Architect Your Business Model

▶ Watch the original lecture

Now we architect to attract those customers. You can be thinking about the business model from day one; it is not something that comes later. And the way I explain it in simple terms is: take all the friction out of it, and reduce the cost of all that sales and marketing spend by how you build the product and how you take friction out of the go-to-market process.

Pricing matters, and it can be architected. How you price your product is super important to that early friction. If there is a way to provide free samples, make your platform free, or create developer kits that cost nothing, people can start using your product, and then you have a way to step them through increasing value at increasing price.

Be careful with free, though. People often equate the value of a product with what you charge for it. If something is free forever, they start to think of it as valueless, and the perceived value is often zero. There is a wonderful older article by the editor at Wired about “free” economics, how free was the new black; it is a useful piece, but free is not the be-all and end-all. “Free fall” is what happens to startups that never figure out how to get beyond it: they reach customers, but those people value it at what they pay, which is nothing. Generally, a free trial period that emerges into something people pay for is preferable to going fully free forever, and if there is a free version you are constantly upselling. LinkedIn did this brilliantly: it started as a place to put an online version of your resume, quickly became a social network, got all the network effects with a free product, then piled on premium products (Sales Navigator, InMail, extra InMail messages, LinkedIn Premium). For you and me the value is the connections; they monetize the recruiters and head hunters who tap that network. Figure out early where you are going to monetize. If you go the free route, give real value so people engage (there has to be a virtuous circle), and if you are smart, build in virality: something that makes people take an immediate action and share it quickly when they see the value.

Time to value. In the enterprise, if you can get time to value under three months, that tends to be great, because it makes it easy for people to build cost models and see when payback begins. People do not expect instant gratification in the enterprise, but they expect rapid payback: three months ideal, certainly within six to nine, and not greater than twelve. If it takes longer than twelve months to deliver value, you are in trouble, because you are going beyond people’s budgeting cycles. Get time to value as short as possible.

The SolidWorks story: architecting a business model on purpose. [Told by John, a founding-team member and later CEO] SolidWorks is a 3D CAD company based in Waltham, today about a $600 million company generating close to a quarter of a billion dollars of profit a year. I joined when we had our very first sale. Let me give you the simple reason SolidWorks existed. People did mechanical design with lines and arcs; AutoCAD moved that onto a PC. But people wanted to see things in 3D. Big companies like Computer Vision let you build 3D solid models, and the biggest benefit was understanding how parts fit together without interference; car and airplane makers spent tens of millions on these systems. Then PTC (parametric technology) did something phenomenal: a brilliant guy named Sam Geisberg let those solid models change, so engineers could make rapid design changes and see them iterate, much like a spreadsheet. That was value, and PTC made a fortune with amazing technology and an amazing sales team. So when we started SolidWorks, people said nobody needs another solid modeler, PTC owns the market. But there was a ton of people with their noses up against the glass who wanted what PTC had and could not afford it: too expensive, too hard to learn, too hard to use. That was the value we created.

The question was how to capitalize on it. Four numbers: 1503. If you go to SolidWorks today and ask anyone, it is part of the DNA. Instead of selling for $20,000-$30,000 through a direct sales force, we sold at $4,000 through a VAR (value-added reseller) channel: independent people who represented our product, got it at a discount, and sold it. At our first sales meeting, about 250 independent businesspeople came to Waltham, and we had to tell them where to focus. We put up a chart: vertical axis, number of seats (engineers); horizontal axis, sales cycle. Most of them wanted to sell 20, 30, 40 seats. A guy from Detroit said, “I have a friend inside Ford who runs powertrain, I want to go sell them, will you come out?” I said I hoped he would not mind if we put two other resellers in his backyard to focus on the opportunities he would not be focusing on. Because here is what happens with the big deal: they bring you in, everyone loves it, powertrain says styling wants to see it too, come back next week, and the two-to-three-month sales cycle stretches to four, five, six months. Then they say, “We love it, let’s do a pilot,” and instead of 60 seats they buy three. So what we wanted people to focus on was one to five seats, zero to three months. That became the 1503 account.

Why so important? Those VARs were massively undercapitalized, barely making payroll, and if they chased the 60-seat deal, they were losing money the whole time it stretched out, only to get a PO for three or four seats. We wanted quick hits. It served our strategic interest to get into a lot of accounts, like weeds popping up through the concrete. From an expected-value view, the big deal looks better, but those VAR salespeople were making $65,000-$70,000 a year, not $150,000-$200,000; we needed them to get quick hits, get into the account, and, most important, qualify accounts out. The perfect opportunity was not a 10-seat Pro/ENGINEER account we would try to convert, and not 10 seats of 2D AutoCAD users we would have to teach 3D from scratch only to sell one or two seats. We called it the modeling saturation index: we wanted an account with two or three seats of Pro/E (already exposed to and sold on 3D) but eight seats of AutoCAD, so they knew the benefits of 3D but had not moved the rest over, probably because it was too hard to learn, too hard to use, and too expensive.

