But today it happened. With all the turmoil in the stock markets, Apple is now the most valuable company in the world, $500M ahead of Exxon when I last looked (so it could go either way in the short term).

The thing Apple seems to do better than anyone else is make profit. It has a small share of the overall mobile phone market, for example (measured by units or revenue) but it makes over 2/3 of all the profit. And I’m not talking about the smartphone market, the entire mobile handset market.

Almost everything they do has a lot of serious critics but it ha all paid off. Retail stores? So 20th century when everything is online. Design your own chips? You should outsource everything you can to experts. Enter the phone market? It’s just a handset (quote from the then CEO of Nokia)

Then Wally told me,  “I immediately sent a copy to Carl Icahn.” So the EDAgraffiti book is up there with poison pills and the rest as an anti-takeover device!

To get your own copy, just click on “book” at the top of the page.

So obviously cash is so important in a startup that it should never be spent? Well, not so fast. I’ve seen some really silly decisions about how to save money in startups over the years.

Most of the cash being burned in a software startup goes on engineers’ salaries. Consequently it makes sense to do everything to make their work environment as productive as possible. Do not force them to use old computers because they are already around. Computers are pretty cheap these days, a few days of an engineer’s salary will buy you something really high end. Do not equip the servers with so little disk space that they have to delete old data that will eventually turn out to be useful. Terabyte disk drives are under a hundred dollars. And don’t forget to make it easy for your engineers to work from home, by having good VPN and paying for them to have good internet connections. You pay your engineers more in an hour than their internet connection costs for a month.

One thing we discovered at Envis was that companies that provide PCs for gamers deliver the most bang for the buck. They overclock the designs, add special fast memory, have custom motherboards and so on. For a lot less than Dell will sell you a machine, you can get one that is half as fast again, with lots of cores. And a bonus, they look really cool.

Do not hire a consultant and then, for egalitarian reasons, give them a day on which it is their job to clean the kitchen at $150/hour. In fact, don’t make your engineers clean the kitchen anyway. That’s pretty pricy labor. And don’t annoy your engineers by charging for coffee or soda.

In a semiconductor company, there is an additional significant cost, namely design tools. Sometimes this is a cost in a software company too since they need tools for quality assurance and integration purposes. This is a hard balance to get right since too few tools again means that what looks like saving money on tools is really burning extra money on engineers’ salaries. Too many tools obviously wastes money more directly. When I was at Cadence we had a venture investment program where we would provide almost unlimited tools to startups for a mixture of cash and an equity position. We’d discovered that most startups underinvested in tools because they were so expensive but that this jeopardized their success.

Benefits, especially medical, are another area where startups can be penny wise and pound foolish. The most cost effective way to handle medical benefits, given that usually everyone is young and fit, is a combination of catastrophic coverage and a health savings account (HSA). In fact this is probably the best way to handle medical period, but that’s a political hot potato right now. It is what Whole Foods does and what John Mackey, the CEO, has recently got into trouble with the left for recommending as better than what congress is attempting to put together.

Bottom line: remember Gordon Bell’s line that cash is more important than your mother. But remember that engineers’  salaries are your biggest investment, and it is foolish not to do everything to make that investment as effective as possible.

Most of these problems are not directly with the license key manager (the most common, almost universal, one is FlexLM). Sometimes there are direct issues because customers want to run a single license server for all the EDA tools they have from all their vendors, something that the individual EDA companies have a hard time testing since they don’t have access to everyone else’s tools. More often license problems are simply because licenses are much more complicated than most people realize.

All sorts of license problems can occur, but here is a typical one. The customer wants some capability and discusses with the salesperson who provides a quote for a particular configuration. Eventually an order gets placed and a license key is cut for that configuration. At this point, and only then, it turns out that the configuration doesn’t actually deliver the capability that the customer thought he’d asked for, and that the salesperson thought she’d provided. Something is missing. The customer calls support to either to report a bug or, if they realize what is going on, to try and get the specific license added. Often an option has been omitted from the configuration (such as a special parser) that everyone assumed was included, or assumed that it wasn’t needed, or that turned out to be bundled with some other capability in a mysterious way.

