Customer-Centric Models

▶ Watch the original lecture

Everything comes back to designing customer-centric models: things that benefit the customer. The secret is to follow the customer journey. When a customer first engages, all they have is a problem, and they spend a lot of time and money figuring out what solutions are out there. Then they buy it, implement it, get trained on it, test it, roll it out, measure whether it is working, and optimize it, and only at some point do they decide they have got enough value that they should keep buying it (the three-up process). All of that is about increasing the dollars you take over the lifetime of your engagement: the more and the longer they engage, the higher your lifetime value, which makes your business model much better.

You will constantly answer questions for your customers: why is this better for them, does it save them money, does it save them time, what is the margin improvement in their business (not yours). Startup Secret: identify the one thing that makes your business profitable that your customer also gets profit or benefit from. That is the key to a great, sustainable business model, because it is not just you succeeding, it is your customer you are making successful. Think about Airbnb: they help homeowners utilize their home and get paid money they would not otherwise have got. Or Salesforce: people get better utilization out of their salespeople, who are more in touch with their customers and can sell more. Basic stuff, but this is what drives businesses and makes them sustainable.

One founder crystallized this: their customer is a low-income consumer in Kenya who currently cannot afford fresh coffee, and the customer benefits from increased reach through access, so the coffee becomes affordable and convenient. That is the whole point: do not just say “increased reach,” say why the customer benefits from it.

Gain/pain equals LCV/CAC (two views of the same principle). For those who were at the value proposition session, gain/pain has two views. The customer’s view is the gain they get from you versus the pain and cost of acquiring your solution. The vendor’s view (yours) is the life cycle value (LCV) versus the cost of acquiring and re-engaging the customer (CAC). When products have low cost of acquisition (low pain for the customer to try and buy) and high gain, and you can keep a relationship for a long time, you get very exciting, very profitable, very valuable business models. It is worth establishing early how you intend to finish, because if you start as you plan to finish, you can be very successful.

The LTV/CAC ratio, and its refinement. The basic startup metric is lifetime value of a customer (how much you get from a customer over the lifetime of working with them) divided by the cost of acquiring that customer. In general, if that ratio is not around 3x or greater, you probably do not have a profitable business model. But over the years I noticed something missing: almost every startup gets in at one point and then has a relationship to maintain, and most founders ignore that latter set of re-engagement. So I refined it: think in terms of life cycle value (LCV), the entire cycle in which you retain the customer, they re-trial, they repurchase, and you upsell them, and think in terms of the cost of acquiring and re-engaging them. Re-engagement costs are typically support, customer service, and, in a complex B2B product, professional services. Factor all of those in.

Cry-buy-fly-die, and extending the life cycle. People talk about product life cycles as a bell curve: customers see it, try it, buy it, run (fly) it, and then it dies. In software and many B2B businesses, if it really looks like that, you have a long, costly customer acquisition, a slow payback, and a very short lifetime value, and you are in trouble. How many of you downloaded an iPhone app recently and never used it again? That green-circle death is a big problem for mobile app developers and for almost every kind of software. The fix: break the simplistic see-try-buy-fly-die model into granular stages (awareness, interest, understanding, engage, try, purchase, and then the blue-circle stages of re-engagement and repurchase), and build in the means to re-engage the customer over and over (a new map for a GPS app, new content for a recommendation service, new data for a dating app). Use the web to get people to find you (SEO, SEM), open source so they can try you free and scratch their own itch, slippery products so they can adopt you at low cost, and on-demand SaaS subscriptions so they buy you as they use you. Each of these is a lever (taking cost out for easier see-try-buy) and a multiplier (extending the value and the life cycle). Summarized in three words: friction-free, rapid adoption, and extended life cycles.

What drives valuation (Goldman Sachs data). There is a wide discrepancy in the values companies get. Data we worked on with Goldman Sachs shows the median value for SaaS companies is about 5.2x revenues, but with a massive range from 1x to 13x (and higher). Demandware went from about 7x to 17x revenue over time. Growth is one reason (companies achieving 30 percent-plus revenue growth get high multiples, because it is hard to find anything on the stock market yielding like that), and it is why founders take venture money, to spend for growth and dominate a market. But two other metrics drive the multiple:

• Retention: companies with high retention (ServiceNow, NetSuite) have much higher multiples, because people can predict a business whose customers stay for a long time. Demandware topped the list at 97 percent retention: their e-commerce platform is so sticky that when a company builds its site and integrates all its back-end catalogs, they live with it for years.
• Upsell: a 2 percent uptick in upsell gives you roughly a 14x leverage in your valuation. Because you already have the customer and can sell to them so cheaply, almost all of that sale drops to the bottom line, compounding the value. This is the payoff of Russian doll packaging, and it is why people place such a premium on companies with multiple products and services they can upsell.

