Sunday, March 23, 2008

DaaS - Data as a Service

I have been preoccupied lately with a software model that is a cousin of Software as a Service (SaaS) - one that I will call Data as a Service (DaaS). Much like SaaS it is more of an evolution of an existing model than something completely different - it just seems to me that the new generation of data delivered as a hosted service are breaking new ground in terms of access to data and analytics. I think more interesting companies will be built around this model.

Definition - DaaS
A software as a service or web service offering that provides customers with access and analytics around a set of proprietary set of aggregated data. The most interesting of DaaS offerings are ones which collect this data from individuals or companies by providing immediate value back for these efforts, but that the aggregated data is able to be resold to a different set of constituents.

Example - collects data by offering individuals the ability to benchmark their compensation levels against others by sharing data into the system. This data is then aggregated and anonomized to be resold to companies (usually HR managers) for hiring and compensation related usage. went public in 2006 and has a market cap of 100M.

Example - Dun and Bradstreet
Dunn and Bradstreet (D&B) is a 165 year old business that provides credit and commercial information. I have used their service in the past to research the credit history of customers and partners before extending them credit. D&B has a team of people who call companies to gather the data and then they resell it back to other companies - both on individual companies and for industry segments. D&B also sells this data for several other uses, for example, to marketing professionals to cleanse contact lists.

Why the DaaS model works

- Virtuous cycle on data - as the DaaS company acrues more data, it can provide increasing more useful data back to both the contributors and the purchasers. The data becomes a competitive advantage that an upstart would have trouble replicating. This is especially true if the data needs scale across different geographies.

- Analytics and actionable business drivers - the data is the basic asset, but it is significantly enhanced by wrapping analytics into the offering. Some DaaS companies pull their data from publically available sources but an idividual user would have trouble pulling the data, navigating the data and building analytics to make it useful. A DaaS company can make this investment as they can sell the data many times over.

Unfortunately I think that the acronym is still up for grabs with a few alternative software meanings:

Development as a Service - Salesforce
Datacenter as a Service - Cloud Computing
Database as a Service - LongJump
Department of Aging and Adult Services - US Government
Data as a Service - StrikeIron (this one is closest to my definition, but is actually DSaaS - Data Service as a Service)

The message for me is that all companies, SaaS or otherwise, should look at what data assets they have (or could easily have) and consider a DaaS offering. If it can provide useful insights - it will be well worth the development costs.

Monday, March 17, 2008

Magic Number for SaaS Companies

Guest post on Will Price's blog - reposted here.

Josh James, CEO of Omniture (a Hummer Winblad portfolio company), gave an inspiring talk on building a SaaS company last week at the Opsource summit. Josh walked through the history of Omniture as a case study for building a SaaS company. He talked about the need to invest in the company with a firm hand on the wheel as the recurring revenue slowly built up over time. He outlined the different stages of evolution of the company:

1) Product: build a rock solid product. Prove you can sell it as founders before moving past this step.
2) Sell: Sell like crazy, build out a team, hire some QBSRs (Quota Bearing Sales Reps)
3) Retention: focus on churn and retention issues, hire more QBSRs
4) Marketing: spend on marketing, hire more QBSRs

The next phases, not surprisingly, also included hiring more QBSRs but interestingly it is not until later that investments in efficient infrastructure and operations hit their ToDo lists. This outline displays a strong focus on finding a product market fit and then adding gas to the fire as the market opened up. The key metric that Omniture used to decide how much gas to pour on the fire was the Magic Number.

The Magic Number

The magic number ("MN") is a metric that can be used to tell you the health of your company from the perspective of growing monthly recurring revenue ("MRR"). It is a common mode metric to compare companies MRR scaled by sales and marketing spend. The MN provides insight into the effectiveness of previous quarter Sales and Marketing spend on MRR growth. Your MN will be penalized if the spend is wasted (bad marketing, bad sales execution), if your churn is high or if the market has issues (saturation, competitive forces). It also has a very high correlation with Q/Q growth rates so in general, high Magic Numbers are good.

To calculate:

QRev[X] = Quarterly Recurring Revenue for period X
QRev[X-1] = Quarterly Recurring Revenue for the period preceding X
ExpSM[X-1] = Total Sales and Marketing Expense for the period preceding X

Magic Number = (QRev[X] – Qrev[X-1])*4/ExpSM[X-1]

For example, consider a hypothetical company with the following financials

Q1 Q2 Q3
Revenue (recurring total) 1M 1.2M 1.5M
S&M Expense 800K 900K

Then the magic number is 1.0 for the end of Q2 and 1.33 for Q3.

Fundamentally, the key insight is that if you are below 0.75 then step back and look at your business, if you are above 0.75 then start pouring on the gas for growth because your business is primed to leverage spend into growth. If you are anywhere above 1.5 call me immediately.

Josh provided the following gas-pouring throttle chart for SaaS companies to evaluate how much to invest in their go-to-market spend. The data on the charts if from Omniture and other public SaaS companies.

Calculate yours…and get back to me if it is interesting! For fun and extra credit take a look at difference in Magic Number for some of the public SaaS companies like Omniture and SuccessFactors. I can be reached at

Friday, March 07, 2008

Eat Your Own Dogfood - SaaS vendors
aka Drink your Own Champagne

Repost from my guest entry on Will Price's blog.

Guest post by Hummer Winblad's Lars Leckie.....Lars drove our firm's investment in Aria Systems and, most recently, vKernel.