Digital Equipment, in the heyday of the Vax, actually had an AI program XCON salespeople had to use to configure Vax computers since otherwise they always had similar problems, although in the hardware domain. The order omitted a required cable, or overloaded a power supply or left out a software driver. Without this error being corrected, the delivered system could not be assembled in a way that would run. This is worse still in the hardware world since it takes from a couple of days to a couple of weeks to get a missing cable to the customer site. It can’t simply be fixed over the phone.

The fundamental problem is that it is hard to map capabilities that marketing wants to sell and price, into the actual control points in the software that permit or deny certain activities, and the ways in which the different components interact. Few people have a good understanding of this, and there is no correct answer to many of the questions.

Here’s an example. Should a long-running tool claim a license when it starts for an optional feature that might be required later? Or should it wait until it has run for hours and then fail if a license is not then available? Which inconveniences the user less? There are pressures on the vendor side to want to claim licenses as early as possible (so the customer needs to buy more licenses) which at least means that if a tool is going to fail due to lack of licenses, it does so immediately without having done a lot of wasted work, and in a part of the code where it is easy to handle. There are pressures from the customer side to want to claim licenses as late as possible (so they don’t get held for long periods when they are not being truly used) but also to expect that the tool will behave gracefully when their paucity of licenses comes to light and the run is deep in the innards of the tool when it finds out it cannot continue.

Interactive tools are worse still. Do you claim a license in order to show the capability on a menu? Or do you show a menu item that may fail due to lack of a license when you click it? Do you behave the same if the customer has licenses but all are currently in use, versus the customer not having any licenses to that product at all?

None of these problems typically affect the engineers developing the product or their AEs. Usually all employees have a “run anything” license. The licenses issues often only come to light when customers run into problems. After all, they may be the only site in the world running that particular configuration. Some testing can be done easily, but exhaustive testing is obviously impossible.

EDA companies want to create incremental revenue for new capabilities, so they don’t want to simply give them to all existing customers even though they may want to make sure that all new customers are “up to date.” This drives an explosion of license options that sometimes interact in ways that nobody has thought of.

Until some poor engineer, in the middle of the night, tries to simulate a design containing two ARM processors. That’s when they discover that nobody thought about whether two ARM simulations should require two licenses or one. The code claims another license every time an ARM model is loaded, in effect it says two. Marketing hadn’t considered the issue. Sales assured the customer that one license would be enough without asking anyone. Nobody had ever tried it before. “Hello, support?”

My experience is that you need to be careful how to structure royalty deals to have any hope of that happening. At VLSI I remember we once had a deal to build a chip for one of the Japanese game companies if we could do it really fast (we were good at that sort of thing). It needed lots of IP so we just signed everyone’s deal with no license fees for as long as possible, but which all had ridiculous royalty rates. We probably had a total of about 25% royalty on the part, more than our margin. But we reasoned as follows: “One in three, the project gets canceled and never goes into production (it did); one in three it goes into production, but we never ship enough volume to reach the point we pay more in royalties than we would in license fees; one in three it is a big success and we tell everyone the royalty they are going to get, and if they don’t accept, we design them out.”

IP is more mature now, so the royalty rates and contracts are more realistic. Back then the lawyers could consume the whole design cycle negotiating terms and there wasn’t enough time to wait. Everyone thought their non-differentiated IP was worth a big royalty of a few percent, even though a big SoC (even then) might have dozens of IP blocks that size. So perhaps the problem has gone away. If you were on a short time to market, you simply didn’t have time to negotiate terms or royalty rates. You had to get going. Hence our strategy of accepting everything on the basis we’d renegotiate if it became important.

