GROWTH EQUITY FOR TECHNOLOGY LEADERS. Cloud Computing: GMPP, rcac and the Importance of Component Level Analysis

Transcription

1 GROWTH EQUITY FOR TECHNOLOGY LEADERS : GMPP, rcac and the Importance of Component Level Analysis April 2011

2 Updata Partners Overview Updata Partners provides growth capital to software and technology driven businesses. We invest at the expansion phase of a company s lifecycle in sectors where we can apply our operating experience and industry network to create value. We work closely with our portfolio companies to enhance their operational, financial and strategic decision making. Since our founding in 1998, we have partnered with exceptional management teams at over 30 leading technology companies. Today, we manage approximately $500 million of committed capital and are currently investing our fourth fund, Updata Partners IV. About the Authors CARTER GRIFFIN General Partner NEIL HARTZ Associate cgriffin@updata.com nhartz@updata.com Prior to joining Updata Partners in 2005, Carter cofounded Brivo Systems in 1999 and served as Chairman and CEO until selling the company to a unit of Duchossois Enterprises Group in December Brivo Systems pioneered the software as a service model in the physical security market by introducing the first ever on demand system for facility access control. Before starting Brivo Systems, Carter spent four years as a Senior Vice President at Kaiser Associates, where he advised Fortune 500 clients on competitive positioning and new market entry strategies. Earlier in his career, Carter worked in London for the Coca Cola Company and held positions at American Management Systems and Arthur Andersen. Carter is a past Co Chair of the Mid Atlantic Venture Association Capital Connection. He holds a B.S. in business administration from the University of North Carolina at Chapel Hill and an M.B.A. in finance and marketing from the J.L. Kellogg Graduate School of Management at Northwestern University. Neil Hartz joined Updata Partners in He supports the firmʹs business development, deal sourcing and due diligence efforts. Prior to joining Updata Partners, Neil worked as a Senior Analyst in the Technology Investment Banking Group at Montgomery & Co., where he provided strategic advisory services to growth stage companies. While at Montgomery, Neil focused on transaction sourcing and execution in the software and technology enabled services sectors, with a specific emphasis on software as a service and recurring revenue businesses. His experience includes private placement and merger and acquisition transactions. Neil holds an A.B. in Economics from Dartmouth College. 2

3 Introduction In 2010 Gartner predicted by 2012, 20% of businesses will own no IT assets. Said another way, by next year 20% of businesses will run entirely on leased or employee owned devices connecting to cloud based services! While the timing may be up for debate there is no denying that cloud computing is revolutionizing the world of software and IT infrastructure. At Updata Partners, we are helping fund this revolution one third of our companies are built on cloud computing business models, and the majority of new software deals we seriously consider are cloud based. We believe a similar transformation is required for business operators to properly analyze these businesses. Through our experience we have developed a framework for breaking down the unit level customer economics underlying these models. For the purposes of this document we focus on SaaS multi tenant application software delivered over the web with a recurring revenue model although the metrics are generally applicable to any cloud company or recurring revenue business. While there are many ways to measure a SaaS business, we believe there are two metrics that matter most: Gross Margin Payback Period (GMPP) and Return on Customer Acquisition Cost (rcac). GMPP is the number of months required to break even on the cost of acquiring a customer. rcac adds the element of customer churn/retention into the equation by calculating the multiple of the acquisition cost provided by the lifetime gross margin. GMPP and rcac offer powerful insights but are insufficient if calculated only at the company level. Company level metrics ignore the fact that most SaaS vendors sell multiple products through a variety of channels and acquire customers over many months, quarters and years. Accordingly, we believe component level analysis is needed to capture nuances across products, channels and vintages. This more granular view allows comparison between and among components so that a company can optimize its product mix and sales and marketing budget. Getting the math right on these straightforward concepts can be tricky, so we lay out our approach below. Our metrics and suggested sequence of analysis are as follows: Step 1. tcac Total Customer Acquisition Cost Step 2. ARPU Average Revenue Per User Step 3. RGP Recurring Gross Profit Step 4. GMPP Gross Margin Payback Period Step 5. elt Expected Lifetime Step 6. LTV Lifetime Value Step 7. rcac Return on Total Customer Acquisition Cost 3

