SaaS Startups: Knobs and Dials And Other Insights

About This Blog

This site is for  entrepreneurs.  A full RSS feed to the articles is available.  Please subscribe so we know you're out there.  If you need more convincing, learn more about the site.



And, you can find me on Google+

Connect on Twitter

Get Articles By Email

Your email:


Blog Navigator

Navigate By : 
[Article Index]

Questions about startups?

If you have questions about startups, you can find me and a bunch of other startup fanatics on the free Q&A website:

Subscribe to Updates


30,000+ subscribers can't all be wrong.  Subscribe to the RSS feed.

Follow me on LinkedIn


Current Articles | RSS Feed RSS Feed

SaaS Startups: Knobs and Dials And Other Insights


If you're building a SaaS (Software as a Service) startup you're in great company.  Most software entrepreneurs today are taking this approach -- including me.  Below are some simple insights that I've learned in the process of being in the trenches with my own startup. onstartups knobs

Insights On SaaS Startups

1.  You are financing your customers.  Most SaaS businesses are subscription-based (there’s usually no big upfront payment).  As a result, sales and marketing costs are front-loaded, but revenue comes in over time.  This can create cash-flow issues.  The higher your sales growth, the larger the gap in cash-flows.  This is why fast-growing SaaS companies often raise large amounts of capital.  My marketing software company is an example.

2.  You’ve got operating costs.  In the shrinkwrapped software business, you shipped disks/CDs/DVDs (or made the software available to download).  There were very few infrastructure costs.  To deliver software as a service, you need to invest in infrastructure — including people to keep things running.  Services like Amazon’s EC2 help a lot (in terms of having flexible scalability), but it still doesn’t obviate the need to have people that will manage the infrastructure.  For a startup, the people cost to manage the IT stuff can be significant (since the team is very small).  So, even though hardware and infrastrucutre are cheap, managing it can take a significant percentage of the startup’s time.

3.  It Pays To Know Your Funnel:  One of the central drivers in the business will be understanding the shape of your marketing/sales funnel.  What channels are driving prospects into your funnel?  What’s the conversion rate of random web visitors to trial?  Trial to purchase?  Purchase to delighted customer?  As a SaaS startup grows, a lot of leverage can be found by understanding the shape of the funnel and removing the “leaks” (i.e. where are you losing business)?  For example, if a lot of people are signing up for the trial, but very few convert to paying customers, you should dig into what the early usage pattern is.  Are people logging on at all?  If so, where are they getting stuck?  Remove the friction that is keeping customers from getting value and you’ll unlock some revenue.  Do this at all stages of the funnel (focusing on the easy stuff first).

4.  Install Knobs and Dials In The Business:  One of the great things about the SaaS business is you have lots of aspects of the business you can tweak (examples include pricing, packaging/features and trial duration).  It’s often tempting to tweak and optimize the business too early.  In the early days, the key is to install the knobs and dials and build gauges to measure as much as you can (without driving yourself crazy).  Get really good at efficient experimentation (i.e. I can turn this knob and see it have this effect).  As with most experiments, don’t change too much at the same time (even though you think several different things will all have positive effects).  The reason is simple:  If you change more than one thing, you won’t really know what really happened.  Unless you have lots of data points, simple tests are usually better.

5.  Value the Visibility:  One of the big benefits of SaaS businesses is that they often operate on a shorter feedback cycle.  You’re dealing in days/weeks/months not in quarters/years/lifetimes.  What this means is that when bad things start to happen (as many experienced during the start of the current economic downturn), you’ll notice it sooner.  This is a very good thing.  It’s like driving a fast car.  Good breaks allow you to go faster (because you know you can slow down if conditions require).  But, great visibility helps too — you can better see what’s happening around you, and what’s coming.  The net result is that the risk of going faster is mitigated.

 Running a SaaS startup is a lot of fun.  There are so many more things under your control than the traditional shrink-wrapped business.  Use this to your advantage.  Keep your feedback cycles short, maniacally track the data and invest in continual (but cheap) experimentation.  In the long term, these things will give you a huge advantage.

