I started my career in enterprise software in the 1980s and after some years working in other areas (outsourcing and online media) I am back in the enterprise software game. This post is my reflection on what is different and what remains the same. I have focused this as advice to entrepreneurs building enterprise software ventures today.

1. Decide whether you want to be: a sailboat or power boat. Most enterprise software companies are self-funded sailboats. You know your direction - you are heading north. But you can tack left when the wind takes you that way and right with a little move on the tiller. You simply cannot go faster than the wind allows.

So it takes longer this way - think in terms of a decade to reach any scale that gets you credibility at the enterprise level. But you don't have any lords and masters to answer to and the pay out goes entirely to the founders and management team.

The big enterprise software companies almost all bootstrapped their way to profitability before they got their first external investors (typically via an IPO). On the other hand, the failure rate of VC funded enterprise software ventures is high. The power boat, loaded with gas, rockets out of the harbor but then runs out of gas. Mixing metaphors, the attempt to accelerate the scaling process is a bit like asking a baby to be incubated in six months (dangerous). But SaaS and freemium are changing the game a bit - as you'll see later in this post.

Guest author Bernard Lunn is a serial entrepreneur and business adviser to start-ups and is ReadWriteWeb's former Chief Operating Officer. He is currently Co-Founder of CapitalMarkets.com and Editor Of SaaSInsights. Bernard is based in New York, was born in Berlin, has lived in and started companies in Asia and Europe as well as America; he is comfortable with globalization. He started in online publishing before the Internet (with Prestel in 1980!). You can reach Bernard Lunn at bernard dot lunn at gmail dot com.

2. You may sell into enterprises from day one but you are not enterprise-scale until you reach enterprise scale and that takes time. Being enterprise scale means that your representatives have regular dialogue at CXO or at least VP level with Global 2000 accounts are they are bringing in millions of revenue from those accounts every year. Don't think that raising a $50 million round makes you enterprise-scale - it just gives you a s***tload of cash to burn. Your entry is likely to be a niche within the enterprise that the big dogs have ignored and that has suddenly become a critical issue to address. You have to be addressing a really nasty pain point.

In simple terms the customer's risk of doing nothing has to be higher than the risk of buying software from an unknown startup. That niche solution gives you your "land" in the account. Later, using internal references, you can expand into adjacent areas and that will eventually get you enterprisey type negotiations with the CIO's team. The good news is that a) enterprises are going through wrenching change today, so there are plenty of new pain points to address, and b) the big dog vendors are conservative and listen to the conservative people in their customer base, so they tend to leave plenty of time for new ventures to make their mark.

3. Learn the rule of 3. The sailboat startups don't invent a product out of pure imagination. They iterate relentlessly, keying off real customer needs. You have to do this; addressing real customer needs releases revenue dollars that you need to survive. But this discipline also means that the product fit to market tends to be very good - you don't end up with solutions looking for problems. But if you go too far in this direction, you end up as a custom software shop and that is way, way less valuable (and fun) than a product business.

The middle way is what I call the "rule of 3" or "once means nothing, twice is coincidence, three times is a trend." The first project means nothing; it may end up as nothing more than a bit of custom code (and some revenue). The second project is when you start to sense that maybe a real market exists. You talk to tons of potential customers to understand their emerging needs and you invest a bit of R&D in generalizing around the second customer's specific requirements. By the time you get to the third customer, you are ready to sell a productized solution with only minor customization. This rule of three does not only apply to the initial core product. It also applies to all the expansions of that core into adjacent areas that eventually get you to an enterprise solution.

4. Change in enterprise behavior takes way, way longer than most startups expect but when it comes it is swift, surprising and fundamental. The slow pace of change is what Ben Horowitz is writing about. My personal experience echoes that. But the change that is happening is very fundamental. This is not just about social media or consumerization of software. That is the surface change. This is about how behemoths adapt to Coase's Theorem now that the Internet is making that a practical consideration. (For more on this subject, here are some earlier posts).

