September 16, 2007 12:19 PM PDT

Open-source mechanics: Improving conversion rates

Open-source business depends upon data. Because so much of the business happens online, it's possible to set and track metrics that would otherwise be difficult to do in a proprietary model. One area that all open-source startups focus on is improving conversion rates.

The difficulty is not in converting an opportunity into a sale. If my open-source peers are like Alfresco, and I'm sure that they are, we actually top the industry in that metric. Why? Because by the time a lead becomes an opportunity, the prospect is already using the software and is effectively "pre-sold" on its value.

No, the real question is how to turn raw downloads into qualified leads. Importantly, this must be done at low cost so that these cost savings can be passed on to customers. A high-cost sales model does not mesh well with low-cost open-source software.

The Alfresco team has been working a lot on this lately. I thought I'd share some of the things that have worked for us, in no particular order:

  1. Good documentation. I've written on this before. If you don't have good documentation, you force too much of a burden on your pre-sales organization. It's expensive and, frankly, with no upfront license fee, foolhardy. Open-source software largely needs to sell itself.

  2. Establish an evaluation framework. With proprietary software, everything points to a license purchase as the "essential next step." not so in open source, where the would-be customer already has the software. An evaluation framework - which might consist of videos showing how to do important things in the software, or how-to guides, or...you get the picture - walks the prospect through the evaluation process, so that they see a solution to their business problems, rather than just a mound of code.

    As with documentation, pre-sales could fill this role, but the name of the game in open source is to enable adoption without heavy lifting, so that you can scale the sales process without linearly scaling the sales organization.

  3. Automate nurturing of leads. Much of what is involved in this is covered above or below, but this step relies, in particular, on demand-generation software (see more on this below). Open-source companies can't afford to call every lead that comes in. Alfresco has many thousands each month, and I know our peers are the same (or higher).

    Before you engage your lead-generation team, you need software that can detect patterns of use (or, sometimes more importantly, non-use) and send emails or trigger other events (perhaps a warning to the sales representative over the territory) that lead to the prospect doing something (e.g., reading documentation on an issue that had stymied their progress). If open-source companies don't automate, they're toast.

  4. Guide the download. This is an amazingly simple, but often overlooked, measure. Basically, you want to remove guesswork for the prospect. When someone downloads Alfresco, we detect the operating system for them and so only present the option that fits them. (More than once I've downloaded the wrong version of OpenOffice, for example, because I downloaded the PowerPC version - software should detect this and present the right choice for me.)

  5. Make as many demos available as possible. Open-source companies need to help prospects see how the software can be used to address their needs. Every online demo is one less that needs to be done by a pre-sales engineer. It's all about scale.

  6. Be active on forums. Prospects will often have questions the documentation won't answer. Answering a question once on one's project forum is a great way to scale an answer to many prospects (and existing customers).

  7. Focus the prospect on the project/product, not (yet) its implementation. It's important for a would-be buyer to be comfortable with both the software and how it will integrate into their environment, of course, but the best customer experience involves commitment from both the prospect and the vendor.

    Things like virtual appliances can help here, as they enable prospects to hone in on the value of one's software without (yet) getting into the details of how to integrate it into one's IT environment. In the proprietary world, customers commit to a license before they spend any time on the integration piece.

    In the open-source world, they get to preview the integration experience and, if they choose, go far down that road. But it's good to keep would-be customers focused on what the software actually can do for them. There will always be pain in implementation, whether they're looking at proprietary or open-source software. Best to keep the focus on the software.

  8. Enable "light" engagements to solidify customer confidence. Alfresco offers a 90-day trial which offers prospects the full Alfresco experience: certified software + support. It's our way of showcasing to prospects what the full Alfresco experience is, and it's a way for customers to get added support during a critical trial phase. One added benefit is that it creates an artificial "sell by" date. Everyone knows that at the end of 90 days, a "go" or "no go" decision is expected.

  9. Use demand-generation software. This is not negotiable in an online business. If open-source businesses aren't tracking how prospects interact with them as companies, they will fail. Period. Regardless of whether you use proprietary Eloqua or open-source Loopfuse, it's critical to know how long it takes for a download to turn into a trial, and at what point prospects start requesting documentation, and what kinds. All of this data makes an open-source company more efficient in fine-tuning its processes to enable it to sell exceptional service...at a fraction the cost of proprietary software.