What have you learned while building your SaaS startup?

Posted by Scott Stephenson on Fri, Sep 11, 2009


We've learned that we are going to likely include SaaS in our future offerings. We do application development but there are some of our solutions that can be delivered via SaaS.  
See you at BoS in SanFran.

posted on Friday, September 11, 2009 at 11:28 AM by Shawn Anderson

Our SaaS is a time tracking and billing sofware. I've learned that when it comes to looking at monthly revenue, don't forget to include the stream of revenue generated by the months prior. It requires a few more steps in the number crunching, but well worth it.

posted on Friday, September 11, 2009 at 11:30 AM by Jessica Eballar

Excellent piece. Thought you might also find the output of this peer learing workshop we ran to share ideas on shortening SaaS sales cycles. Very useful tip from Salesforce incuded.

posted on Friday, September 11, 2009 at 11:39 AM by Mark Littlewood

Also know your limits. It easy to get carry away with set and leave. You need to know at what stages you will need to invest significantly more in the infrastructure and at what stage you will need to investigate.

posted on Friday, September 11, 2009 at 12:02 PM by Scopulus

Just started development on a SaaS model for e-learning. Any suggestions on capital providers?

posted on Friday, September 11, 2009 at 12:37 PM by Vin

The most important thing we have learned is to keep development cycles short as it helps you quickly respond to changing customer demands.

posted on Friday, September 11, 2009 at 12:38 PM by Shad

Not to bash you but this is nothing new SAAS. 
If I where doing a IT startup I'd probably go the way of Cloud computing smae thing sort of. 
no offense though

posted on Friday, September 11, 2009 at 12:43 PM by Dan

Thanks for sharing. Gung-Ho is in the "shrinkwrapped software" manufacturing and fulfillment business. We have found that many companies that start with SaaS, ESD downloads, and shareware later want to add physical manufacturing and fulfillment (even if it's just for backup CDS and manuals). If you have those requirements, feel free to contact me - my direct number is 775 831 3750.

posted on Friday, September 11, 2009 at 12:43 PM by John L Wade

We just launched our product and have been playing around with our initial price structure in order to get customers into the system. Does anyone have any feedback on what strategy worked best for them (free trials w/credit card info, w/o credit card info, reduced pricing, rebates, etc.).

posted on Friday, September 11, 2009 at 12:47 PM by Jordan

Thanks for the great insight, Dharmesh. I'm working on a bootstrapped startup SaaS right now -- very bootstrapped. We do app development (we're also a Hubspot partner) and we're working on a service built 100% on the Amazon cloud - Amazon even bills our customers using the same pay-per-use model as S3 and EC2. For us, infrastructure costs are very low because the costs are built right into our product's pricing. We've been working on the project almost a year now. We would work on it whenever there was downtime from our clients' projects. I've learned this is a terrible way to stay on schedule and stick to goals, as our clients' demands are always changing.

posted on Friday, September 11, 2009 at 12:53 PM by Daniel Lynton

Gr8 Article. The company in which i m currently work is also a SaaS Start up in Rajasthan. 
They have also been working in launching their own online products which of which 

posted on Friday, September 11, 2009 at 12:54 PM by Jitendra Singh

My company has been deploying SaaS for over 10 years. If you remember, 10 years ago the Internet was not reliable and a whole lot of people didn't have email. 
A business that offers a SaaS product is not any different from any other business - in that, they all require "effective management" and customer focus. Reliability is a HUGE issue and you have to prepare your organization to deal with issues 24x7x365. 
SaaS is not a golden bullet, it is just another service - that can either be done well or poorly.

posted on Friday, September 11, 2009 at 12:59 PM by Mike G

Very interesting piece. We are developing several products that we may position as SaaS, but we need to be careful not to empower our clients and competitors too much. Our core asset is consumer and business data. You hit is right on the nose with the initial costs. They can be very high.

