Summary: Questions asked of the Enterprise SaaS Working Group

October 1, 2009 Appirio

Ryan Nichols


Great kickoff today of the Enterprise SaaS Working Group, sponsored by Conformity. With Scott Bils (Conformity), Peter Coffee (Salesforce.com), Tom Fisher (SuccessFactors), Steve Coplan (451 Group), Doug Harr (Ingres), Scott Carruth (Initiate), and Michael Amend (Dell) as panelists, of course we ran out of time long before we were able to explore all of the questions asked of the panel.


So we thought we’d continue the conversation here– below is Appirio’s take on the issues raised in today’s webinar (if you missed it, feel free to watch here). Looking forward to your feedback!


What will it take for SaaS and cloud apps to become ‘mainstream’ in the enterprise?

We think about enterprise adoption of the cloud in 3 layers–-Software, platform, and infrastructure as a service. SaaS apps are already “mainstream” in many enterprises, as are pilots of custom applications built on cloud platforms like Force.com, and burst and test / development on cloud infrastructure like Amazon. That’s the first column in this diagram:

















Now, one common pattern of adoption we see is enterprises who do a point SaaS implementation, and see it come in on time and under budget, with off the charts usage. This causes the CIO to ask “what else can I do with this platform?” This results in the migration of more and more custom applications– that’s a big part of the activity we see today. That’s the second column.


Where is this heading? Well, our perspective is that over the next 5 years you’ll see even the largest enterprises running the majority of their IT infrastructure in the cloud. That’s how we run our business at Appirio– no servers– and we want to help other companies do the same thing. We call it cloudsourcing– where cloud computing meets outsourcing.


What’s holding that back? We highlight 3 things that are missing. The first is a strategy and roadmap– a business-case-driven answer to “what do I do next”. The second are solutions to connect and extend cloud platforms– making the cloud easier to consume. Third is a new type of partner for enterprise IT. Traditional partners have too much to lose in the transition, and their core skill set is irrelevant: it doesn’t matter how good you are at managing or working with a server farm anymore.


PaaS vs SaaS: What will be the preferred model in the future for how cloud apps will be delivered/consumed in the enterprise?

PaaS is as disruptive to standalone SaaS as SaaS has been to traditional on-premise software. For example, Appirio has built a solution for professional service firms on Force.com that is winning customers from PSA vendors that have been building out their own SaaS stack for 10 years. It’s not that we’re smarter– its just that 100% of our R&D is targeted to the specific needs of our customers. Force.com’s R&D team is doing the mobility, the reporting, the analytics, etc. Read more at “Now that’s a big PaaS market”

SaaS industry evolution: Do you anticipate long term cloud vendor consolidation or the realization of the ‘best of breed’ vision for business apps?
As an integrator in the space, we see a proliferation of packaged and custom applications built on only a handful of cloud platforms. It’s hard work to build out a standalone SaaS stack We think you’re going to see less and less of that in the marketplace. Read more at “NetSuite Buys Quickarrow”
Private / hybrid clouds: Are they part of the long term answer in the enterprise, or just indicative of an early market? Would the model benefit SaaS or legacy on-prem vendors?
Private cloud is just a data center with a fancy name. Of course anyone with a data center should be thinking about optimization, virtualization, etc. But NOT at the expense of thinking about the public cloud. The real question is how you can get out of the business of running a data center. Read more at “Rise and Fall of the Private Cloud”
SaaS TCO in the enterprise: Is it compelling enough to justify ‘rip and replace’ of existing on-prem apps?
Yes, rip and replace can absolutely make sense, but not without a business reason to switch. The TCO benefits are necessary but not sufficient. One thing to keep in mind is that the savings are significant, but “chunky”– for example, you can’t shut down a server until everything has been moved off. You need to have a roadmap to actually capture these savings. Read more at “Cloud Computing Savings: Real or Imaginary?”
Enterprise security / control requirements: Are cloud vendors doing enough? Will enterprises be willing to meet vendors ‘halfway’?
Enterprises sometime ask things of their cloud vendors that they’d never ask of their on-premise vendors. You’d never get a hard-drive manufacturer to compensate you for the deals you lost if the drive crashed… but we’ve heard CIO’s ask exactly that of Amazon S3. Read more at “Cloud Insurance?”
Application penetration: Are there any application areas that SaaS will not be able to penetrate? If not, why haven’t they penetrated yet?
We run our own business 100% in the cloud– we’re a serverless enterprise. We’re only 150 people, but could we do the same thing at 10X our size ? Absolutely– we’re talking to customers about helping them make exactly that switch. Could we do so at 100X or 1000X our size, or if we were in an industry with factories and warehouses to manage? Probably not yet, but it’s not a technology problem, it’s a market maturity problem. Read more at “Cloudsourcing: Where Cloud Computing meets Outsourcing.”
Are SaaS/cloud vendor outages overblown, or are they symptomatic of problems that may be barriers to enterprise adoption?
Totally overblown, but that’s a good thing in the long run. Anyone who’s worked at a big company using exchange knows how often email is unavailable for one reason or another, sometimes scheduled, sometimes not. But nobody notices or even measures the impact. When gmail is unavailable, on the other hand, it’ a big deal, it makes headlines, and Google fixes it. That’s part of the reason why SaaS solutions will become better and better suited to enterprise adoption over time.
Are high-transaction-volume applications appropriate for the cloud? If so, can you please provide some examples that have transitioned to the cloud successfully?
The cloud is absolutely ready for high-transaction-volume applications, if architected carefully. Here are some examples:
  • We’ve helped Japan Post manage 150M customer relationships using Salesforce.com
  • We’ve helped a public agency in Japan using Force.com and Google AppEngine together to send a hundred million workflow-generated emails a day
  • We’ve helped Starbucks Pledge 5 use Force.com Sites to match a half million volunteer opportunities on Obama’s inauguration day
  • We’ve helped a 5000 person consulting team crunch their timecards in realtime to forecast revenue and utilization by region, practice and group
Summary
Of course, there were too many great comments throughout the session to summarize here– definitely worth watching the full replay here. But we wanted to close with a paraphrase from Peter Coffee in today’s webinar:
There are 2 words that are holding back the adoption of cloud computing in the enterprise: “CAN’T” and “OR.” The first is an assumption that something can’t be done on the cloud, the second is an assumption that you need to choose between cloud solutions and on-premise solutions. Both are false.

Previous Article
Larry Ellison takes on the Cloud – Entertaining but far from Accurate

Balakrishna Narasimhan The blogosphere is abuzz with the news of Larry Ellison’s hilarious evisceration of ...

Next Article
Enterprise SaaS Working Group: The cloud is ready for your enterprise, is your enterprise ready for the cloud?
Enterprise SaaS Working Group: The cloud is ready for your enterprise, is your enterprise ready for the cloud?

Ryan Nichols We’re honored to be participating in this week’s kickoff of the Enteprise SaaS Working Group, ...