Why a VAR channel and not web or affiliate sales? When we started SolidWorks, the internet essentially did not exist for this; we actually debated whether to even have a website (“What would we put on it? Directions? Maybe someday people will download software updates,” and we all cracked up laughing, and then we were one of the first 100,000 websites registered in the world). We needed the VARs: independent businesses already selling other products, consulting, and training, who needed to survive and wanted to expand their customer base. It was critical to get customers to try the product, use it, and succeed with it, because we knew there would be viral adoption once people started shipping files around.

How did we know our 1503? We were lucky and got it roughly right early, but we knew the tendency was for people to pull us up into bigger accounts, and with a $4,000 price and a VAR channel we could not afford a direct sales force. So we intentionally forced them to go smaller, and kept the message incredibly simple: focus and repetition. The predictability of revenue was incredible. Our first-year revenue plan was about $3.8 million, we upped it to $4 million, and we ended up doing something like $12.8 million. The philosophy went right down into operations: take an order, process it, ship the product, and be profitable at a $2,400 net price. When we shipped upgrade boxes, they weighed 15.9 ounces, because at 16 ounces UPS charged a higher freight weight. That became the cultural mentality: get inside the account, land and expand. Did that mean we only went after small companies? No. As we grew, we ended up displacing Pro/ENGINEER at EMC, several hundred seats, three or four seats at a time, until at about 100 seats they had to deal with us.

Pricing methodology: align every step. We had a subscription-service program that is today a $300 million business. How did we price? You can build price on what it costs, or figure out what the market will pay. The installed base charged list price plus 18% maintenance (a shrink-wrap model), so we knew we would charge at least 18%. But we had to align not just our interests but our VARs’, because we needed them to support customers. So we started from the end user: what would a customer pay, how many customers could a VAR support on calls, how much money did the VAR need to make, and what would our margin be? We worked from the end user back to the middle and back again. The point: it is not just about the money you want to receive; you have to align each step in the process.

Bundling to win a price war. With analytics (a performance-monitoring tool, a little nugget of software on each PC sending back aggregate data), we could see which version people ran and which add-in applications they used. Plotting customers against number of applications used, it was a massive ski slope: hardly anyone used more than two applications with SolidWorks, and the popular ones were partner applications, not ours. Meanwhile Autodesk was undercutting us by bundling their 3D solution with AutoCAD 2D. We realized the value of these partner applications was not being used, partly because it was not worth a VAR’s hassle to go back and sell a $400-$500 add-on. So we bundled groups of them into SolidWorks Office Professional and Office Premium. In the middle of a price war, we took our price from $4,000 for the base product to $5,400 and $7,500, and increased subscription along the way. To the application partners I said, “You’re only making $15,000-$20,000 from us; give me a blanket license and we’ll pay you $100K.” We raised our average selling price by over 50% in the middle of a price war and grew volume tremendously. And it was defensible: once we had a higher subscription revenue base, VAR owners were waking up on January 1st with a $1.5-$2 million subscription business they would not walk away from, so it became a barrier against competitors trying to steal our channel. That all came from analytics we introduced to solve a problem, and only later realized was a huge business enabler.

Other hard-won lessons.

• Meet your enemy: time. In a startup the only advantage you have is time, the calendar. The installed-base incumbents have customers, capital, presence, and a megaphone. You have to move quickly.
• Events force actions. Why did you pay your taxes on April 15th? Because the government created an event that makes it painful not to. Inside your company, put deadlines and events in place. We ran SolidWorks World, 5,000 people, once a year, at a cost of about $2 million. Even if you strip away all the media exposure and the shaved-head zealots who go tell everyone about you, forcing people to get together on one day and align and make decisions was hugely worth it.
• The perfect is the enemy of the good. Get your product done, get it out, and iterate.
• What you think versus what you know. Everybody says beta test, but if people are not paying you, they will tell you they love it all day long. At CloudSwitch tons of people loved it, but when they finally paid, we realized they were paying for something different (education), not the product.
• Hire well. Hire people who scare you with their confidence. On salespeople: if a salesperson says they do not really care about money, they are not a good salesperson, fire them. Salespeople are coin-operated; reach behind their back and look for the coin slot. The other thing that motivates them is recognition: if a salesperson is not concerned with how they rank versus others, get rid of them. They are naturally aggressive and competitive; those are the two drivers.
• Culture and alignment. 1503 was a culture as much as a strategy: at every step we made money by being efficient, and we cared about customers whose names we did not even know. We built a company on the backs of suppliers to the Nikes of the world, mold shops, not the logos, and we were proud of them. We aligned the company consistently across product, pricing, packaging, and culture.
• It is easy to start a company; it is hard to build a business. And remember it really is about the journey, which is why most entrepreneurs do it: they see the world and how it should be.

The one thing to take away: SolidWorks got scale because the go-to-market methodology was reduced to something so simple that even a $60,000 VAR salesperson could carry the message and execute on it repeatably. It had nothing to do with the technology in the end, although the technology was critical at the beginning. Architect that road map beyond just the product and technology.