Guest Post: Lloyd McKenzie on HL7s IP Announcement
Sep 5, 2012Note: This is a guest post from Lloyd McKenzie, a follow up to my previous post. These are not necessarily my opinions, though I sort of generally agree. I’ll happily post other guest posts on this subject if people submit them to me. I’m going to be a bit more blunt than Grahame.
I don’t object to HL7 content being free to non-members, but I’m hugely nervous (using polite words, seeing as this is a public blog) about the potential outcomes and the timing of the announcement. Specifically I’m concerned about the announcement of a decision with huge consequences for the viability and direction of the organization with a side note that “we’re still working out the details”.
This is not an announcement that’s reversible. If HL7 changes its mind, we lose all market credibility, so we’re now in the position of having to release the IP for free whether we can come up with a viable business model or not. And from the sounds of things, the viable business model is part of the details to be worked out. That doesn’t seem like wise planning in my books.
However, for better or worse, the decision’s now been made. Given that I have a strong intellectual, emotional and financial attachment to the HL7 organization and would like to see it succeed for many more years, it becomes all the more important to figure out what a viable business model could look like. There’s not a lot of time to get this right because the revenue stream from the old business model is going to start drying up tomorrow (regardless of the fact that the change to IP rights won’t take effect for “months”).
This guest blog is going to be a bit longer than a blog entry probably ought to be. That’s because the issues here aren’t simple. (If they were, HL7 would have adopted a new business model long ago.) My intention is to get a discussion started now so that we can make good progress at the WGM and move quickly from the analysis stage to the implementation stage.
An HL7 business model
A business model has two sides. What do we need to spend money on and where does that money need to be spent?
HL7 has costs, and they’re not insignificant. Last year they were over $4 million. Services include:
- Managing the various mediums for member communication and standards development (teleconferences, list servers, website, wiki, gForge, etc.)
- Coordinating the various governance structures including secretariat duties for the board, TSC, ARB, FHIR governance groups, tooling, project management, etc. These are essential to ensure the organization can coordinate, respond in a semi-timely fashion and that the right hand knows what the left hand is doing most of the time. In theory, this also includes paying for roles such as the CEO and CTO, though those are currently gratis.
- Organizing and executing the 3 Working Group Meetings (WGMs) each year
- Supporting the legal structures necessary for being a standards organization, including managing processes required for ANSII and dealing with the pleasures of ANSII audits
- Publishing content for ballot, managing the ballot process and actually publishing the standards specifications. This include developing and maintaining the tools needed to perform these functions.
If these things don’t continue to exist, then HL7 (and the content it produces) doesn’t exist anymore either. Some of these functions might be reduced or eliminated, but any significant decrease is going to negatively impact how well the organization works, and that in turn negatively impacts the benefits resulting from membership in the organization risking a downward spiral.
On the revenue side, fees come in from membership fees and from WGM registration fees. I’m going to largely discount the latter. The objective of the fees is to make WGMs a revenue neutral activity. Sometimes we run a profit. Sometimes (particularly non-US meetings) we run a significant deficit. Increasing meeting fees will seriously interfere with the work of the organization and given the demand curve is unlikely to significantly increase available funds.
In theory rather than just having additional funds come from membership, we could look at a “fee for service” approach where in addition to or in parallel with membership, HL7 could offer certain benefits for an additional cost. How best to go to market with a particular service – bundled in membership, bundled in something else, stand-alone or some combination there-of – is best left to those who know marketing strategy. I’m going to cross my fingers and hope that HL7 can get some good advice in that space. For now, let’s just look at the services the organization does or can offer and figure out which ones might drive someone to pay HL7 International money.
1. IP Use
This has been one of the main drivers for HL7 membership – even more so recently, since HL7 put such a large focus on defending its IP combined with the mandated use of HL7 standards by various organizations. And it doesn’t necessarily all go away.
“But wait, didn’t HL7 just say it wasn’t going to charge for IP anymore?”