These are just a few of the things that Alfresco does, and to good effect. I'm sure there are a range of things we can do better, so if you have suggestions based on what your own company has done, please share them. Open-source sales models are still an art, rather than a science. It's good to share best practices because, as you can see from the above, the direction is to make the process more science than art.

Recent posts from The Open Road
New startups explore new niches for open source
Analysts as a lagging indicator of success
Symbian on the decline: Time to move on open source
The key to making money: Charge for your product
MySQL's Monty Widenius leaves Sun
Add a Comment (Log in or register) 14 comments
A chink of light on......
by PACSferret September 17, 2007 12:28 AM PDT
.... the inner workings of Alfresco. Nice post, Matt, although I have qualms on your point about lead generation. You seem to be suggesting that it is critical to have a facility to monitor the status of usage - downloaded/installed/in use/in widespread eval etc. An essay on google and privacy aside, I know as a user/evaluator/integrator of open source software if I am ever given the choice to provide such feedback automatically the first thing I would do is wrap the whole thing up into a firewall until I knew - firstly what the 'feedback' entailed, and secondly, that I intend to move to the next step.
Reply to this comment
A chink of light on......
by PACSferret September 17, 2007 12:28 AM PDT
.... the inner workings of Alfresco. Nice post, Matt, although I have qualms on your point about lead generation. You seem to be suggesting that it is critical to have a facility to monitor the status of usage - downloaded/installed/in use/in widespread eval etc. An essay on google and privacy aside, I know as a user/evaluator/integrator of open source software if I am ever given the choice to provide such feedback automatically the first thing I would do is wrap the whole thing up into a firewall until I knew - firstly what the 'feedback' entailed, and secondly, that I intend to move to the next step.
Reply to this comment
Missed step 1
by ian.waring September 17, 2007 2:17 AM PDT
Another metric is the conversion rate from ad-hoc market knowledge to download attempts, which is like traditional web marketing "fold factors" (number of page views divided by the number of unique visitors to your web site). Or if you do print advertising, opportunities to see (OTS) divided by the actual response. In the latter case, given that high OTS's drive advertising cost, just measuring cost per response for each advert placement (we even had different 0800 phone numbers per publication to measure CPR off the switch, as we found most operators in the call centre were picking the first choice on a drop down menu - whatever it was!).

When I ran the marketing team for an ISP, we were landing consumer and small business customers for £30 each (AOL were £180 at the time). The main challenge is getting marketing types to get at least one test in each campaign, to measure the waterfall down to "landed long term customer" as accurately as we could, and then split run testing every implementation (even just changing a few words around to see if they pulled differently). Done correctly, it should be a continuously learning process (for everyone) rather than art.

The best media we found in three years of trying was paying the UK Royal Mail to cancel their stamps on letters with a stamp that said "For your Free 30-day Trial, call Demon Internet 0800 666 666". £4 per landed customer.

The main difficulty with Open Source is that the target audience is completely different than that for proprietary vendors (we sell to both). When we did some profiling work with MySQL, our sales teams in one specific large bank knew no-one listed as the folks implementing 14 different projects there (gleaned from details volunteered at download time). The large proprietary vendors are selling "platform strategy" and high ticket items like mad at and around CIO level, whereas open source projects percolate up the organisation as "instant pudding"; by that time, the line of least resistance is just to use what's already working. Surprise arrivals rather than CIO mandated.

Developers are the key community in most large organisations for open source - and where folks like Microsoft, Oracle and indeed even Red Hat now are focussing much more of their efforts. One of our challenges is to spot them systematically and to engage the 4P's and 2D's - Product, Place, Price, Position, Database and Dialogue.

The latter two are the most difficult of all. A lot of Open Source software just "works"... until 12 months ago, only 30% of our Red Hat base even bothered to activate their susbcriptions (and hence would never have called Red Hat for technical support nor received any updates). We now activate everyone at point of initial subscription sale. Our next challenge is to keep in touch with all those developers as they play musical chairs around various projects in their organisations. Way to go...

Ian W.
Reply to this comment
Missed step 1
by ian.waring September 17, 2007 2:17 AM PDT
Another metric is the conversion rate from ad-hoc market knowledge to download attempts, which is like traditional web marketing "fold factors" (number of page views divided by the number of unique visitors to your web site). Or if you do print advertising, opportunities to see (OTS) divided by the actual response. In the latter case, given that high OTS's drive advertising cost, just measuring cost per response for each advert placement (we even had different 0800 phone numbers per publication to measure CPR off the switch, as we found most operators in the call centre were picking the first choice on a drop down menu - whatever it was!).