Last week Phil Wainewright wrote a great post (as he always does) called SaaS vendors, eat your own dogfood, or die. In the post, he describes how SaaS companies need to embrace the SaaS services available in the ecosystem.

I support the viewpoint from the post that SaaS companies need to have religion and leverage SaaS in everyway they can – further, I believe if they don’t they leave themselves open for other SaaS companies to disrupt them. From discussions with the infrastructure management of many leading SaaS companies I often hear how they are forced to use some on-premise pieces on the back-end reluctantly. This has provided the motivation for a few of our SaaS infrastructure investments (eg. Aria – SaaS billing and customer management).

Phil’s title had me thinking along another important vein for SaaS companies…literally to eat THEIR own dogfood and use their own product. For example, Salesforce aggressively uses Salesforce to manage prospects, Omniture eats their own analytics for online marketing, Teleo uses their own product for recruiting and SuccessFactors brags about their own talent management. Luckily we don’t all work for tobacco companies…

My belief is that SaaS companies must use their product for a few reasons:

1) Point of View: it puts everyone in the company in the seat of the customer. This means that the internal teams will live the same pain, the same experience and the same leverage that you are espousing as a vendor. This POV extends the advantages SaaS has of bringing the vendor closer to the customer by putting the customer in the office next door.

2) Analytics: By using your own product you will think about the product extensions and depth of how to apply the benefits of analytics to build stronger products and best practices. SaaS companies have a huge edge with analytics so it makes sense to use them internally as well.

3) Sales Roadblock: a fair question for a prospect to ask would be, “if your solution is so great, why aren’t you using it?" SaaS vendors can do their sales team a great favor by getting ahead of this question.

Any interesting SaaS companies that are using their own product and want to reach out – I’d love to hear from you. Please email me at

Labels: ,

Thursday, March 06, 2008

VKernel and Hummer Winblad Join forces on Virtualization Infrastructure

This is a repost of a blog entry I contributed as a guest writer on Will Price's blog.

Hummer Winblad is pleased to welcome Alex Bakman and his VKernel team to the Hummer Winblad family. We are very excited for the opportunities ahead and look forward to working together. The recent press release and further details can be found here.

Enterprise IT environments are undergoing one of the biggest shifts in 25 years – shifts that have not been seen since the move from mainframes to client-server. Big shifts in infrastructure open up large gaps in the current management solutions. Virtualized environments provide many benefits to the enterprise from flexibility to lower TCO but along with that have introduced a few headaches too…

- “vmotion” – servers growing legs and dynamically moving
- “vm sprawl” – servers multiplying at unmanageable rates
- Shared resources – performance, monitoring, resource accounting, etc
- Cost visibility – no longer tied to physical resources

vKernel is a new platform for system management in enterprises that are embracing virtual infrastructure to power their businesses. Enterprises IT managers face increasing challenges as virtualization spreads within the datacenter. vKernel provides an essential suite of virtual appliances tailored to meet the virtualized datacenter including chargeback and capacity planning management. These appliances require zero installation, are quick to deploy and provide insights within minutes.

More information can be found on vKernel’s website and webinars – or download the latest virtual appliance to try it out today.

Hummer Winblad’s enthusiasm in virtualization is captured by our investment in vKernel as well as several other companies including Scalent and Akimbi (acquired by VMWare).
Aria Systems Raises Series A - Leave the Billing to Us

This is a re-post from a post I did on Will Price's blog.

Last week, Hummer Winblad Venture Partners announced the Series A funding for Aria Systems, the leading on-demand billing and customer management provider.
HWVP’s investment in Aria is the firm’s fourteenth SaaS investment (others include Omniture, Adforce, Employease – some of which date back to 1998).

The investment in Aria is a double down on the SaaS market – Aria is not only itself a SaaS ISV, but also the company is a core component of any SaaS company trying to efficiently scale. RBC Capital Markets predicts that the SaaS market will grow at 43% CAGR to be $28bn by 2010. That’s a lot of billing.

As the SaaS market grows, the need for ISV’s to make bespoke investments in data center operations, billing infrastructure, and other core, but non-differentiated, functions is lessening by the day. Providers like Aria allow ISV’s to focus 100% of their human and financial capital on differentiated application logic, while using best-of-breed platform vendors for infrastructure services. The company puts it best when it tells customers it will help them, "get customers, get paid, get control."

Why invest in Aria? Simply put, traditional billing solutions, independent of expensive on-premise software, are not well suited to subscription and usage based billing models. New revenue models bring about the need for new billing architectures. Emerging SaaS companies who don’t use Aria can either choose to build a billing system themselves, thereby diverting engineering resources to a non-core engineering function, or they can use existing on-premise solutions with a myriad of spreadsheets and manual accounting. Aria is a solution that can scale and provide the best infrastructure through a strong focus on the billing market.
The Aria solution is a complete turnkey SaaS (single instance, multi-tenant) solution to manage recurring revenue billing. They manage the quote-to-cash system for a company including hosted customer sign-up, customer service, provisioning of service, dunning and all the other components needed for a full billing solution.

Aria is already in the market with strong customer traction and key partnerships in place. The vertical markets that Aria addresses today are SaaS, On-line Gaming and Telco. A few of their partners in these markets are Opsource and GNi but you can see the full list of over 25 partners here.

Congratulations to Ed Sullivan and the rest of the Aria team - Welcome to the Hummer Winblad family.