4 Step 1: Calculate tcac Total Customer Acquisition Cost tcac is the fully burdened unit level investment required to sign up a new customer, including net one time onboarding costs. We believe a proper CAC calculation involves consideration of all departmental costs of sales and marketing plus one time costs. Oftentimes companies look only at the variable cost of customer acquisition, such as sales commissions and marketing campaign expenses. While this is an appropriate way to calculate the economics of acquiring the next marginal customer, we do not believe it reflects the full cost of customer acquisition after all, the segmentation work by the product manager and the cut sheet from the marcom group also helped bring in the deal. Variable only CAC also fails to recognize that fixed costs must scale over time as the company grows, usually in a stair step function as infrastructure is added. Fully burdened sales and marketing acquisition cost is a good start but doesn t tell the whole story don t forget about onboarding that customer! These are the implementation or provisioning processes that are required to light up a new account (e.g., training, data migration). Any upfront expense or capex outlay (net of what is billed back to the customer) should be rolled into the onboarding cost and included in tcac. Lastly, tcac must be reported by component to represent the cost of acquiring specific customers rather than the theoretical average customer. In Figure 1 we show tcac by channel and by product. Figure 1: tcac calculation looking at a channel component view Organic will often appear more favorable than it really is some of these customers may have initially come in through another channel but delayed their purchase tcac (per customer) $18,300 $14,340 $11,600 $11,000 $2,000 Two difficulties encountered when isolating tcac by component are attribution and cost allocation. Attribution is difficult, especially in marketing, because one channel often drives the end result in another. For example, a display advertising campaign may generate customer interest that results in a sale through another channel. These untraceable accounts often fall into the organic channel, inflating its apparent efficacy. The second issue of cost allocation arises because it s not always clear how to allocate items like marketing overhead or onboarding costs. After all, some customers will require extra handholding at the outset. To solve these issues we are not advocating hiring a statistician for attribution or a consultant for an activity based accounting exercise! Rather, we recommend the development of repeatable, clearly defined methods for attribution and cost allocation to accurately benchmark tcacs across different components. 4

5 As noted above, the focus on tcac should not obviate the need to look at variable CAC. Variable acquisition spend, specifically online marketing, is one of the easiest levers a SaaS business can exploit to accelerate nearterm growth and should still be monitored and managed closely. Note: For the purposes of this exercise we are not accounting for variability in sales cycle duration in our calculation of tcac. Sales cycles can fluctuate across companies, not to mention between products and channels if it typically takes three months to move from lead to close, then we would look at three months of acquisition expense against the corresponding three month total of acquired customers to determine tcac. Step 2: Calculate ARPU Average Revenue Per User ARPU is the average revenue per user on a monthly basis. While company level ARPU will tell you the average revenue per user across the entire customer base, it ignores variability across the product, channel and vintage components. For example, different products have different economics a $300 Basic product should not be lumped in with a $1,500 Premium offering to assess ARPU. Also, tracking ARPU over time within an individual vintage will illuminate upsell/downsell trends a significant factor in the efficacy of the SaaS business model. Figure 2: Expanding the analysis to include ARPU, calculated independently for each component ARPU (Monthly) $2,000 $1,700 $1,200 $1,100 $1,000 ARPUs can vary significantly across products and channels Step 3: Calculate RGP Recurring Gross Profit RGP is the gross profit generated each month. ARPU less recurring Cost of Goods Sold (COGS) will yield RGP. Typical recurring COGS items include the cost of customer delivery (e.g., datacenter usage), the cost to support the customer (e.g., call centers) and payments to 3 rd parties (e.g., software license fees). The key is to include all the month to month costs required to maintain a customer that is already live on the software, but to exclude the initial expenses necessary to light up a customer (those one time expenses were captured in tcac). There will be a mixture of fixed COGS (e.g., servers) and variable COGS (e.g., merchant fees) here. 5

6 Figure 3: Adding recurring gross profit to the equation by netting out recurring variable costs from ARPU Recurring COGS Data Center $15 $9 $12 $9 $6 Customer Support $45 $48 $51 $42 $33 Merchant Fees $9 $15 $6 $11 $9 Total Recurring COGS $69 $72 $69 $62 $48 Recurring Gross Profit (RGP) $306 $183 $171 $118 $102 Recurring Gross Margin 82% 72% 71% 66% 68% Recurring Gross Margin 83% 81% 78% 79% 77% It is critical to separate tcac related upfront costs from recurring COGS Note: RGP does not necessarily conform to GAAP accounting and neither do many of the metrics in this paper. Instead we are trying to focus on the intrinsic unit level economics. For example, we include onetime costs such as onboarding in tcac, but they would likely fall under COGS with GAAP accounting. Similarly, items that typically are capitalized and then depreciated over their lifetime (e.g., devices shipped to the customer) are instead recognized as an upfront cash expense in our framework. If billing monthly by credit card, we also need to make sure that merchant fees associated with the transactions are included within recurring COGS. Step 4: Calculate GMPP Gross Margin Payback Period GMPP is the number of months required to break even on the cost of acquiring a customer. Dividing tcac by RGP allows us to determine how many months it takes to break even on the acquisition investment (GMPP = tcac / RGP). Utilizing GMPP to compare components is one of the first levels of analysis that pulls multiple metrics together to derive actionable insight. Assuming proper cost allocation and attribution analysis, GMPP offers intelligence into which channels to feed and which channels to starve. 6