When I ran the marketing team for an ISP, we were landing consumer and small business customers for £30 each (AOL were £180 at the time). The main challenge is getting marketing types to get at least one test in each campaign, to measure the waterfall down to "landed long term customer" as accurately as we could, and then split run testing every implementation (even just changing a few words around to see if they pulled differently). Done correctly, it should be a continuously learning process (for everyone) rather than art.

The best media we found in three years of trying was paying the UK Royal Mail to cancel their stamps on letters with a stamp that said "For your Free 30-day Trial, call Demon Internet 0800 666 666". £4 per landed customer.

The main difficulty with Open Source is that the target audience is completely different than that for proprietary vendors (we sell to both). When we did some profiling work with MySQL, our sales teams in one specific large bank knew no-one listed as the folks implementing 14 different projects there (gleaned from details volunteered at download time). The large proprietary vendors are selling "platform strategy" and high ticket items like mad at and around CIO level, whereas open source projects percolate up the organisation as "instant pudding"; by that time, the line of least resistance is just to use what's already working. Surprise arrivals rather than CIO mandated.

Developers are the key community in most large organisations for open source - and where folks like Microsoft, Oracle and indeed even Red Hat now are focussing much more of their efforts. One of our challenges is to spot them systematically and to engage the 4P's and 2D's - Product, Place, Price, Position, Database and Dialogue.

The latter two are the most difficult of all. A lot of Open Source software just "works"... until 12 months ago, only 30% of our Red Hat base even bothered to activate their susbcriptions (and hence would never have called Red Hat for technical support nor received any updates). We now activate everyone at point of initial subscription sale. Our next challenge is to keep in touch with all those developers as they play musical chairs around various projects in their organisations. Way to go...

Ian W.
Reply to this comment
Look after those geeks
by PACSferret September 17, 2007 2:44 AM PDT
Ian makes a fine point - it's the guys who can be loosely described as 'geeks' (not just developers) that bring in an OSS product, take a look, and propose it up the line. Particularly in SMBs, even now, that may be the first time to 'C' level gets much exposure to OSS software at all (Linux has gone to mainstream - doesn't count anymore!).

But, not being at the 'C' level, the 'proposers' tend to get only a handful of chances & it'll only be the products they feel most comfortable with that will get put onto the table, so it's critical that those guys are looked after - and in that sense if you don't do (number 1) documentation right, you won't get a chance put over-burden pre-sales.
Reply to this comment
Look after those geeks
by PACSferret September 17, 2007 2:44 AM PDT
Ian makes a fine point - it's the guys who can be loosely described as 'geeks' (not just developers) that bring in an OSS product, take a look, and propose it up the line. Particularly in SMBs, even now, that may be the first time to 'C' level gets much exposure to OSS software at all (Linux has gone to mainstream - doesn't count anymore!).

But, not being at the 'C' level, the 'proposers' tend to get only a handful of chances & it'll only be the products they feel most comfortable with that will get put onto the table, so it's critical that those guys are looked after - and in that sense if you don't do (number 1) documentation right, you won't get a chance put over-burden pre-sales.
Reply to this comment
Agreed on the privacy
by Matt Asay September 17, 2007 6:06 AM PDT
@PACSferret: I should have been a bit more explicit on that. We don't actually track people as people until they explicitly allow it. We try to provide a lot of value for that information so that people have an incentive to share it, but we don't glean any of it without permission.
Reply to this comment
Agreed on the privacy
by Matt Asay September 17, 2007 6:06 AM PDT
@PACSferret: I should have been a bit more explicit on that. We don't actually track people as people until they explicitly allow it. We try to provide a lot of value for that information so that people have an incentive to share it, but we don't glean any of it without permission.
Reply to this comment
Hit one of my pet raw nerves :-)
by ian.waring September 17, 2007 7:51 AM PDT
PACSferret: if we're talking SMBs, I tend to think of organisations staffed with folks with pretty constrained time resources and where business survival is a much more laudable goal than growth. That's where "instant pudding", "customise as you go" and "hosted services" tend to intersect.

If I could slap a company logo/colour scheme on a hosted version of Alfresco and have me publishing shared documents to external companies using the same thing, I think i'd have a chance of affecting my company's strategy (which is articulated back to me as "Microsoft .net"! - a strategy based on techie talk rather than business outcomes). If there's a tick in the box for single internal sign-on using Active Directory later, then i'd have ticks in the right boxes.