The best form of royalty is one that is paid out elsewhere in the value chain. If TSMC pays a royalty to Artisan (now ARM) or Blaze (now Tela) then the customer ends up paying since TSMC adds it to their bill. But if the chip is a success then the customer doesn’t get to re-negotiate from a position of strength. If the overall deal is a huge success then TSMC has probably negotiated percentage breaks anyway (I don’t know any details so I’m guessing) and can choose either to reduce the cost it charges on or take a bigger profit.

However, when Artisan was bought by ARM for $900M every VC wanted a deal like that. And they saw royalties as the secret sauce. But royalties only work really well when they are much higher with unexpected success, otherwise they are just moving payments around in time. Plus they only work well with unexpected success if they can’t be renegotiated as a result.

The reality is that they are often disappointing. Mike Muller, CTO of ARM once told me that “royalties are always later and less than you expect.” It took Artisan 15 years to get to the stage it was sold to ARM, those royalties were largely licenses fees foregone in the early years. ARM was, in effect, paying for money pushed into the future that it would then collect.

At one level, many people say that it is a pity that EDA doesn’t get a percentage of semiconductor revenue. But in a way, they do. EDA has been around 1.5-2% of semiconductor revenue for years. Of course EDA would like that number to be 4% but it’s unlikely that semiconductor companies would have signed up for the deal of no upfront money and big royalties, even though they would probably have been served well by it. They would have avoided any of the business impact that many companies suffer due to having inadequate numbers of licenses.

When you put together a spreadsheet to show how the bookings will build up, there are two factors to which the amount of money turns out to be very sensitive:

• How long after you hire a salesperson before they start to produce revenue?
• How fast does a salesperson ramp up to “full power”?

The answer to these two questions governs how much you need to invest in a sales team before they are a net positive for the company, and the total and timing of that investment governs how big a hole you have to cover and thus how much money you need to raise.

You might think that how big you plan to get is a critical variable, but in fact the answer is about a $50M run-rate after 5 years. If you don’t have a plan like this then nobody will fund you (they probably won’t anyway at present, but let’s leave that to one side). You almost certainly won’t grow that fast, and everyone knows it, but they will “take a haircut” to any numbers you give them anyway, so you’d better play along and give them big ones.

Other assumptions you’d better bake in. Any bookings you have will come in the last week of the quarter. This means that any cash associated with that order will not come until the following quarter. So every quarter, for every sales team, you need to pay all their salaries without the cash from the business they are generating that quarter to offset those expenses.

Each salesperson needs two application engineers to be effective. Or at least 1½. This means that a sales team costs approximately $800K per year in salaries, travel and so on, which is $200K per quarter, perhaps a little less if you don’t have the full 2 AEs per salesperson.

As for sales productivity, at capacity a sales team brings in $2M/year. If you put in much more than this then you are simply being unrealistic. If you put in much less you’ll find that the business never gets cash-flow positive.

EDA tends to have a 6 month sales cycle. So normally  a new salesperson won’t close business in less than 6 months, and probably 9 months (3 months to understand the product and set up initial meetings, 6 months of sales cycle). I like to use a ramp of $0, $0, $250, $250, $500 for the first 5 quarters, which assumes a salesperson sells nothing for two quarters, is at half speed for two quarters and then hits the full $2M/year rate. Later this may be conservative since a new salesperson can inherit some funnel from other existing salespeople in the same territory and so hit the ground if not running then at least not at a standing start. In the early days it might be optimistic since I’ve assumed that the product really is ready for sale and it is just a case of adding sales teams. But realistically it probably isn’t.

So those are the variables. In 5 years you need to be at $50M which means about 25 sales teams at the end of year 4 (because only those sales teams really bring in revenue in year 5). Some may be through distribution, especially in Asia, but it turns out not to make all that much difference to the numbers.

In the meantime, the rest of the company has to be paid for and don’t directly bring in revenue. So if you ramp sales too slowly, the rest of the company will burn more money in the meantime. This makes the model less sensitive than you would expect to the rate at which you hire sales people, within reason.