posted on Friday, September 11, 2009 at 1:04 PM by Ray

Could'nt agree more with your observations. To the first point I would add that a SaaS company assumes all the risks for the buyer who can now determine the actual value of the software in period one of the subscription. In case of shrink wrapped software the buyer assumes all risk of determining value after paying the full price. 
One lesson for me has been that enterprise software ( at least right now) sold as a service has the same long sales cycles as the shrink wrapped version despite the lower setup costs. 

posted on Friday, September 11, 2009 at 1:05 PM by Ranjit Nayak

Great article! Solid business pointers regardless if you're doing SaaS or not. Regarding point 1: not many business models exist w/o a good deal of front loading.

posted on Friday, September 11, 2009 at 1:07 PM by Steve Yakoban

What I've learned from the dot bomb was not to reinvent the wheel.. whatever you're doing has been done before. Thousands of firms building their own security model, their own search engine, their own personalization, etc. etc. Don't waste your time or money creating and perfecting something that can be leveraged from other sources, sometimes for free.

posted on Friday, September 11, 2009 at 1:08 PM by Richard Steere

I believe you're right on these points. We believed in the SaaS model as a the future of services so much, we built a Platform as a Service that allows companies to quickly translate their industry knowledge into a SaaS solution with little to no development. Soon, the quality of service, and the ability to provide effecient workflow within the service, will determine the winners and losers of the Saas field.  
Great acticle! Thank you for adding to the SaaS conversation.

posted on Friday, September 11, 2009 at 1:08 PM by William Hodges

Thank you Dharmesh, this gives me some more ideas to work my SaaS website out. 
Dan, maybe you built 10 successful SaaS before? Do you own Maybe some people are beginners and still need to hear about such things?

posted on Friday, September 11, 2009 at 1:36 PM by Alexis Wilke

Nice post, the stuff about understanding your sales funnel is the real gold in here. 
As developers and designers it's often easier for us to focus on the feature set and building it out. 
Understanding your customers and how they're finding you I find is more of a challenge than the technical aspects. 
Shape of your funnel  
Encouraging user feedback 
These are the two things that I'm focused on in building out the various flavors of our Hosted CRM & CMS application, XLsuite.

posted on Friday, September 11, 2009 at 1:57 PM by XLsuite

Good article - we've just hit our 3rd birthday as a SAAS company and the recurring revenues are starting to coe thru so its well worth sticking wih this model through the bad times!

posted on Friday, September 11, 2009 at 2:22 PM by Paul MElvaney

Hi Darmesh, 
Nice article!, One of the things that we get the benefit of being a SaaS enablement company is that we get to talk to a lot of folks that are building SaaS applications and learn all the different issues and details of their business and journeys as SaaS companies. A lot of times their reaching to us for architectural help and want to know if SaaSGrid can help them but many times they also engage with us because of our SaaS expertise and overall strategy guidance. You touch on several important points some of which we’ve written a lot about <a>here.  
Cheers, and keep up the excellent work 

posted on Friday, September 11, 2009 at 2:32 PM by Abe Sultan

Good insight as usual. I couldnt agree more with your points. Here are some of my insights 
1. Research your target market (in great detail) BEFORE developing your SaaS business model. Study the trends in the market (Google trends)  
2. Always try to stay lean and mean 
3. People make all the difference in the start-up and remember to change your leadership style to accommodate / incentivize

posted on Friday, September 11, 2009 at 2:35 PM by Justin R French

thank you for the great insights and the group for their helpful comments. 
Good analogy for the Value the Visibility section, but as a car enthusiast it would be good to have both breaks and brakes. ;-) 
Keep up the great writing.

posted on Friday, September 11, 2009 at 2:40 PM by R Nikora

Good post. We sell SaaS to local government and are seeing the rate of adoption increasing with the economic stimulus funds and also Obama's endorsements of cloud computing. 
Check out our blog post and comment.