A lot of vendors tend to focus on "SMB", whereas the businesses themselves never think of themselves as "small" - just more focussed (on an industry sector, application or geography). In the case of IBM SWG, they think anyone who does less than $4m/year of license fees with them is an SMB!

For a genuine SMB, a hosted, customise as you go, online service would probably give you more early traction than relying on an inhouse developer - IMHO of course.

For another SMB snippet, any idea who the largest UK reseller is for both Sage and for Intuit Quickbooks? It's achieved as a £15/month subscription service by... Barclays Bank.

Ian W.
Reply to this comment
Hit one of my pet raw nerves :-)
by ian.waring September 17, 2007 7:51 AM PDT
PACSferret: if we're talking SMBs, I tend to think of organisations staffed with folks with pretty constrained time resources and where business survival is a much more laudable goal than growth. That's where "instant pudding", "customise as you go" and "hosted services" tend to intersect.

If I could slap a company logo/colour scheme on a hosted version of Alfresco and have me publishing shared documents to external companies using the same thing, I think i'd have a chance of affecting my company's strategy (which is articulated back to me as "Microsoft .net"! - a strategy based on techie talk rather than business outcomes). If there's a tick in the box for single internal sign-on using Active Directory later, then i'd have ticks in the right boxes.

A lot of vendors tend to focus on "SMB", whereas the businesses themselves never think of themselves as "small" - just more focussed (on an industry sector, application or geography). In the case of IBM SWG, they think anyone who does less than $4m/year of license fees with them is an SMB!

For a genuine SMB, a hosted, customise as you go, online service would probably give you more early traction than relying on an inhouse developer - IMHO of course.

For another SMB snippet, any idea who the largest UK reseller is for both Sage and for Intuit Quickbooks? It's achieved as a £15/month subscription service by... Barclays Bank.

Ian W.
Reply to this comment
Pros/Cons of forced registration
by tristanbob September 17, 2007 8:32 AM PDT
Matt,

Am I correct in assuming that your "tracking" is accomplished by the registration that is required on your website? It must be, or you could not contact your leads.

Anyway, there are obvious pros and cons to this argument. Personally, I think that registration is a barrier to investigators. I understand that registration does not cost an investigator anything except time and privacy, but it still deters people.

When I am researching open source products, I want instant access to online resources (like whitepapers, documentation, etc). Registration leaves a bad taste in my mouth when I run into it, since that is how many traditional software companies work.

You must have discovered that the benefits of forced registration exceed the negative of turning people away, otherwise you would not be following that strategy. The benefits include of forced registration are that you can see what your investigators are doing, and are able to contact them.

Any thoughts you want to share on this issue?

Tristan
Reply to this comment
Pros/Cons of forced registration
by tristanbob September 17, 2007 8:32 AM PDT
Matt,

Am I correct in assuming that your "tracking" is accomplished by the registration that is required on your website? It must be, or you could not contact your leads.

Anyway, there are obvious pros and cons to this argument. Personally, I think that registration is a barrier to investigators. I understand that registration does not cost an investigator anything except time and privacy, but it still deters people.

When I am researching open source products, I want instant access to online resources (like whitepapers, documentation, etc). Registration leaves a bad taste in my mouth when I run into it, since that is how many traditional software companies work.

You must have discovered that the benefits of forced registration exceed the negative of turning people away, otherwise you would not be following that strategy. The benefits include of forced registration are that you can see what your investigators are doing, and are able to contact them.

Any thoughts you want to share on this issue?

Tristan
Reply to this comment
"Forced Registration"
by royrusso September 17, 2007 8:45 AM PDT
@Tristan

Having deployed our Marketing Suite to many an OSS customer, I can share with you a bit of best practices, regarding "forced registration".

The single-most important thing to remember is to strive for an equitable exchange of information. i.e. For someone wanting to register on your forums for free support, you may want to ask more questions in the registration form, than someone wanting to download/eval your product. Sometimes a form with name, email, phone works fine and equitably in the user's eyes. (I would ask people to enter in "Company", but LoopFuse already knows what Company you're in ;-)

Consider your basic OSS company with a message board, wiki, downloads, hosted-trial, and PDF/whitepapers. Those are 5 separate opportunities for you to grab some "nugget" of information from your user and assign an identity to him/her. With automated marketing tools, you can then further qualify/score that lead and import him/her in to your CRM.
Reply to this comment
"Forced Registration"
by royrusso September 17, 2007 8:45 AM PDT
@Tristan

Having deployed our Marketing Suite to many an OSS customer, I can share with you a bit of best practices, regarding "forced registration".

