check out this. They’ve taken one core asset, or platform (copper) and sold at least 3 core services down it. Voice, access / rental and now data. They are doing the same with mobile networks. Voice, SMS, data, and content.  They’ve also sorted out sales, provisioning, billing and support (to a greater or lesser degree I).

These traits are the mark of a great platform provider. Telecommunications is a $1.2 trillion dollar business built on platforms. If you want to know what the future looks like, do your history.

Microsoft, Platforms, Opensource and SaaS

Ok, I'm playing the meta information game, apologies its kind of busy here at the moment.

Following on from my prediction about Microsoft opensourcing its code base and the implications on SaaS. I read with interest these two blogs.

The first, the war in the clouds, discusses how Force.com and Microsoft are going head to head on their platform plays. The second by Mary Jo Foley  goes into much more detail about opensource, platforms and the different plays. They are both in general agreement with my post. MS is doing what its always done, getting a bunch of software developers to use its code and build customer centric services onto the core code, and in this instance into the Live platform.

An interesting time ahead

 

Ok, Appliances aren’t SaaS, they are managed appliances – duh!

I stand corrected. In the face of a good deal of logic from “the other Ben”, Bob Warfield and Ben Kepes I am going to amend my position.  Using the age old wisdom that things are what they usually appear to be, then managed appliances are just that, managed appliances and SaaS is completely different.

I do think that globally we need to agree to some fundamental traits for SaaS to be SaaS,

 

  • The application sits in some sort of cloud – normally the internet, but can be a private cloud
  • It is multi-tenanted
  • It requires no / minimal onsite configuration
  • Bills on some sort of user pays or subscription model.

 

If this is the definition, then we can rule out appliances as SaaS. The bits that the appliance model misses from my view of best of breed SaaS plays are some sort of development platform, the reach associated with the cloud and the full blown centralised management that SaaS provides, and of course the hype.

 
I do hold to the view appliance based models are viable business models. As Bob suggests they are a step of from Legacy Software Vendors (LSV’s), and that’s got to be valuable. They can in fact deliver some of the benefits of SaaS, which is why I held my previous view.

The appliance can be a staging environment for the application allowing for centralised updates to occur (that is remove multiple versions). Appliances allow for a subscription based model, especially if they lease the kit out to. The support can also be done centrally, that is you are buying the outcome.

 

 

SaaS, Clouds and other ways of achieving the same outcome

 

Over on Ben Kepes blog, he asked the question Does SaaS have to be in the clouds???. To me the answer is no.

I personally think that his example:

He proceeded to tell me that one of their products constitutes SaaS because, although it is installed on server within the organisation, from th users perspective it has all the appearances of SaaS;

  • It is updated centrally by someone else
  • They just use it and don’t need to think about updates, sysadmin etc”

Is just another way of delivering SaaS aka the appliance based method.  SaaS being tied to the cloud is just a commonality of the service, not a pre-requisite. To me this  commonality exists because the internet (the cloud) is a economical way of delivering services. Its got large reach, is redundant, low cost (compared to private networking) and is largely global. To that end it’s a great way of providing the service. But there are others

  • The private network variant – a good example is managed security providers like Safecom or counterpane. They literally deliver “firewall as a service and content filtering as a service” just over a private connection
  • The managed service variant – Kaseya is a good example. They use software but sell a remote monitoring and central updating service.
  • The appliance or staging environment variant – desktop antivirus vendors do this currently.  Makonetworks do this with a firewall appliance currently.
  • Damon Edwards at dev2ops claims that internal IT are SaaS providers too. In a theoretical sense I agree. Having a large outsourcer as our IT provider here the reality is a little different.

The one line I love in Damon’s post is this

“SaaS is about the relationship between the service provider and the service consumer. The physical characteristics of service delivery are almost inconsequential.”