posted on Friday, September 11, 2009 at 3:00 PM by Susan Kirk

Never under estimate the power of good design. I spent 5 years designing a distributed platform for building efficient vertical specific applications on top that has implicit global fine grained security, is scalable across nodes, has per object notification and life cycle management, can be scaled at run time without any down time to existing nodes, has dynamic resource allocation and throttling with regard to memory and database connections and can be extended with new functionality at run time.  
I also made sure to generalize access to file services and data (db) services through API's. This allows the running system to easily swap out data stores or file storage locations thus it was designed with cloud sourcing in mind. 
These choices allow the platform to run very efficiently enabling more users per server and thus reducing the operating costs. Screen casts of these efficiencies in action can be seen here:

posted on Friday, September 11, 2009 at 3:20 PM by David Saintloth

My Service providers are Saa S companies and I can relate to each of your observations. Great insight.

posted on Friday, September 11, 2009 at 3:48 PM by Lalitha Brahma

I think what I enjoyed most about this article was the enthusiastic responses from the many cool companies out there doing great things with the SaaS model. 
As Dharmesh has outlined, SaaS has its challenges. Because we have entered a mature market with our service offering; the approach we have taken @iwmn is to diversify our revenue stream. Not only do we offer the world's most innovative domain registrar service, but we also make the underlying technology available to other registrars. 

posted on Friday, September 11, 2009 at 3:54 PM by Paul Spence

Great Post - I just received the best compliment from my sister - mostly about recommending as a starting place for her Zurich based photography business. 
In the past I've used your stuff to teach entrepreneurship to college students (with your permission - thanks again!) 
One lesson I came up with relates to the dials and gauges idea. I put up a picture of a car dashboard, and ask the students to think about how much time they spend looking at the road vs. the dash. Then I post a picture of the cockpit of a Cessna 172. I get them thinking about how a pilot spends at least as much time scanning their gauges and dials as they do looking outside the aircraft. 
One value to a SaaS business model is that there are so many built in performance indicators to the system. Entrepreneurs HAVE to be able to focus on many different variables while running their companies - and not to lose site of the big picture. 
My experience is that too many start-ups fixate on one thing, and lose track of their "altitude." (mixed metaphor?...) 

posted on Friday, September 11, 2009 at 4:11 PM by Mike Nolan

Awesome article and insite into the SaaS market. I too am a startup and the development cycle has been tough so far. However, one of the things I have learned as you mention in your article "Get really good at efficient experimentation (i.e. I can turn this knob and see it have this effect)" is this: too much development time is quite simply, the life and possibly, the death of your application.  
The service is always there and you need users to debug it for you. What is not working today, could very well be working tomorrow. 
Once again, an excellent article@

posted on Friday, September 11, 2009 at 4:19 PM by Tom cullen

Alexis Wilke  
I do not know what the snide remark was for I am just stating an opinion. 
SAAS was good 5 years ago , still good today. 
My company runs a CLOUD to each their own.

posted on Friday, September 11, 2009 at 4:20 PM by Dan

The infrastructure costs are a definite challenge to the SaaS model. Particularly if you're trying to convert from a consulting to a service model. The two big ones are IT and Customer Support. 
At the outset, it seems like an easy transition, but without dedicate IT and support staff, I've watched billable hours shrink as we ramp up early customers. While a consulting client might have no problem with a 10-20 hour a week consulting bill, getting them to commit to even a tiny fraction of the equivalent in a subscription to pay for support and IT seems nearly impossible. 
But it's no doubt that that's where the value lies. The reason people want SaaS is because that infrastructure is offloaded to the vendor, who (it seems) can no longer get consulting rates. It feels like if I outsourced IT & support to India for $10/hour or less, we'd still be losing money on subscriptions.