7 Figure 4: Combining the elements above to derive the GMPP Recurring COGS Data Center $15 $9 $12 $9 $6 Customer Support $45 $48 $51 $42 $33 Merchant Fees $9 $15 $6 $11 $9 Total Recurring COGS $69 $72 $69 $62 $48 Recurring Gross Profit (RGP) $306 $183 $171 $118 $102 Recurring Gross Margin 82% 72% 71% 66% 68% Gross Margin Payback Period (GMPP) Gross Margin Payback Period (GMPP) Looking at tcac alone suggests affiliates are the most efficient paid channel, but CPC with the highest tcac actually has the most rapid GMPP Before jumping to conclusions about marketing budget reallocations, however, it s important for high growth SaaS businesses to realize that acquisition channels are not perfectly elastic. Channels that appear scalable at 100 monthly additions with rapid GMPP may not scale to 1,000 monthly additions due to the increasing marginal tcac. For example, CPC campaigns that work at low volumes can get prohibitively expensive at high volumes due to lack of online inventory. Having said that, we are enthusiastic advocates of A/B testing, new channel development, and searching for incremental gains. Armed with this component level analysis, SaaS companies have a better chance at getting closer to the elusive efficient frontier of channel mix. Note: For simplicity in our example we are assuming nominal cash flows, but in reality the future cash flows should be discounted for the time value of money. This factor has a compounding effect on churn, a metric we will explore in the next section. 7

8 Step 5: Calculate elt Expected Lifetime elt is the length of time a company expects to keep a paying customer. While some customers will churn out immediately and others will stick around for an eternity, elt is concerned with the average lifetime across a group of customers. Converting the familiar churn metric into elt (elt = 1/churn) allows for an intuitive comparison to Gross Margin Payback Period. Recognizing that companies might not have detailed cohort analysis to model long term retention curves, month to month calculations of churn within a segment of customers is an adequate proxy. (Note, we have a serious problem if elt is shorter than GMPP!) While some SaaS companies have long term contracts that suggest a predetermined minimum customer life, we care more about customer stickiness, not contract length. After all, contracts can be renewed and broken. Different acquisition channels often yield customers with different profiles. Organic customers are often sticky because they sought the company directly; customers from affiliates are often flighty because a third party did some or all of the sales work. In our example, the component view demonstrates that Premium customers are more loyal than Basic customers. Using company level elt rather than component level elt would cripple our ability to draw real conclusions about existing customer behavior. Figure 5: Expected Lifetime (elt) is derived from churn and added to our analysis Recurring COGS Data Center $15 $9 $12 $9 $6 Customer Support $45 $48 $51 $42 $33 Merchant Fees $9 $15 $6 $11 $9 Total Recurring COGS $69 $72 $69 $62 $48 Recurring Gross Profit (RGP) $306 $183 $171 $118 $102 Recurring Gross Margin 82% 72% 71% 66% 68% Gross Margin Payback Period (GMPP) Monthly Churn 2.5% 2.9% 3.1% 3.3% 4.5% Expected Lifetime (elt months) Expected Lifetime (elt months) Can t ignore the profound effect of churn a delta in monthly churn of less than 1% between CPC and Affiliate reduces elt by 25% We also need to discuss a concept even more important than account churn the idea of dollar churn. While SaaS companies will constantly be fighting a losing battle against account churn (at best, breakeven), the good news is that customers who stick around often increase the size of their subscription over time. Account 8

9 growth can occur for several reasons: periodic price increases; a growing customer requires more seats; or satisfaction with the product leads to the purchase of additional modules. Furthermore, there is such a thing as good churn that occurs when low ARPU, support intensive customers leave and resources can be reallocated to higher profit customers that have a greater chance for upsell. Step 6: Calculate LTV Lifetime Value LTV is the economic value, net of costs, delivered over the life of a customer. While GMPP is a great tool for comparing the efficiency of different components from a time to payback perspective, LTV takes it one step further to incorporate elt. LTV can be derived by multiplying the RGP by elt, and then subtracting tcac (LTV = RGP x elt tcac). As previously noted, we encourage discounting future cash flows for the time value of money, but for simplicity we have not shown that here. Because the fully burdened cost to acquire a customer and any variable recurring cost required to support the customer has already been removed, LTV goes towards paying off the remaining fixed costs in the business G&A and R&D where significant operating leverage can be found with scale. Figure 6: Determining LTV based on the expected lifetime Recurring COGS Data Center $15 $9 $12 $9 $6 Customer Support $45 $48 $51 $42 $33 Merchant Fees $9 $15 $6 $11 $9 Total Recurring COGS $69 $72 $69 $62 $48 Recurring Gross Profit (RGP) $306 $183 $171 $118 $102 Recurring Gross Margin 82% 72% 71% 66% 68% Gross Margin Payback Period (GMPP) Monthly Churn 2.5% 2.9% 3.1% 3.3% 4.5% Expected Lifetime (elt months) Aggregate Gross Profit Contribution $12,240 $6,395 $5,472 $3,546 $2,244 Lifetime Value (LTV) $7,815 $2,270 $2,397 $1,596 $1,854 Lifetime Value (LTV) $59,720 $55,887 $23,968 $24,629 $28,030 Channel level analysis reveals major variability in LTV and customer acquisition efficiency 9