This will get the SaaS purists a bit upset. And I guess to some extent that’s a good thing. With any new market you need your zealots, the people with the passion to make it reality.  (Who IS it that that decides what is SaaS and what isn’t anyway?). But from a customers perspective it’s almost nothing to do with the technology and almost all about the outcome they want and how the offer matches that.

 
The various benefits of SaaS appeal to customers in different ways and can be delivered in many different ways. It’s the outcome that’s important. That’s why I stick by my claim that crusty old Telco’s have been doing (Voice) Software as a Service for a very long time.

 

What does MS open sourcing .net have to do with SaaS?

I'm willing to bet that the announcement by Microsoft today that they would opensource their .net application went fairly unnoticed by the wider SaaS community. I think this is a grave mistake and a fundimental unestimation of Microsofts attempt to get into SaaS big time (in their own way according to Phil Wainwright)

Marshall Kirkpatrick on R/WW states that " It's hard to say what the incentive was for this move" and openly admits that they at R/WW are "still chewing on the significance of this announcement"

Well here's my pick. It links nicely with my bit about opensource and SaaS from the other day. Stay with me because this is a little convoluted.

.Net is a critical component of Microsoft's web services play. One of the key ways that Microsoft has commercialised it and is using it in a SaaS way is through its Connected Services Framework .   Here's where SaaS companies should sit up and take notice.
 
CSF is being sold to Telco's as the underlying framework to connect communications AND SERVICEs together. Its a platform, a webservice platform to make a Telco's existing stuff better and to run new services.  The new services they allude to are SaaS services.  (Told you this would get interesting).
 
The other thing (and this is a guess), but i'm willing to bet that Live.come is built on CSF/.Net 
 
To my mind this puts .Net up there as a clear contender to other application development platforms like Force.com  . Only better, better for a tonnes of reasons.
  • There's a massive developer community already skilled in building .Net stuff.
  • By openning it up to the world, Microsoft has essentially given its developer partner community an easy leg into the SaaS world. Something they clearly wanted after they announced thier S+S play.
  • Its also given its chances of developing a lot of different applications that appeal to the masses a huge boost. (remember when i talked about Scale and Microsoft ?)
  • Its also given its Telco partners a way of getting access to that same developer community / ISV's. Given they too have scale, know about platforms and are seeing SaaS as a blue ocean, thats incredibly valuable.

 

Watch this space, i think you'll see a whole bunch of SaaS action out of it

Is Larry right to stay On-demand focused?

I think Larry Ellison knows what he’s about.

 

I must admit, I’ve been thinking a bit about the commentary about oracle not getting into SaaS because there's more money to be made in on-premise. Because of this  I re-read Phil Waineright’s post about how Oracle was spurning SaaS for ERP easy pickings.

The thing I noticed second time around was that the quote wasn’t from Larry, it was in fact from the Information weeks reporter and was her interpretation of what Larry said.  So I went searching for another article and found this one which presents a channel centric, less sensationalist view of what Larry actually said.

The interesting thing is that the high points Larry mentions, are almost exactly the same as Steve Singh in his really insightful interview with Bob at Smoothspan.

 

“ Steve:  This is just one guy’s prediction.  I don’t believe large companies can make the conversion.  Forget their genetic code.  How many will take the pain?  Companies won’t reinvent themselves. 

Think of taking a $40B company to On-demand.  The value of the business will go through huge negative change.  It will get crushed.  Cash flow will get crushed.  You have to layoff.  The transition is really hard and its very sudden.

If you’re north of $100M its hard.  Over $1B its impossible.”

Overlay that with Larry’s points about building a new channel, marketing to new customers, and the cost of building an entirely new code set and you could call Larry’s statement that it would be “enormously challenging” the understatement of the year!  You could as a shareholder also say that it was good decision and correct corporate governance. 

The two largest software companies to try and do this are of course SAP and Microsoft. From what I can gather from press coverage , they haven’t exactly been successful todate.

So then (bearing in mind I am a SaaS advocate), why then is everyone bagging Larry?