posted on Friday, September 11, 2009 at 7:09 PM by Aaron Evans

Some really good insights in this article, especially the point about installing knobs and dials.  
At our SaaS company, SmashFly Technologies, we track EVERYTHING. We have real-time alerts and monitors for basically every transaction in the system. This provides us the ability to stay ahead of any big issues and ensure a quality experience. 
When building a SaaS company, a BIG factor in your success is understanding how to build scalable, solid software that is easy to monitor and maintain. That involves expertise in both software & hardware architecture.  
But, as we've seen, if you have the expertise and build your application right the first time....the SaaS recurring business model is awesome, and the upside is huge.

posted on Friday, September 11, 2009 at 10:28 PM by Mike

Clients and Management Consultants perspective. 
Great posting Dharmesh! We wish there are more successful Saas. 
Saas is a great solution for Small to Medium Size business owners, especially if it is done right. 
There is little up front cost, back up is not an issue nor access to software update. Very importantly its ubiquitous access. 
As a management consultant who focuses on business transformation and improvement, when I look for a great IT solution for our clients’ needs I always prefer a good Saas solution.  
Eric Tjoeng 

posted on Friday, September 11, 2009 at 10:31 PM by Eric Tjoeng

WE have just started on Saas model for our online recruitment software RecruitPro Live -  
Thanks for all these inputs. The market response for a hosted solution on recruitment front is still confusing for me. 
Supriya Shankar 

posted on Friday, September 11, 2009 at 11:40 PM by Supriya Shankar

Not to sound cynical........but going with the fast car analogy, i can't find the equivalent of good brakes in Saas business?? or is it a business specific invention done by the founder??

posted on Saturday, September 12, 2009 at 3:18 AM by anand

We started our SaaS business of providing Intra College Portals called 2 yrs back in Bhopal (India).. and must say we benefited a lot by another of Dharmesh's article on SaaS maturity level. Again a gr8 post. 
- Raghu Pandey

posted on Saturday, September 12, 2009 at 4:59 AM by Intra College Portal

You can offer so many more features and instantly install them. As the leader in SAAS Business plan and due diligence reporting with a video elevator pitching plaform, we have been able to transform the business planning and fund raising process to a whole new level online.Creating a more level playing field for access to capital. 
Great article. 
Ruth E. Hedges

posted on Saturday, September 12, 2009 at 11:36 AM by Ruth E Hedges

timely post. I am interested in # 2 - if this group can share soem options on which platform/development tools to pick to build our SaaS application (its an enterprise software). 

posted on Saturday, September 12, 2009 at 12:34 PM by rajiv c

For the hardware, a cloud is a good idea, but if your needs are quite small at the start, your own computer at your offices is not a bad idea so that way you can quickly setup the computer they way you need it... 
For the software, I strongly suggest you start with an existing CMS such as Drupal (actually, only Drupal is worth it, but anyway...) That already includes everything you need to handle your customers account, including payments to have access to different features. If you want more info on Drupal, feel free to contact me: 
I have many sites running Drupal and it works!

posted on Saturday, September 12, 2009 at 1:18 PM by Alexis Wilke

Having worked in a few startups offering SaaS solutions, I have the following to offer: 
1) Outsource your infrastructure, it is not your core business, let the experts handle it, even if you are an IT rock stars. Since the IT cost needs to scale proportionally with revenue, be very cautious on trial offerings which get your zero revenue. 
2) Invest time and discipline in the backend. The web front end would go through many many iterations, but you really don't want any backend architectural changes and dataset migration. It is very painful. 
3) Get a usable version out as soon as you can, even if it is limited in functionality. More importantly make sure it is clear to end user how to submit feedbacks, which guide your web front end tweaking.

posted on Saturday, September 12, 2009 at 1:31 PM by Wai

One of the challenges we are experiencing from our mid to large size customers / prospects is the fear of providing too much feedback for product improvement. Many of them feel, and they openly say to us - if they share what changes need to be made in product, example new reports etc., they believe, they are giving away their competitive edge, their secret sauce. They know, once those features are implemented they will be available to their competitor also through our product.  
>> Any suggestion how you dealt with it during the early stages of business? 