Technically, no. The press release talked about “standards and select IP”, not all HL7 IP. If you want to mince words, HL7 produces a bunch of IP that isn’t necessarily a “standard”. And the organization could get more creative in the future to cause fewer things to fall under the “standards” banner. For example, we might say that CDA and v2 are standards, but the implementation guides based on them are not. So if you wanted to do CCDA or v2 immunization, you’d still need to be a member. The organization could do that. But they’d be foolish if they did.
First, the market would not look kindly on the bait and switch. We’d lose a lot of face and credibility. And in the long run, other organizations such as IHE that already do a pretty decent job of putting together implementation guides would simply take on more of that work. There are problems with that of course – having the IGs developed by a separate organization from those who create the standard tends to create disconnects that don’t serve either side well, but it would be a natural consequence of charging for the implementation guide when the standard was free. The approach would likely result in a net decline in revenues, certainly not much of an increase.
HL7 has other IP too. Things like code system definitions, domain analysis models, functional profiles, etc. that might not qualify as “standards”. We could reverse course and try charging for those, but it’s not clear that we would get significant revenue for them, nor that the organization provides such a significant value add that someone else couldn’t do it as well or better than we do.
2. Influence the standards
This is the big one. This is what drives most of the participation in HL7. All of the big contributors – Benefactors, Supporters and even quite a few of the individual members participate in the standards organization because they want to affect the outcome. Whether it’s getting a feature into a specification (or keeping one out) or shifting the focus of the organization to something new, or even getting a new specification built, participating is the way this happens. The drivers can vary – consultants seek to have influence and expertise they can market to clients, vendors to ensure a good fit between the standards and their products, healthcare institutions and academics to ensure that standards meet evolving clinical requirements, governmental bodies to ensure alignment of standards with jurisdictional objectives, etc. In theory, patients could even be at the table, though they usually aren’t.
The challenge here is that engagement isn’t a short-term thing. If you want to influence a specification, you’re looking at a minimum of a 1-year turn-around. And to have serious influence or get something big done, two or three years is more realistic. That’s not because HL7’s slow. It’s because building standards means getting consensus. Influencing consensus means learning the ins and outs of the organization and building a reputation so that people listen when you talk. And those aren’t quick activities.
That means that the average physician office system vendor who’s in the throes of trying to meet Meaningful Use requirements is unlikely to see a business case for the level of investment required vs. the amount of influence they’d likely have. We’ll likely pick up a few, but fewer than we’ll lose who are currently only members for IP reasons (especially those who only signed up in the last year or two because we sent them nasty letters).
In practice, at present it’s possible to have significant influence on HL7 specifications without being a member. HL7 conference calls are open. So is the wiki. So are our list servers. So are WGMs (though slightly more expensive for non-members). The only limits on participation come if you want to be a committee co-chair, want to run for the board of directors or other governance position, or if you want to vote on specifications.
If HL7 wanted to further grow revenue in this space, there are few possible approaches:
- We could change policies on list server, conference call, WGM and other participation to only permit members to play. However, this would be self-defeating. Those individuals and organizations that are now HL7 members are members because they were once non-members and decided to “check out” HL7. They got hooked, got engaged and became members. If outsiders can’t see in, they don’t have a chance to get hooked or engaged. As well, the quality of HL7 standards is dependent on the breadth of the review and input they receive. Limiting review to members only hurts both members and non-members.
- We can increase the number of specifications we produce and/or the breadth of the communities that use them. If a percentage of those implementers decide to get involved, that grows revenue. This should certainly be a long-term focus, and is one of the arguments made by the proponents of free IP. However, it’s a long-term thing. The new members will accrue over many years, while those we lose who were just paying for IP will be gone in months (1 year max).
- We can charge differently for “greater influence”. For example, creating implementation guides or adding features “for pay”. This has been bandied about, but is quite challenging to accomplish without completely changing the character of (and likely the value of) the HL7 organization. Most work at HL7 is done by volunteers pursuing their own interests and the interests of their sponsors. The fact HL7 might have received money from an external source to get something done doesn’t necessarily mean that these “volunteers” will have any incentive to get that work done quickly – or even at all
- We can follow a “pay for governance” model, similar to the OMG where the TSC or FHIR governance board or ARB or similar bodies have their membership based on financial composition. If this was done extremely carefully and the right people paid, it could work. But if the people in the governance positions didn’t have the respect of the organization, HL7 would become dysfunctional very quickly. And I have no magic recipe for how to ensure we get the “right” people. (Not to mention that most wise organizations would want to wait a few years to see if HL7 was going to implode before making much investment in this space, making it a longer term proposition too.)