If you hire people too fast on day one, then the hole gets huge before your first teams start to bring in any money to cover the cost of the later guys. You need to get to about $7M of bookings before things get a bit easier and the early salespeople are bringing in enough to cover the costs of the rest of the company. However, if you bring in people too slowly then you will not get to a high enough number in the out years. The trick is to hire in a very measured way early and then accelerate hiring later. This will give a hole of about $4-5M meaning you should raise about $6M to give yourself some cushion to cover all the inevitable delays.

The R&D team should be about 10 people. It will be less in the early days but it shouldn’t really be more unless the company truly must develop a range of products in parallel. With more than 10 people, engineering will be off developing a range of products even if that isn’t the plan!

With a CEO and another “person” in the form of an accountant, an office manager, a little marketing (they may be one person or more likely a few people part-time) that makes a total of 12-13 people, which is a fixed cost of around $2.5-3M per year. A single sales team is around $800K-1M per year as we saw earlier. With that headcount in place it takes about $3.5-4M to break even.

But a sales team only brings in $2M so $3.5M is more than one sales team can bring in so we need a second, at another $800K-1M pushing the breakeven up to $4.5M. This is just about doable but more likely a third sales team will be required, pushing revenue to $6M and expenses to $5.5M. Add in some inefficiencies in training salespeople, filling the funnel and the rule of thumb is that you need to get to about $7M to become cash-flow netural and the company start to be able to fund its own growth, albeit slowly.

But breakeven isn’t the end goal, being profitable enough to have options is. Then we can be acquired, or continue to grow the business or even just pay our shareholders nicely out of the profits. This means getting the business up to about $10-11M, which means about 5 sales teams. The 5 sales teams will cost about $4-5M, leaving $6M. That will pay the $3M original (non-sales) fixed cost with $1M for some additions to the corporate team: a marketing person, maybe some non-bag-carrying sales management, and after a couple of years somebody might want a pay raise. That leaves $2M to either take as profit or use to fund further growth, start a second product and so on.

All of this makes one big assumption. That the product is really ready at the point that the channel expenses are ramped up. It assumes that each salesperson rapidly makes it to the $2M per sales team level. This is where companies die though. If the sales teams are added too early then they will burn all the cash. If the product is not ready for the mainstream then the sales guys will not make it to the $2M level and burn all the cash. But if everything is in place then the company can get to $10M rapidly. The first year I was at Ambit we did $840K in revenue; the second year, $10.4M.

This is the point at which a company is very attractive for acquisition. It has traction in the market ($10M in sales and growing), the technology is proven (people are using lots of it; look, $10M in sales), the acquisition price hasn’t got too expensive yet (only $10M in sales), it is probably the market leader in its niche ($10M in sales and growing). Of  course if the company continues to grow it will typically take in more investment at this point in order to grow even faster. Value of a software company is some multiple of forward earnings, and the greater the growth the greater the multiple.

But the biggest problem, especially for startups, comes about from section 302 and 404. Section 302 says that companies must have internal financial controls, that the management of the company must have evaluated them in the previous 90 days. Section 404 says that management and the auditors must report on the adequacy and effectiveness of the internal controls.

In practice this means that auditors must repeatedly go over every minute piece of data, such as every cell in a spreadsheet, every line on every invoice, before they can sign off. For a small company, the audit fees for doing this are a minimum of $3M per year. For larger companies the amount grows, of course, but slowly so that it is much less burdensome for a large established company (where it might be 0.06% revenue) than for a small one.

Only public companies are required to comply with SOX so you could argue that it doesn’t matter that much for a small venture funded startup. At one level that is true. But it has also meant that a company has to be much larger to go public.