The single-most important thing to remember is to strive for an equitable exchange of information. i.e. For someone wanting to register on your forums for free support, you may want to ask more questions in the registration form, than someone wanting to download/eval your product. Sometimes a form with name, email, phone works fine and equitably in the user's eyes. (I would ask people to enter in "Company", but LoopFuse already knows what Company you're in ;-)

Consider your basic OSS company with a message board, wiki, downloads, hosted-trial, and PDF/whitepapers. Those are 5 separate opportunities for you to grab some "nugget" of information from your user and assign an identity to him/her. With automated marketing tools, you can then further qualify/score that lead and import him/her in to your CRM.
Reply to this comment
Powered by Jive Software
advertisement

Latest tech news headlines

Resource center from News.com sponsors
What you need in business class email.
Mailtrust

Click Here!
Never worry about email again. From mobility and shared calendaring to virus and spam protection starting at only $3 per mailbox. more>

Rackspace Mailtrust
Total Email Relief

We'll take care of your email so you can take care of your business.

14 Day Free Trial

With expert support 24x7x365 we guarentee 100% uptime. Try us for free for 14 days. Never worry about your email again.

Just $3 per mailbox

Choose the plan that is right for your company and only pay for what you need.

About The Open Road

Matt Asay brings a decade of in-the-trenches open-source business and legal experience to the Open Road, with an emphasis on emerging open-source business strategies and opportunities. Matt is general manager of the Americas division and vice president of business development at Alfresco, a company that develops open-source software for content management. He is a member of the CNET Blog Network and is not an employee of CNET. Disclosure.

Add this feed to your online news reader

The Open Road topics

Featured blogs

advertisement
advertisement

Inside CNET News

Scroll Left Scroll Right
  • News - Business Tech

    Chrome's JavaScript challenge to Silverlight

    The advent of Google's Chrome browser, software pros say, should spur a big speedup for JavaScript, which would raise its standing against Microsoft's Silverlight technology.

  • Gallery

    Photos: Top 10 reviews of the week

    Here are CNET Reviews' 10 favorite items from the past week, including the TiVo HD XL, Sony Cyber-shot DSC-H50, and the Dish Network's newest digital TV converter box.

  • News - Apple

    Apple watchers spot 'iPod Nano' pix, iTunes hints

    The rumor mill has long been predicting a longer, leaner new version of the iPod Nano, and now it's conjuring up some pictures.

  • Coop's Corner

    Chris Shipley 1, Internet lynch mob 0

    Demo's impresario goes public with a tart and smartly written riposte to the shoot-from-the-lip crowd.

  • Video

    Katie Couric reflects on first Webcast

    The political conventions are over and so are CBS Evening News anchor Katie Couric's first series of Webcasts. CNET's Kara Tsuboi sat down with Couric on the final night of the Republican National Convention to discuss what she liked about Webcasting, some of her most memorable guests, and whether TV news will still be around by the next round of conventions.

  • News - Digital Media

    Google-focused satellite enters orbit

    The search titan has exclusive rights among online mapping sites to images from the new GeoEye-1 satellite, which launched Saturday.

  • Video

    YouTube plays party politics

    During the presidential campaigning four years ago, YouTube didn't even exist. Now it's a tool candidates must master to get their message across. CNET's Kara Tsuboi stops by the YouTube upload booths at the Democratic and Republican conventions to find out why Google's video site has such a big presence in Denver and St. Paul, Minn.

  • News - Gaming and Culture

    Are Demo and TechCrunch50 fragmenting their audiences?

    With both events scheduled to start Monday, many press, as well as venture capitalists and others are having to choose which one to attend.

  • News - Cutting Edge

    Execs predict next Google-like tech

    On eve of company's 10-year anniversary, researchers and business pundits speculate about what technologies might someday have as much impact as Google.

  • Gallery

    Images: The art of 'Spore' prototypes

    Will Wright and his Maxis team worked on dozens of prototypes to test the elements of their soon-to-be-released evolution game. Here's a sampling.

  • Webware

    At the TechCrunch50, an unfair advantage?

    Inside baseball: How Webware and other blogs can compete with TechCrunch in covering the TechCrunch50 event.

  • Green Tech

    Duke Energy to invest in mini solar power plants

    Can hundreds of rooftop solar panels collectively operate like a central power plant? Duke Energy launches $100 million distributed solar program to find out.