posted on Saturday, September 12, 2009 at 3:10 PM by Raju Gupta

Your problem can be mitigated against by simply allowing the aspects of the service that can be tailored to a specific client to remain private to that client. As an example, the AgilEntity platform that the web site is built on is a multi-tenant platform. The client designated administrators are free to use the services privately and keep their business processes and workflows proprietary. Now of course, your ability to do this will vary with the type of service you are providing. Numeroom provides business focused collaboration and workflow services. As long as the separation of concerns is built into the lowest levels of your platform allowing clients to retain competitive advantage should be easy to do. 

posted on Saturday, September 12, 2009 at 3:41 PM by David Saintloth

Building and running a SaaS business is a roller coaster ride. One minute you are on top of the world and the next minute you are panicking because a customer found a bug. To me, there is no better feeling than building a brand from scratch and experiencing it blossom resulting in happy customers. 
Important things I have learned 
Might sound obvious, but you have to find good people to work with. It's easy to sniff out the bad ones - if they bring up contracts or money before a single line of code is typed. Better yet, if money comes before their passion, big red flag. 
I heard it preached a million times, "If you launch your software and aren't embarassed by it, then you waited too long." With both me and my business partner being somewhat perfectionist developers, we made sure most everything was polished before launching. Even so, we got early feedback and now looking back 5 months, so much has changed between then and now and our software is 100x better now. Lesson learned, launch as soon as you can and get that feedback. Know ahead of time that most of your software functionality and design, and even pricing will change with fast iterations in the beginning. 
In the beginning, don't get too wild building features that are outside of your core services. For example, we hurried up and built an affiliate program. To this day, we've only had about 5 people sign up and how many new customers have come by way of affiliates? Zippo. We could of and should of spent that time improving the core product. 
Really good customer support goes a long way and really sticks out to customers. My partner and I are by default curtious and we respond to customers in a timely manner. With us always focusing on the software, we were kind of shocked to hear 'top notch customer service' as one of the number one things that our customers like about us. 
It's been roughly 6 months since we launched. As our software has greatly improved, we have aquired a few really happy customers who have then felt compelled to tell others about our services. We have seen this compounding effect increase as the software improves and more customers use us. 

posted on Sunday, September 13, 2009 at 1:29 AM by Matt Brown

Dharmesh ,  
Nice article ,i have worked a lot of PAAS was wondering if i can write to you and get some insights i am planning to start building a product i am based out of Pune(india)

posted on Sunday, September 13, 2009 at 2:19 AM by arvind

Great points - liked the knobs and dials. Too many people don't prove they have failed before they move on to the next sales campaign, feature, project etc.

posted on Sunday, September 13, 2009 at 12:55 PM by ian farmer

All your observations are correct and valuable! SaaS platforms showed their value for some time now and they are still good. In combination with the cloud datacenters you can think it would be a killer of the traditional way of delivering software. In my experience I have learned that having an alternative traditional way of delivering the product, next to your SaaS cloud platform, is not only necessary to insure that customers can forever use their data independent of the platform, but it also brings in valuable revenues that will help with the initial cash-flow issues mentioned in the first point of the article.

posted on Monday, September 14, 2009 at 4:24 AM by Bogdan Calin

Most SaaS businesses are subscription-based (there’s usually no big upfront payment). As a result, sales and marketing costs are front-loaded, but revenue comes in over time.  
Sorry Dharmesh, but your offering is more expensive, per month, than what we'd charge our customers for life (assuming they don't wish to pay for yearly maintenance, which is a fraction of the purchase price). Your product is most certainly, not cheap, or front loaded. Its not like your pricing is $10/month, its more like 50x. 
And we probably add more value to the end customer. Sorry, but I disagree. We looked at your product, thought nice but expensive, then you increased the price, making it a definite non-purchase. 
Sorry. I normally like your articles, but this is so far off the mark. 