In the past, an EDA company with $20M in revenue and $3M in profit (with good growth) could go public. But now a private company like that must comply with SOX to go public, so that $3M cost suddenly hits and the company has $20M in revenue and no profit at all. It must wait until it is, perhaps, $80M in revenue with $12M in profit (which would have been $15M without SOX). In EDA, in particular, this is extremely difficult to achieve with a single product since most sub-markets are not that large. In effect, EDA companies can no longer go public.

This applies to many venture-backed startups in whatever industry. Since the introduction of SOX most IPOs have taken place in either London or Hong-Kong. It is controversial just how much of that is directly due to SOX but clear that a lot of companies that could have gone public in the past in the US have not done so, and as a result have been acquired for lower valuations that would otherwise have been the case.

In the early 2000s (SOX was passed in 2002) the stock market was not friendly to IPOs as it recovered from the downturn. But venture backed IPOs in 2005 and 2006 were way below what they were in the 1990s, and Q2 2008 was the first quarter in 30 years in which no venture-backed IPOs took place in the US. This has been another reason that VCs are rarely willing to invest in EDA companies.

There are two big problems with a large expensive fixed cost. One is just that it is expensive and so it ties up a lot of capital (or a lot of expense budget for a “fixed” cost like employees) for which there may well be more profitable uses. Second, the fixed cost usually puts in place a fixed capacity of some sort, and that capacity risks always being either more than the market need is, or less than the market need is.

TSMC makes money as a foundry, of course (well, maybe not right now). It’s scale is enormous so it may well be able to make money selling wafers for the same price as you can get wafers out of your own fab, even if you have one running at capacity. But that’s the point. Your fab is never running at capacity. It is either below capacity, in which case wafers cost more than the “standard price” because all that depreciation needs to be spread over fewer wafers. Or else it is above capacity, meaning that there are wafers that you could sell profitably that are not being built (if your planning is poor, you may even have orders for them, and commitment dates that you are going to miss). Even if you pay a price higher than your standard price for wafers, it is worth a lot to avoid having to absorb fab variances when the fab is not full, and to gain the capability to sell more than capacity when you have a strong order book.

In the web space, you no longer need to build your own high-capacity server farm. Amazon, Google and others will sell you server time and disk space on a purely variable cost basis. If you website becomes a big hit then scaling should be much more straightforward.

In some ways you can look at Amazon S3 or TSMC as companies that are in the business of making the up-front investment in fixed cost assets and then charging you a variable cost to use them. Lots of other companies do the same. It doesn’t cost an airline anything (well, not much) extra to fly an extra passenger; it is basically in the job of taking airplanes (fixed cost) and working out good business models to sell trips (variable cost). Cell-phone companies largely have a network of base-stations (fixed cost) and work out how to charge each customer for using them (variable cost). It’s not always obvious what the best model is for making the cost variable: do you charge data per megabyte, or unlimited data for a month? How does the money get split when you are roaming on other people’s networks? Is data the same price as the digitized data underlying a voice-call?

When supply chains disaggregate, usually one thing that happens is that non-core areas, especially ones involving fixed costs such as equipment or full-time employees, are divested. New companies spring up to specialize in providing that non-core activity as their core competence. Ross Perot made his fortune at EDS taking companies’ IT departments off their hands and created a big specialist company to provide those services. Semiconductor companies get rid of their EDA groups and an EDA industry comes into existence (Cadence, Synopsys, Mentor etc). Semiconductor companies get rid of some of their fabs and a foundry industry comes into existence (TSMC, UMC, Chartered etc). Semiconductor companies get rid of their technology development (TD) groups and rely on the foundry industry for that too. One interesting area of debate right now is whether design is next, and how much of design. Nokia already moved its chip development group into ST. eSilicon, according to Jack Harding its CEO, is doing very well. Faraday is (or at least was) doing about 200 designs a year.

When semiconductor companies design chips about as often as they reconfigure buildings, does it make any more sense to have their own not-very-expert employees floor-planning their chips than their own building architects floor-planning their offices.