I'm flying back from Alfresco's quarterly management meeting, and looking through some of the data shared by the team. We're just one company in a rising tide of open source companies, so I'm not sure how much to infer from our numbers about open source, generally, but I thought I'd share, anyway. Perhaps the data will be useful for you, too, as you plan your own open source business.
A few points:
- Lead Generation. Alfresco's website (Downloads/product trials) drives 72% of our leads, and documentation drives another 21%. Cold calls from sales people, direct mail, email, Google Adwords, and other sources you might imagine...almost nothing. To be fair, we don't really do direct mail or email campaigns yet, so it's not surprising that these bring in few leads.
But we found (as SugarCRM and MySQL found before us) that PR, which leads to visits to our website and subsequent downloads, is the biggest driver for our business. Lesson? Invest in great PR and ensure your web presence is well-designed to receive the visitors.
- Geography. US, France, Germany, Spain, UK, Italy, Belgium: these countries download our product (and associated documentation), and come back to buy from us more than other geographies. The US and France are strongest for us. I was surprised to find that France exceeded Germany, Spain, and the UK for us.
- "Free" is good business. We only recently started measuring unpaid implementations of Alfresco, and have over 6,000 "Community" implementations. (Note: We use a Fedora/RHEL model, and call our free product "Alfresco Community" and our paid product "Alfresco Enterprise" - very original of us, I know. :-) Given that we've had over 430,000 downloads in our first year of business, I suspect the actual number is much higher.
These Community sites prove to be our best source of qualified leads - that Community pool returns to buy Enterprise consistently and relatively easily. Our sales cycle - start to finish - is ~2.5 months, compared to our ECM peers that take 9+ months to consummate a sale. They are (or were, I should say - the $1M deals of yesteryear are largely gone for everyone, and good riddance) doing larger deals, to be sure, but if you divide the size of their deals by the time/resources required to close, I think we come out ahead.
I was worried about a "free-riding" market at first, but it seems to be paying off very well. As Martin Mickos has said, it's much better to have prospects freely using your product than for them to be paying for someone else's product.
- Competitors. For all the proprietary world's facade of nonchalance about open source, they sure do spend a lot of time on our site. And yours, too, I would imagine. (Dave recently wrote about a certain competitor that expends inordinate amounts of energy studying and spreading untruths about his company. I think it's probably more prevalent than we imagine. Why a billion-dollar company with hordes of marketeers would worry about little open source companies is beyond me...could they be feeling Fear, Uncertainty, and Doubt? ;-)
Btw, we did a quick survey of media attention paid to our competitors, and it turns out that they each (excepting Microsoft and EMC-Documentum, both of whom - especially Microsoft - are doing interesting things with ECM) get more press for their acquisitions than they do for their products. (And we get more than their product-related coverage.) I think that shows who is actually doing the interesting work in the industry - there's just not much to say about most proprietary software companies, if ours is a representative sample.
- Bugs. We get support emails/calls regularly from our customers. I just figured that they were all due to bugs. Our support manager (I would give you her name, but she's exceptional, and you might try to hire her) disabused me of this notion, however, when she pulled the data on why people contact us. Almost no customer or partner contact stems from bugs in the product, but rather configuration assistance and such. (Bugs account for a low single digit percentage of support calls.)
I don't think we necessarily write any better code than anyone else, so I'm assuming this is typical of all software companies. It surprised me, ignorant as I am of the support side of software businesses. I just assumed customers would be calling about bugs, nine times out of 10. Nope.
Btw, because our business is essentially support (as this is the primary ongoing value we provide to a customer, just like any software company), we've found it imperative to have support inextricably connected with engineering. The engineers who write the code take the calls. As we scale, we're hiring people whose primary job is support, not programming, but even here everyone must be contributors to the product. It means our customers spend less time with us on the phone/over email because we grok their problems more quickly.
I suspect this is critical for all open source businesses to achieve scale. It puts a big premium on engineers or super-technical quasi-engineers. If you fall into either category, your job prospects look bright.
I continue to be impressed with how different open source is as a business model. It requires a complete shift in where one invests resources (We treat Helen, our support manager, as a goddess), and how one thinks about customer acquisition and retention. I compare notes often with Red Hat/JBoss, MySQL, SugarCRM, JasperSoft, and other open source companies, and our experience is akin to theirs.
Open source is all about making software more like 99% of the world's industries: customer-focused and driven instead of vendor-focused and driven.
Posted by Matt Asay on December 8, 2006 12:52 PM