The P&L reality. At the earliest stage, almost all expense is R&D (about 100 percent-plus of revenue, so your loss equals your development spend). At a target software model, you aim for gross margin of at least 80 percent, R&D drops to about 15 percent, and (a surprise to most people) sales and marketing is typically about triple the R&D cost (around 40 to 45 percent). Your real cost in building a company is reaching customers, which is why business model matters: use levers to reduce that cost, or you never get to a bottom-line net margin of about 20 percent. G&A settles at about 8 to 10 percent (spiking during the IPO process). Acquia’s real numbers illustrate the journey: in 2010 they were still spending about 48 percent on R&D with cost of goods sold over 60 percent, and the long-term model targets gross margins around 75 percent (a little lower than a pure proprietary software company because of the cloud), R&D at 15 to 18 percent, sales and marketing at 35 to 40 percent, G&A around 10 percent, and an operating margin of about 15 to 20 percent.

Benchmarking (Lauren Kelly, OpexEngine). OpexEngine benchmarks companies’ financials and operating metrics. Their own disruptive business model: take in clean data, aggregate it, and give it back as good-quality benchmarks. Almost every industry is being disrupted; software was disrupted by software-as-a-service. In traditional software you sold perpetual licenses and started from zero every quarter, with salespeople controlling the pipeline and visibility. In SaaS, with long-term subscriptions and complex revenue analyses run by finance, the CFO (like Dennis) now tells the rest of the company what the business will do, with high accuracy, based on the number of subscriptions, the churn rate (you assume you keep 85 or 90 percent), and upsell percentages. Customer lifetime value is calculated as subscription revenue to a customer, multiplied by a lifetime (maybe 5 years or 60 months), multiplied by your renewal rate (say 85 percent), multiplied by your gross margin, compared to your cost of customer acquisition. A high CAC may not matter if lifetime value is $2 million, but if lifetime value is only $80,000 a year and CAC is $6,000, that is maybe not very profitable. And definitions matter: contracted monthly recurring revenue versus plain recurring revenue can make a $1 million business look like a $5 million business because of one-off deals or installation services. These terms are not legally or accounting-defined; they developed over the past 10 years, so one company’s definition can differ from another’s. In OpexEngine’s annual survey, there is a huge difference between venture-backed firms (with access to capital, chasing revenue growth, median around 200 percent growth, but unprofitable) and bootstrapped firms (which cannot spend over their revenue, so lower growth but profitable). Interestingly, the gross margin often ends up the same; how you get there is very different.

Demandware: the shared success model. Demandware is a public company at the confluence of three big markets (a $12 trillion retail market, e-commerce, and global SaaS), so it has a huge total addressable market. The problem it solves: people want to buy across many channels (web, mobile phone, tablet, point of sale, and increasingly TV and mobile point of sale) with one integrated experience, and it would be enormously expensive for each retailer or brand to build all of that. Demandware provides it as software as a service: at the core is an infrastructure more reliable than Amazon (less outage), handling hundreds of millions of customers and billions of dollars of commerce, but that is now table stakes. The key differentiators are what customers actually want to spend time on: merchandising and marketing that make brands like Gucci or L’Oreal stand out. So Demandware shifted its core from reliable delivery of commerce sites to making merchandisers more effective. The multiplier and lever that matter most: a shared success model. Instead of charging tens of millions for the infrastructure, Demandware charges nothing up front and takes a percentage of the revenue flowing through the platform. As the customer is more successful selling online and opening new channels, so is Demandware. It is a “why wouldn’t you” model, and because Demandware delivers the function (unlike models where the customer has to deliver it), customers readily agree. It literally changed the game and is the single biggest reason for the company’s success. On top, the link program is co-creation: rather than build ratings, reviews, recommendations, and back-end order-management integrations themselves, Demandware created an open platform so partners connect and build in their capabilities, giving customers a broader, pre-integrated whole product with faster time to value. Results: content growth rates over 50 percent, customers going live increasing dramatically, and average revenue per user nearly doubling, with a highly sticky subscription model and stable, high subscription gross margins, ending around a 25 percent drop to the bottom line. And self-service is part of slippery: Crocs went live in a dozen countries on their own without Demandware even touching the software, so the number of sites (and revenue, of which Demandware takes a share) grows exponentially, which is leverage in the business model.

A note on profitability and market share. You have a good business model when you are headed toward profitability, but many companies worth over a billion dollars at IPO are not profitable, and investors are okay with that, because they understand the business model can be switched to profitability at any time (by slowing growth or stopping investment in new markets). With an RSVPD-type strategy that people trust will lead to profitability, they will continue to let you invest and build a leading position. And you build valuable businesses by doing the right things: deliver value to your customers, manage the right business model, and value follows.