3. Tooling
The FAQs for the IP change made reference to “Implementation Tools” as one of the benefits of membership. This was somewhat of a surprise. I’m pretty involved with tooling and by and large, HL7 produces no implementation tools. There are some, but they’re all from third-party sources and thus aren’t HL7 IP (and certainly not subject to membership restrictions). The tools that HL7 does create are generally produced by volunteers and are mostly freely available.
HL7 has made noises about getting into the business of developing implementation tools. However, this is a hard business. If it’s being done on a volunteer basis, then charging for them is hard. If they’re created on a commercial basis, then making it cost effective is hard. Most of the big HL7 tooling projects, whether sponsored by HL7 International or one of the affiliates have had both schedule and budget issues.
In short – tooling is not HL7’s core competence. Expecting it to suddenly become a revenue source, particularly in a timely enough fashion to offset short term losses seems unwise.
There are other issues here too. Implementation tooling can come with liability issues and liability around healthcare isn’t a good thing. There’s also the question of market impression. If we move to a space where the standards are free, but the tools needed to implement them are not, then there’s either a perception or reality of conflict of interest and the market is going to see HL7 as being proprietary regardless of the status of our IP
4. Certification
This is not a space HL7 has operated in before. In fact, it’s a space we’ve worked hard to stay far away from, in part because those who pay for membership don’t look kindly at being told that their products aren’t compliant (and perhaps that their competitors are). However, if we need to look at new revenue sources, this is a possibility. Anyone can use HL7, but if you want to be “certified”, you need to pay.
There are two challenges here: What’s the value of certification, and how do we do it?
Vendors will only pay for certification if there’s value in it. That generally means that their customers either demand it (possibly because regulators demand it) or see enough value in it they’ll pay extra for it. None of these are likely to develop prior to a track record of success in the certification process being demonstrated and aren’t guaranteed even then.
Those who have done anything in the certification space can tell you it’s not easy and not cheap. Or if it is easy and cheap, it probably doesn’t accomplish much. The best certification process out there in terms of speed and cost is probably IHE’s Connectathons, largely because they offload a lot of the work to the implementers themselves. It’s not clear that HL7 can do a better job of something that the IHE’s been doing for years, though it’s an option for us to try.
5. Standards Plus
While HL7 has committed to making it standards and selected IP available for free, it hasn’t committed to making it available in only one way or having the “free” way be the “best” way. It could make a “basic” version, perhaps as an un-indexed PDF available for free, while the fancier indexed version with wire format instance examples, etc. would only be available to paying members. Or perhaps paying members would receive access to a specification 6 or 12 months earlier than non-members.
This would likely encourage some, though not all, implementers to apply for membership. However, realistically this would only encourage the purchase of individual memberships rather than the more lucrative organizational memberships.
6. Restricted Advertising
While HL7 might allow free use of its specifications, it can still place restrictions on the use of its trademarks and logos. I’m not a legal expert, so exactly how restrictive we can get I’m not sure. I suspect that we can’t keep people from using the term “CDA” or “FHIR” in their marketing literature, but we could likely restrict the use of the HL7 logo and some of our other marks. Of course, this only makes a difference if customers perceive a difference between those who use the marks and those who don’t, which will take a degree of marketing effort by HL7. The benefit here is that these could easily be restricted to organizational members only – individual members would have little use for them.
One consideration here might be whether such constraints should be limited to commercial interests but perhaps relaxed for open source efforts (which would have limited or no financial resources to pay membership fees.)
7. Other options?
This isn’t necessarily an exhaustive list. If you have other ideas, or disagree with some of the assertions about pros and cons of the options above, feel free to respond in the comments section.
The sooner HL7 can put non-productive ideas aside and move forward with those that are likely to be successful, the faster we can help the organization recover from the dip in funding that’s likely to follow this announcement and move back into growth mode.