posted on Monday, September 14, 2009 at 5:29 PM by Stephen Kellett

Remove the friction that is keeping customers from getting value and you’ll unlock some revenue. 
The friction, in your case, is the $500/month fee to use the software. Its not as if what you are doing is particularly hard, you aren't reverse engineering executables to find safe inject points and so on. Then compare your operation to 37 signals and its $10/month (and upwards price points) and compare to some of your lesser competitors that also price from $10 upwards and its a non-starter. 
Seriously, we have a deal with a very good consultant that has twice built multi-million dollar businesses and sold them. We mentioned your fees to him, then he pointed at a $10/month competitor and said "All the information I need it provided by these people (and yes he was using them, with good results), why should I spend 50x more?". Says it all really.  
I don't understand why you think I should give you $6000/year for your service. That is not value for money.  
Sorry Dharmesh, but as someone that thought they may become a customer of yours and found your pricing so wide of the mark, I find this article really hard to swallow. 

posted on Monday, September 14, 2009 at 5:45 PM by Stephen Kellett

Stephen: Pricing is difficult. You are, of course, entitled to your opinion on HubSpot pricing. 
We do not compare ourselves to lower-end (and sometimes free) offerings. There are some great products at that end of the spectrum. We are actually postitioned against much pricier offerings -- that are usually several thousand a month.  
We have more customers tell us we're not charging enough than those that tell us we're charging too much. 
Having said all that, I still stand by my points in the article. There are no black-and-white answers on pricing. It's tricky to get it right.

posted on Monday, September 14, 2009 at 8:52 PM by Dharmesh Shah

We provide managed hosting and managed services for many ISV's offering a SaaS model. As a result we've seen a number of issues related to delivering the service. Here's one to consider - SLA's. We've seen issues arise when ISV's try to build one infrastucture and sell it with varying Service Level Agreements. They get themselves into problems they didn't consider at the outset. That's just one example. we can help if anyone is looking at hosting related issues of SaaS.

posted on Monday, September 14, 2009 at 10:10 PM by Barry Cohen

@2. Software projects like open source Scalr (on which I work) take a lot of the time element out of managing infrastructure. It provides automation, mysql scaling, and dns handling so that all you have to do is push your code to the cloud. 
Good point about creating the knobs early so you can experiment.

posted on Tuesday, September 15, 2009 at 1:27 AM by Sebastian Stadil

This is a nice concise post about SaaS delivery. I'm the CTO of a SaaS company that delivers eProcurement and sourcing software to large international companies. Even though the model has been around for a little while, there are still quite a few people who don't really understand the difference between the SaaS model and the older ASP model. 
One of the more interesting challenges with enterprise SaaS software is that while each customer gets the same "version" of the software, each customer has slightly different needs. It's the quality of those "knobs and dials" that allows you to deliver a custom environment to each customer from a single multi-tenant instance of your product. 
It's a very challenging to product management to weigh the customer requested enhancements against all other customers since you are basically forcing all customers to upgrade to the new version every time you do a release. It's especially important when you change user interface elements which might require new training since you have to make sure that you get all of end users trained prior to the rollout. Otherwise you end up with some pretty unhappy customers. 
Also, it's worth pointing out that the flip-side of financing your customers is that the SaaS model is very easy to scale. Once you have the infrastructure set up, you can essentially add new customers with very little additional hardware or software expense. 
Another key is to provide rock-solid security and reliability. Make sure you do a SAS 70 Type II audit each year. That is the only way that your customers will feel comfortable letting someone else run a critical piece of their infrastructure. 
If anyone is interested in chatting about this more, feel free to drop me a line. I'm very interested in this topic, especially as it applies to early stage businesses.

posted on Sunday, September 20, 2009 at 8:01 AM by Del Putnam

Being a quintessential customer-driven operations person I think your point number 4 is really crucial for business growth. 
We are a SaaS company in making -- it was good advice. 
Please read the follwoing for similar points: 

posted on Saturday, September 26, 2009 at 10:39 PM by Rini

Comments have been closed for this article.