10 As Figure 6 illustrates, LTV can vary significantly between channels. In our example, the CPC channel provides more than four times the lifetime value of the affiliate channel, despite having a total cost of acquisition that is more than twice as expensive. Step 7: Calculate rcac Return on Total Customer Acquisition Spending rcac is the multiple of the acquisition cost provided by the lifetime gross margin, and can be calculated by dividing the aggregate gross profit contribution by acquisition cost (rcac = RGP x elt / tcac). rcac brings it all together by combining the raw unit economics of GMPP with expected customer lifetime. By utilizing GMPP and rcac together, we can quickly determine 1) the time required to recoup the cost of acquiring a customer, and 2) expected leverage on that acquisition spend. The two metrics together are crucial when making decisions about the most efficient ways to allocate resources. A channel with a rapid GMPP but very little rcac ultimately means little profit will be realized from customers because they churn soon after the breakeven point. Conversely, customers with great rcacs but very long GMPPs create a substantial need for capital to weather the storm until they become profitable (and don t forget to discount those future cash flows). Figure 7: The complete equation now we can use the rcac paired with the GMPP to quickly compare channels Recurring COGS Data Center $15 $9 $12 $9 $6 Customer Support $45 $48 $51 $42 $33 Merchant Fees $9 $15 $6 $11 $9 Total Recurring COGS $69 $72 $69 $62 $48 Recurring Gross Profit (RGP) $306 $183 $171 $118 $102 Recurring Gross Margin 82% 72% 71% 66% 68% Gross Margin Payback Period (GMPP) Monthly Churn 2.5% 2.9% 3.1% 3.3% 4.5% Expected Lifetime (elt months) Aggregate Gross Profit Contribution $12,240 $6,395 $5,472 $3,546 $2,244 Lifetime Value (LTV) $7,815 $2,270 $2,397 $1,596 $1,854 Return on tcac (rcac) 2.8x 1.6x 1.8x 1.8x 5.8x Return on tcac (rcac) 4.3x 4.9x 3.1x 3.2x Look at both GMPP and rcac together to compare channel efficiency 15.0x 10

11 Applying the Framework A few rules of thumb help put the framework into action. GMPPs under 12 months are great, and are acceptable up to about 18 months. After 18 months, however, the present value of those cash flows way out in the future is harder to justify against the upfront acquisition expense. For example, a GMPP of 36 months suggests that three years are required to break even on the customer acquisition cost. Even with monthly churn as low as 3%, these customers will never recover the tcac, let alone make contributions towards the G&A and R&D expenses (lifetime is only 33.3 months against a 36 month GMPP), so a channel displaying this sub par performance should be starved. A GMPP of 12 months or less assumes tcac is repaid within a year. Assuming manageable churn, a company with such a short GMPP should be throwing fuel on the fire. For rcac benchmarks we like to see a return of at least 2.5x, with 3x or more preferred. rcacs below 2.5x don t leave much to cover operational expenses beyond the acquisition expense and recurring COGS keep in mind that an rcac of 1x is equivalent to breaking even on COGS and tcac. Low rcacs mean a company earns little over the life of each customer and new customers have to be added quickly just to replace ones that churn. Higher rcacs, on the other hand, provide more headroom to cover expenses and reinvestment. Ultimately, realizing multiples of tcac is how cloud based companies build enterprise value. Turning to our SaaS example, the Basic offering only yields one acquisition channel with passing economics CPC with a GMPP of 14.5 months and an rcac of 2.8x. As is often the case, the organic channel has superior economics, but it s hard to invest in this type of customer acquisition. If we look at the Premium offering, however, every single channel is producing an rcac of at least 3x with a GMPP of less than 13 months. Not considering the possibility that Basic customers are upgrading to the Premium offering, these results would suggest that we should be refocusing our resources on growing the Premium side of the business at the expense of the down market offering, with the possible exception of reinvesting in CPC for the Base product. * * * We hope entrepreneurs find this framework useful for analyzing the unit economics of their cloud based businesses. Cloud computing has exploded quickly, and we at Updata Partners are enthusiastic supporters of this transformational shift. The market is dynamic, and we expect this framework to be as well, so please contact us with any feedback or questions. 11