In the forum thread about Altium’s move to China, I was asked to comment on the “negative” aspects of the EDA space in general, and how that relates to Altium’s situation. We certainly don’t want to come off as “Pollyanna Engineering Journal,” so here goes.
First, as to why we write so frequently about Altium – it is because they make a lot of aggressive strategy moves. Aggressive moves – whether they’re good or bad and whether they work or not – make much more exciting stories than long slow decay due to inherently flawed business models, so I write about them. However, I think my ideas about the EDA industry are pretty well documented, and they’re certainly not all sunshine and roses.
I have little hope for the long-term viability of the EDA industry. That makes me sad, because I do think that EDA brings huge value to the electronics industry, and without independent EDA suppliers critical progress in tool development could slow to a crawl, crippling all of electronics.
The large EDA companies of today – Synopsys, Mentor, Cadence – all exist because there was economy of scale in developing tools for ASIC design in the 1980s and 1990s. One EDA company could develop sophisticated tools that would be applicable across a wide range of ASIC vendors. A design team could buy those tools and design chips across all those ASIC vendors. Independent EDA could amortize the development costs of those sophisticated tools across many customers of many ASIC vendors and could therefore do a better job with tools than the ASIC vendors’ own internal CAD development groups. Internal CAD groups at ASIC companies largely wilted and died, and commercial independent EDA flourished.
Unfortunately, EDA built itself into a corner with a business model optimized for that situation. The big money came from the biggest customers – the 25 or so largest electronics system houses in the world. Therefore, EDA companies invented themselves around pleasing those key customers – who accounted for the majority of their revenue. On the sales side, this meant assigning a major account manager to each one of these cash pipelines. The job of the major account manager was… golf. Lots of it. EDA account managers were just about the only professional golfers on the planet who were paid to play badly. (You always want the customer to win.) Once every three years, the major account manager and the corporate CAD executive from the big systems house would come in from the golf course and re-negotiate the three-year, all-you-can-eat (meaning as many licenses as you need of any of the products our company has), term license deal. The negotiation probably went something like this:
EDA guy: OK, We think we need $15 million per year for the next 3 years.
Corporate CAD guy: Make it $10 million and you’ve got a deal.
EDA guy: How about 13?
Corporate CAD guy: How about 13 if you win the next round of golf, and 12 if I win?
EDA guy: Let’s hit the links!
…three hours later, the EDA guy – who _always_ lost the golf matches, “miraculously” pulled out a win. Instead of insisting on the 13M, though, he generously splits the difference at 12.5. You don’t want to pull off a million-dollar golf hustle. People could get wise.
A week later, the press release was issued, and it looked something like this:
“Cadenopsys Graphics signs $37M 3-year deal with Intemotonokisson to supply design tools for next-generation anadigital celltabletower development.”
The EDA guy was a hero because he just brought in a $37M deal. The Corporate CAD guy was a hero because he just got the whole company unlimited design tools and huge amounts of FAE support at a screaming deal with one stroke of the pen – freeing all the company’s engineers to spend their time engineering instead of evaluating and choosing EDA tools. Win, win. Both companies’ heroes are given big bonuses and sent back to the golf course for another three years.
Small customers lose because their needs are not addressed at all by this process. EDA account managers have no time for them, they’ve got to keep the corporate CAD guy happy and looking good for the next big negotiation. Back in the EDA company factory, EDA product marketing teams are working hard to be sure the major account customers are happy, and that their product is NOT the one that makes the customer mad and sours the next big deal.
Product evolution loses because the critical feedback link between product development and customer use and satisfaction is broken – or even inverted. Every product engineering team at the EDA company is now confused because they just sold … an unknown number of seats for an unknown share of revenue to some engineers at a huge company – who might or might not ever actually use their particular tool. For some teams, their tool will be one of the ones actually used by those engineers. However, they’ll probably find this out when they get raked over the coals because their tool has bugs or missing features that are making the BIG customer unhappy. These teams will get the feedback that they’re doing everything wrong. For other teams, even though their tool is part of the big deal, the customer never even fires it up. These teams get no heat or complaints – just their share of the big revenue.
Tool pricing loses all touch with reality as well. Since the EDA company gets one big check, EDA product teams have no real idea what share of that revenue belongs to them. The EDA company management will somewhat arbitrarily apportion the money to various product lines based on a number of possible criteria. For example, if your Widget-Verification Tool (WVT) is listed as #2 in the market by Dataquest or by Gary Smith, and you’d really like to be #1, you may want to count a big chunk of this year’s mega-deal as sales of WVTs so that you can become the industry’s #1 supplier of WVTs. The irony, of course, is that this customer might end up never even using your WVT at all. However, back at the EDA company the WVT team are now heroes – they’ve hit #1, they get big bonuses, bigger budgets, more reqs to hire new engineers, and clearly whatever they’re doing down there in WVT engineering is working well. Go WVT Team!
If you’re at the same EDA company, and you’re working on a new, leading-edge product, you may not fare as well. The golf-playing account management team doesn’t want your leading-edge product anywhere NEAR their big-deal customer. New products have lots of bugs, require tons of FAE support, and often make the customer unhappy, even if they are providing huge benefits or essential services to that customer. As account manager, you’d much rather your big customer buy their leading-edge, risky tools from start-ups and vent all their leading-edge rage at them, while writing the big checks for the big-‘ol stable happy non-controversial tools to you between rounds of golf.
What does this have to do with a company like Altium? For all I know, Altium may suck at meeting their customers’ needs. Browsing their forums might easily give one that feeling – particularly lately. On the other hand, I’d get the same idea about Apple from browsing their forums, or Microsoft, or just about any company that has open forums for users to vent their issues and opinions on products. However, Altium is one of the very few EDA companies actually trying to serve the ~90% of the electronic design market that is NOT one of the largest 25 system houses in the world doing big-budget custom SoC design.
Now, you may be saying “Hey, the big 3 EDA companies care about the small customer too – that’s why they have third-party distribution channels. We can discuss that separately if you want. Or, if you’ve bought EDA tools through third-party distributors – tell us how your experience was.
As we all can see – Altium is doing a lot of stuff to try to make money in the business of serving that other 90%. It appears that none of that is working very well so far, as the company certainly isn’t reporting any big profits. They are, however, rapidly experimenting with their business model to try to get traction in this difficult market. I can imagine that much of this experimentation is extremely frustrating for existing customers. I certainly hate it when I buy a product and the price drops shortly thereafter. I hate it when I negotiate a deal for a serious product and it feels like I’m buying a used car. I hate it when I’m trying to work and the tools I’m using keep changing underneath me. I’d hate it if I was depending on bug fixes and enhancements in the tools I was using, and learned that the core engineering team from that company was going to spend a good chunk of the next year re-locating to a different country. I see all of these things happening to Altium customers and I hate it for them.
Of course, I hate it more when I buy a product that is critical to my job and requires lots of support – and then the company goes out of business.