Studying ExactTarget – Post #2: Cost of Revenues and Operating Expenses

In this post, I wanted to explore a bit about how a high growth SaaS company has been spending in various functions against top line revenues.  I’m using consolidated financial statement data from their S1 (link here) between 2002 and 2006.  Given that ExactTarget was founded in 2000, this represents years 2 through 6 of their business.

Cost of Revenues – 2002: 1% -> 2011: 34%  |  First 5 Years Average: 13%

It’s clear that as ExactTarget scaled, two major costs did as well.  The first was the size and complexity of their data center operations, and the high-availability requirement of their service.  The second was the professional and support services required to enable customers, particularly high end clients.  Cost of Revenues seems to be hovering in the 30%-35% range for now.

Note that the first five years, Cost of Revenues averaged 13%, presumably including whatever support and professional services existed in those early years.

Cost of revenues consists primarily of wages and benefits for software operations and optimization services, as well as depreciation, licensing, maintenance and support for hardware and software used in production, and co-location facilities, bandwidth and infrastructure expenses. The expenses related to co-location, bandwidth and infrastructure are affected by the number of clients using our application, the complexity and frequency of their use, the volume of messages sent and the amount of stored data. In addition, these expenses are impacted by our requirement to maintain high application availability and redundancy. We expect these expenses to increase in absolute dollars as we continue to expand our business and serve the needs of larger and more sophisticated enterprise clients.

Sales & Marketing – 2002: 101% -> 2011: 45%  |  First 5 Years Average: 71%

As top line revenue scaled, the percentage of dollars spent on sales and marketing dropped from 101% of top line revenue to 45%.  The latest 10-Q (Q212) shows 39% spend in this function, so one might be able to say that it’s leveled off around 40%.

In the first five years, the average spend was 71% of to line revenues.

A side note about growth of top line revenue: It’s interesting to me that it took 5 years to get to $11M, given that VCs often ask me how we’re going to be a $100M company in 5 years.  This poster child of a SaaS company focused on the marketing vertical was at $11M in year 5, and any VC out there would want to be a part of this syndicate.

Here’s management’s commentary on Sales and Marketing expense:

Sales and marketing expenses consist primarily of wages and benefits for sales and marketing personnel, sales commissions, travel and entertainment expenses, and lead generation marketing programs. All sales and marketing costs are expensed as incurred, including sales incentives and commissions. Sales incentives are expensed in the period of contract signing and commissions are expensed upon receipt of payment. Our sales and marketing expenditures have historically been highest in the last two quarters of each year, which are periods of increased sales and marketing activity. In order to continue to grow our business and increase our brand awareness, we expect to continue investing substantial resources in our sales and marketing efforts. As a result, we expect that sales and marketing expenses will increase in absolute dollars as we invest to acquire new clients, but decrease as a percentage of revenues as our existing client base represents an increasing portion of our total revenues.

Research and Development – 2002: 38% -> 2011: 20%  |  First 5 Years Average: 24%

R&D expenses, likes sales and marketing, start high and drop by roughly half over 5 years of operations, where they stabilize.  I would have guessed R&D expense to be higher, especially in the early years.  Here’s how ExactTarget categorizes R&D expenses:

Development expenses consist primarily of wages and benefits for product strategy, product architecture, product design and development and quality assurance personnel. We focus our development efforts on usability, application performance, new features and functionality, and development of emerging one-to-one marketing technologies. We expense development costs as incurred. Direct development costs specific to new functionality is minimal due to the relatively short development cycle. Such costs are instead recognized as a component of development costs when incurred. We expect that development expenses will increase in absolute dollars as we continue to enhance our product offerings, but decrease as a percentage of revenues as we continue to scale our business.

General & Administrative – 2002: 45% -> 2011: 13%  |  First 5 Years Average: 21%

While most of my analysis is focused on the early years (which averaged 21%), its interesting in the management notes below that they are projecting an increase in both absolute dollars and as a percentage of top line revenues that G&A expenses will go up as a result of being a public company.

Management’s commentary:

    General and Administrative.    General and administrative expenses consist primarily of wages and benefits for executive, finance and accounting, legal, human resources, internal information technology support, and administrative personnel. In addition, general and administrative expenses include professional fees, bad debt expenses and other corporate expenses. We expect that general and administrative expenses will increase as we continue to add personnel to support our growth. We also anticipate that we will incur additional costs for personnel, professional services including auditing and legal services, insurance, and other corporate governance related costs specific to operations as a public company. We expect that our general and administrative expenses, in total, will increase in absolute dollars as we grow and operate as a public company. We expect that the costs as a percentage of revenues will increase in the next twelve months.

 

 

Here’s my spreadsheet.

Advertisements

Studying ExactTarget – Post #1: Cost of Revenue, Subscription v Prof Svcs, Gross Profit Margins

ExactTarget (ET) is currently on a tear.  Founded in 2000, they raised $75M on the way to going public in April of this year, and have already made $115M+ worth of acquisitions (Pardot for $95M, and iGoDigital for $21M).  As of this post, they have $1.5B market cap and rumors of buyouts from SAP, Oracle, Salesforce.com and Microsoft (Bloomberg article).

ExactTarget serves as an interesting SaaS model in the interactive marketing space, and one that many SaaS startups (Windsor Circle included) can learn from.  In this post, I’d like to explore how quickly ExactTarget grew, its gross profits as it did so, and the amount spent on various functions in the business.

Cost of Revenue

For the past 4 years of reported financial data, ExactTarget has consistently reported 33%-34% CoR.  Raw data can be seen here.

Subscription Revenue v. Professional Services Revenue

Average 84% of Revenue from Subscription Software, Cost of Revenue 23%

ExactTarget’s subscription revenue has been shifting a bit.  In 2009, 87% of its revenue was from software subscriptions.  For the 6 months ending June 30, 2012, it was 80%.  On average, looks to be about 84%.

Management notes in the S1 show this as them going up market and needing to handle more professional services on behalf of larger clients.

On average, the Cost of Revenue (CoR) for the SaaS software revenue is about 23%.

Professional Services Revenue Averages 16%, Cost of Revenue Averages 89%.

The inverse of the revenue model is obvious.  What you’re not making in software revenue falls to professional services.  But the interesting thing here is the CoR…  They only make about 11% gross margin on this.  It’d appear that they are basically covering their costs of professional services, showing that this is an enablement play to ensure that their customers are getting enough value from full usage of the software to stick around year after year.

I’ll dig a bit deeper into amounts spent in each function in the following post.

Here’s my spreadsheet.