The global economy is currently going through a "perfect storm" of change, so we can expect even more creative destruction than normal. The classic mistake is to over estimate change in the shortterm and under estimate change in the longterm. Enterprises are great at presenting a facade of normality, looking like nothing is changing, until that inflection point happens where everything changes very, very fast and fundamentally.

5. Founding teams must be balanced. You need both a hustler and techie. In a consumer Web venture, the hustler/sales guy is non-core (a few biz dev deals are important but not core). But in an enterprise software venture, nothing really happens until customers sign on the dotted line. So if you are techie, find a founding partner that is a hustler. If you are hustler, find a superb techie as founding partner. The technology has to be superb to get into the enterprise. The architecture has to deal with the constant pivots and directional changes to adapt to the market needs.

Photo by LotusHead

6. The transition from a founder who does all the sales to a scalable sales team is really, really hard and many ventures fail at this transition. It is equally true for the brilliant developer who has to migrate to becoming a leader of managers. Ventures fail at this transition for two reasons - too much process and too little process. Mark Suster writes very well about this trade-off/dilemma. For more on building "age appropriate process", see this post from StartUp 101.

7. The consumerization of software primarily impacts the UI, but don't mistake that for only a surface change. The new style of user interface - one click leads to another, no training needed, end users get results very, very quickly and that motivates them to learn the more complex functions - is fundamentally different from traditional enterprise software. You cannot use freemium strategies without this type of user experience (and freemium is the first fundamental change to the cost of customer acquisition). Nor should you think of the UI as a pretty layer on old software. That is just lipstick on a pig.

The architecture has to enable this type of user interface for all types of users, including developers, administrators, managers, partners and so on. These users require complex, feature-rich systems. But they also now demand the types of user interfaces they experience online with Google, Facebook and Twitter. This is where some foundational technology around adaptive case management, real time/event driven, open data/semantic web and loosely coupled web services are starting to have an impact.

8. SaaS reduces your sales cost but also means that VC is needed. SaaS transfers risk from the buyer to the seller. In the end, SaaS benefits both buyers and sellers. Sellers get a lower cost of sale/shorter sales cycle and better revenue visibility. But the enterprise software bootstrapping past was based on upfront license fees where the customer took the financing risk. So VCs will need to invest. Given the successful track record of SaaS ventures (as tracked by SaaS Insights), VCs are not holding back.

9. Freemium is the only new marketing tool that is proven to accelerate market entry since God invented sales people. Lincoln Murphy is the best source for specific freemium techniques, but the basics are easy to understand. You use a free version to entice the user to eventually try the paid version. If that free version also generates viral growth (it is communication related) and if the free users generate data that serves the paid users, you have a great formula for success.

You are trading the cost of push sales (expensive folks with Rolex watches, the cost never goes down) for the cost of system infrastructure (Amazon Web Services or equivalent, constantly falling in obeisance to Moore's Law). When you reach enterprise scale you will still need sales executives to negotiate with the CIO's team, but you will be doing that from a base of traction within the customer. Freemium helps at the critical market entry phase when ventures often fail. The case study to watch is Yammer.

10. Remember the other 80/20 rule: 80% of enterprise IT budgets just "keep the lights on;" Only 20% goes to new stuff. I learned this in the technology nuclear winter in 2002, when a 20% cut in IT budgets meant that no (zero, nada) new projects were approved. If you can show how to reduce that 80%, you get a better shot at the 20%. That 80% market is a replacement market. You need to know what cost you are replacing. The incumbents are looking at the 20% budget as well and they have the inside track. You have to attack the 80% to make it big.

The good news is that CIOs (and their bosses, the CFO and CEO) are increasingly asking why the company spends 80% on technology that does not improve competitive advantage. This has benefited outsourcing (if it is non-core, outsource it) but also enables disruptive technology to get a serious hearing in the corner office.