This is part 3 of my interview with Xactly CEO Chris Cabrera who tells us what it takes to start a new SaaS company and why you’d even want to try.  If you missed Part 1 or Part 2 be sure to go back and check them out.  As always in these interviews, my remarks are parenthetical, any good ideas are those of Chris Cabrera, and any foolishness is my responsibility alone.

Sales and Marketing

Tell us about your use of the Web as a marketing medium.  How are you using the Web? 

Chris:  We do a lot.  The AppExchange was brand new and we were one of the first on it.  We do a lot of webinars constantly.  We do the majority of our selling over the web.  We never see most of our customers before the deal is closed.  We regularly see $500K deals done entirely over the web.  That was unheard of in my previous enterprise experience.  SaaS marketed over the web is way cheaper and more effective.  Cost of sales is much less.

How does SaaS affect the sales process?  Walk us through a typical SaaS sales cycle from initial lead generation to closing.

Chris:  Because of the size of the order and the relatively small commitment from the customer, the hurdles are lower.  For us, we identify a need over the telephone through a discovery session.  How many are paid, plan complexity, etc.  We then sign NDA, get copies of their plans, and often we will build their compensation plans on the fly on a web demo so they see how easy it is.  In the same demo we show the web based results and reporting.  The customer is often ready right at that point to buy and you jump into talking terms and conditions.  Sometimes a 2nd or 3rd demo is needed for large audiences.  Our reps can do it all from their desks instead of getting on planes.

Bob:  (The process Chris describes is as different from the conventional Enterprise process as night and day.  No expensive air travel.  No regional sales offices.  Everything is happening via telephone and the web.  The cost to the customer to get started is so low, it doesn’t make sense to engage in a protracted evaluation at their end either.  Buy it, start using it, and junk it if it doesn’t work.  The customer will have spent less to do that than a normal enterprise evaluation, and if the SaaS solution works, they’re done and recognizing ROI almost immediately.  I’ve shown concrete statistics on how the cost to sell SaaS is less than conventional software,  but Chris Cabrera really brings those dry numbers to life when he walks through his sales cycle.)

SaaS Technology and Operational Issues

What platform is your software running on? 

Chris:  We’re a Java app running RedHat Linux and Oracle.

Is Xactly a multitenant app?  Why is multitenancy so important for SaaS companies?

Chris:  Both our app and database are multitenant.  You can have a multitenant app that still has a single instance database where every customer has their own DB even though the app is multitenant.  Each customer has their own schema.   That’s not a true multitenant architecture.  You need to support one line of code so all the customers are on that line.  I can then pass savings along that way.

Bob:  (There’s a lot of controversy about what multitenancy means.  Chris has an interesting definition:  every customer must have the same schema and the same software version.  The schema issue goes to using metadata for customization instead of modifying the database tables with expensive custom programming.  The software version goes to keeping every customer up on bug fixes and making sure the ISV only has to keep one code line running. )

Chris:  My last company had 80 engineers, but 30-40 spent their time supporting old releases and many platforms.  So I can accomplish with half the engineers just as much.  You get benefits throughout the company from this focus that are similar.  And by the way, it’s one reason it is so hard to mix the two religions together because you lose that without multitenancy.

Bob:  (If you’re wondering what the 30-40 engineers have to do in a conventional company, Chris is right on.  You support many platforms–different flavors of Unix, different application/web servers, different database servers, and a myriad of other platform related issues:  LDAP, Active Directory, the list goes on and on.  On top of that, you’re supporting many versions of the software.  Given the pain of upgrading conventional enterprise software to a new release, most customers wait years to do it.  Meanwhile, they need minor enhancements and bug fixes to the version line they’re on.  All of that easily adds up to the figures Chris is quoting–a substantial savings for SaaS that makes smaller SaaS companies competitive with larger conventional companies.)

Chris:  One of the big differences in these religions has to do with the culture of the companies.  If I look honestly at conventional Enteprise, the focus is closing the current deal.  When it closes, we move on to skin the next big fish.  The SaaS world forces the culture of the vendor to be focused ground up not on the next deal but keeping the customers you have happy.  No single deal can make or break your company. It’s all the deals.  It’s pervasive everywhere.  This culture is so much more customer friendly.  The old way had us on opposite sides of the table from the customer.  Here, the customer is always in control. 

Bob:  (Steve Singh and other SaaS executives I’ve talked to are passionate about this issue.  Anyone familiar with conventional ISV’s has also seen the syndrome Chris describes where a sales person will promise anything to sell the big deal, and then they just move on to the next one while the ISV struggles to deliver.  It’s not a happy situation for customer satisfaction.)

Tell us about your trials and tribulations with hosting and data centers:

Chris:  This is our main cost of sale and one of our biggest expenses.  The hardest thing is getting to critical mass.  You have to invest ahead of the curve on a lot of it.  We’re just getting to volumes where we can see economies of scale.  For 2 years that wasn’t the case.

Bob:  (At some point I will do some interviewing and write some posts to try to nail down these costs for SaaS.  They’re a significant aspect of the business economics of every SaaS vendor I talk to, and have a huge effect on profitability, so it’s worth chewing on it.)

What’s next for Xactly?

Chris:  We’re going to keep building out our Sales Performance Management vision, which is leveraging the data.  Every company should be thinking not just comp automation, but tools to provide analytics, modeling, programs, and really using behavior to drive the business.  Not enough VP’s of Sales and senior finance people wake up and see this vision as a top priority.  We want to be the company at the forefront of that emerging space to show companies how to dramtically increase top line revenue.

Conclusion

I’ll close with an anecdote that builds on Chris Cabrera’s contention that driving sales behaviour is highly strategic.  Steve Singh told me that originally, his salesforce was incented to push a balanced proportion of the deals to on-premises to cushion the shock of the SaaS transition.  This was built into their sales compensation, and it resulted in a nice 50/50 balance of sales for both models.  A little later, they changed the comp plans so that neither model was favored and it was all up to the customer to choose.  Suddenly, 90% of the deals went SaaS.

Imagine being able to use compensation changes to help your sales force change customer behaviour in nearly half the deals you engage in.  What would you do with such a tool?  What are your urgent priorities as a business, and how do you achieve sales (and the rest of your organizations) alignment behind those priorities?  There’s nothing quite like a paycheck to make things happen.

Reed Taussig, the original CEO at Callidus, used to say that compensation was the key to herding cats.  “Trying to get cats to follow you is hard”, he’d say.  The trick is to carry a saucer of milk and then they’ll follow you anywhere.