On-Premise Installation Vs. Hosted Solutions.

Every so often we (still) get some sales inquiries about from potential customers who want to install a survey tool/solution on their own servers. Back in 2003 we used to get a lot more inquiries like these. In fact back in the day, users were even developing their own survey engine from scratch. Today we get about one or two a week.

I’ve been in the IT/Software business for some time now. Even before starting QuestionPro, I was a technology consultant for a few startups mostly in SF and Seattle during the dot-com boom and then did some work for larger shops like Washingon Mutual during the dot-com bust here in Seattle 😉 So, I do have a good grasp on what the actual costs of a in-house application development/installation/maintenance is.

For the most part (thanks to Google, Salesforce, GotoMeeting, etc.) most people really have understood the value and real cost differential between the TCO (total cost of ownership) between hosted solutions (On-Demand/ASP) and on-premise installed software. For example, see this blog post by by Peter Yared about using only hosted solutions for his new company :

Quote from the Peter’s Blog:

Back in 2003, we still had to buy our own servers and hire IT people to
get some basic services. This mindset prevailed into 2007, where to
upgrade our Wiki into something much more functional, I learned that IT
had physically installed two iterations of wikis on our servers. It was
beyond me what features we would get from our own installation vs. a
hosted one, so I suggested a policy of only using hosted infrastructure
moving forwards. This led to a suggestion of having some meetings to
discuss the concept, which in my experience means “not going to
happen.” 🙂

Starting with a clean slate at wdgtbldr, there is
definitely an “everything must be hosted” policy, and I am amazed at
how cheaply and easily all of the functions of a small business can be
set up and shared between employees. There are no servers, no VPN to
get to the servers, no software to install, configure, and maintain,
and definitely no part time IT people. Everything works as advertised,
since it is not our installation of Bugzilla/wiki/etc. that has to be
maintained, rather proven infrastructure shared by many other companies.


We also understand your concerns about going with a hosted model for your survey software solution. They typically revolve around data ownership, portability and flexibility. What you also need to consider is some of the issues that are now obvious and apparent like server OS patching (especially if you are in Windows), monitoring and redundancy.

Here are a few points to consider when making a build vs. buy decision on survey software that we’ve collated over the years. These apply not just to survey software, but to any SAS (Software as a Service) purchase decision:

  • Core Value Add — Support and maintenance of software cannot be a core value add nor should it be for a market research firm or a company who’s core business is not collecting data. All the market leaders in the industry know this and have adopted the simple model of keeping only their core business in house.
  • Staying Competitive —  If you decide to maintain the software in house, you will put yourself behind your competitors, as your company will incur an additional cost/risk associated with maintenance of software. Software is one of those things where the maintenance to capital ratio is very high. For example, if you compare this to the Auto business, you buy a car for about 30K, and you spend about 3K/Year (gas, oil changes, break pads etc.) to keep the car running. If you buy a software (on-premise install) for about 30K, it’ll cost you about 10K to keep that running. 30% maintenance fees is very typical and minimum. In fact in many cases the operational costs actually surpass the capital expenditure via. costs like server installation fees, product upgrades, bandwidth provisioning etc.
  • Quicker ROI — From an initial ROI perspective, hosted solutions shorten time-to-value by eliminating software implementation and cost issues. You focus on collecting your data, refining your processes, and defining your business goals. Not worrying about how Microsoft Vista SP1 patch will affect your installation, or how the next virus threat may affect your connectivity.
  • Total-cost-of-ownership. Users of hosted software solutions need not ever implement software upgrades, pay for maintenance, or add hardware. The net effect is to keep total cost of ownership in check. More along the lines of “hidden” costs. If you spend $49/Month on Salesforce or $15/Month on QuestionPro you have a clear understanding on what the total cost is. In on-premise installations, the total costs can never be nailed down. Server hard drive failure can cause outages that may involve one of your engineers to go to the data-center to replace a failed hard drive. These costs are ad-hoc and can only be estimated at best.
  • Security Team — Most SaaS vendors have infrastructure devoted entirely to staying up to date with the latest software/hardware security issues. Its our job to keep your data secure – if it wasn’t we wouldn’t be in business. We can do this efficiently because our upgrades/changes affect all our customers — we are essentially distributing the costs across all our customers — Economies of scale.
  • Data Portability — This is obviously a key issue. What if the company that you rely on for your core business stops servicing you, or goes out of business. In almost all SaaS solutions, there is usually ways of getting your data out of the system (locally) as a backup measure. Companies like Salesforce, QuestionPro, Netsuite, ConstantContact, etc. all offer up the ability to download data in a standard Excel/CSV /XML format. These output models serve as a good backup measure just in case things go south. There are however issues around system portability. For example, data from Salesforce cannot be simply taken out and put into Netsuite, or data from ConstantContact cannot be ported to VerticalResponse. This would involve setting data structure standards and everyone adhering to them — Not realistic in the near future.

Now, in many cases a SaaS solution may not be right for some companies. Typically we used  to think that larger (Fortune 500) companies go with in-house/on-premise installations simply because they could afford it. With Web 2.0, what we are now seeing is that even for larger companies, it’s not a question of cost, but time-to-market is becoming a bigger factor. This is the driving force behind the Enterprise 2.0 model. For example, with a hosted solution you can get whatever you want done in a few days. A few days for an internal IT team to put a solution in place is simply not possible. It takes about 2/3 months just to order a set of servers and have it installed in the data-center, configured and patched. We know about this first hand — We have to do this too!

I know there are hard circumstances and directives (“thou shall not use hosted software”) that many companies have, and most likely there are some reasons why such policies were put in place. What I would question is _when_ was this policy put in place — In 2000? or 2008? Many times we get into conversations, especially with companies in Govt., Healthcare and Banking/Financial — which are heavily regulated industries about compliance and issues surrounding privacy, data protection and risk mitigation etc. They are all valid issues and concerns, but lets analyze them. If the issues are not technical, then most the compliance issues are legal directives. For example HIPPA (HealthCare) certification does have anything to do with real technology. Many legal and complianc
e issues can be simply handled with a customized SLA (Service Level Agreement)  — An SLA that is specific and addresses concerns surround data storage, uptime, privacy and confidentiality.

The fact is SLA’s is much easier to negotiate than hosting and running your own servers and services. SLA’s are a one-time cost.We’ve had custom SLA’s and agreements with many companies in regulated industries like heathcare and banking. Custom SLA’s provide an easier mechanism for enterprises to have hosting agreements that capture all the core needs of uptime, availability and regulatory compliance while maintaining the flexibility and cost savings on a hosted platform.

It is also not uncommon for companies to conduct stringent (both technical and financial) due-diligence on service providers. We’ve gone through these a few times now and almost expect this from large companies now. I recently had a chat with one of our clients — Tim O’Conner who is the Senior VP Marketing & Strategic Sourcing for Unisource (a 4 billion dollar company) — and he said in no unequivocal terms the benefits of a hosting solution — Not only in terms of cost savings (about 5-10x cheaper) but more importantly in terms of time-to-market. With a hosted solution (especially self-service) he is in total control of the execution parameters and can get solutions out of the door as and when he wants.

Comments are closed.