Is Freemium Compatible with Enterprise Software?

Sure it is… despite the gnashing of teeth that regularly flares up when the discussion of freemium is engaged in.

Freemium is not magical pixel dust, it is a way to deliver a multi-channel business model that intersects customer segments regardless of size, complexity, and revenue opportunity. It does require you to rethink your business from top to bottom, beginning with how you meter your pricing model and ending with how you use your website.

The metering discussion is critical and you have a range of implements you can take advantage of, including metering by number of users, time, type of organization, features, and specific capacity dimensions. If you are offering free to high priced enterprise license agreements the metering by feature model creates landmines that you will have to deal with, most notably the bias that will emerge in your development process which results in packaging of high value features to defend your high price points which then results in the most interesting stuff you build being available to the smallest audience in your customer base.

Pricing for growth and pricing for margin are on different ends of the spectrum and if metering features supports margin then capacity must be in support of growth, right? Wrong, pricing on pure capacity creates a different kind of problem that you have to consider, which is leaving a lot of money on the table as a result of product realities that frustrate consumption. An example is a complex signup process that frustrates casual users or a deficient getting started process that creates barriers in the initial trial process. When pricing by capacity everything your development organization does must be viewed through the lens of creating consumption, even at the expense of value.

By the way, speaking of the initial trial process… will you have one? The trial process is valuable only if you are using it to facilitate a purchase decision otherwise you are better off having a “buy now” process or at a minimum have it as an option. If someone coming into a trial product experience isn’t using it in the first 30 minutes then it’s unlikely you will get them to use it over 30 days (which by the way is optimal if for no other reason than managing your internal reporting… all trials created in one month convert in the next).

Will the free product experience initiate in the trial process or as an explicit free product signup? If I were you I would go with a single trial product that converts into a paid or free product at the end of the trial, which works only if you don’t front load the trial process with payment information. The conversion rate for trial-to-paid will be low, probably 3-6%, at which point the debate shifts to getting people into the website and as efficiently as possible into a trial experience… it’s a pure numbers game at this point, feed the funnel with x number of site visitors to get to y number of trials to z number of paid customers.

The website is where the conflict between enterprise and monthly subscription customer segments will be realized. Enterprise marketing – the traditional kind – is entirely focused on content and getting contact information from a site visitor in order to have a sales resource follow up with them. This doesn’t work well for online freemium goto market strategies because it frustrates the goal of moving site visitors from the top of the funnel (your homepage) into a trial experience.

The conflict gets exacerbated when you realize that enterprise leads are opting into the low friction trial experience and once in that buying path the effort to shift them into the traditional enterprise path is cumbersome, mostly because of the pricing disparity. However, this fails to acknowledge that you, as a business, should not care how prospects come into your funnel and if your pricing model is appropriate for your business then prospects will naturally coagulate around the pricing plan most appropriate for them… pricing for growth.

The other dimension of the perceived channel conflict that you should consider is that it is unreasonable to think that the majority of your customers will start in free and end up in enterprise. Customers who select into one buying path are doing so because of what they want rather than what you are getting them to do… a customer who comes in via self-service trial, automatic conversion and monthly subscription renewal, and never talks with someone on your team is doing exactly what they want…. and that is not talking to you. Get over it, you can still serve them well and they will be happy, as evidenced by the fact that they renew each month.

You can help customers find the best fit for them among your product portfolio but doing so successfully at scale is as much dependent on in application marketing as it is good content on your website. Once you acquire a customer spend money designing and delivering a compelling product experience that facilitates upgrades and add-ons, rather than extensive email marketing campaigns and call centers.

Some enterprise products are not appropriate for freemium models but almost exclusively the result of high COGS and/or specific industry vertical issues (e.g. compliance). These are edge cases but you do need to be aware of them.

In summary, freemium works for enterprise software if you:

  1. Carefully consider all of the consequences of various packaging and metering models.
  2. Build your product to maximize the Hour-1 customer experience and then in support of the metering model you select.
  3. Build your website for throughput and efficiency in support of customer acquisition through the trial experience.

Columnar Data Storage

Hasso Plattner, SAP
Image by dfarber via Flickr

At the SAP academic research conference yesterday Hasso Plattner spent a lot of time talking about database design and why it’s still important. More significantly, he drilled into why re-architecting applications to take advantage of a fundamentally differently database than what we are used to with relational databases is critical if we are to simplify code bases and develop new generations of applications that take off where the current state of the art ends.

This area has a pretty steep learning curve so to get started here’s a good explanation of the distinction between columnar vs. row database architecture.

Reblog this post [with Zemanta]

Jive SBS Launches

I spoke with Sam Lawrence at Jive about their new Social Business Software (SBS) product and came away impressed on two fronts, the first being that the product is wicked cool and perhaps more significantly they are skating to the proverbial puck rather than following in the footsteps of other companies.

Longtime Jive followers will notice something immediately, Clearspace and Clearspace Community have been retired as naming conventions. For SBS, the technologies represented in both of these products are now referenced as “Jive Foundation” which forms the underpinnings for the new products and initiatives.

200903100946.jpg Jive is looking at the market opportunity from the standpoint of what people do with the software, and that represents the work centers which map to a neatly presented perspective on what happens in all companies. Within each of these centers is a business process in which a social component is integral. Based on my own experience in very large companies, I think this is a realistic perspective and it’s worth noting that the overlap between centers is probably proportional not by design but based on what actually happens.

In addition to process centers there are cross application modules that allow for top down functions across the entire suite of services. Analytics represent an obvious cross application module but it was the Bridging Module that really captured my attention.

200903100920.jpg

What the Bridging Module enables is a federation of related communities for an integrated view. As an example, Kaiser is a Jive customer and with the Bridging Module any Kaiser user could add components that represent content and functionality in the American Heart Association community.

To be clear, this federation capability works exclusively with other communities that are built on Jive technology, but with 2,500 customers this is a significant list and represents the greatest strategic opportunity for Jive, to become a vertical industry standard where they have strong representation. This is class Law of Accelerating Returns stuff, a vendor will win more new business as a consequence of being perceived as the accepted standard by a group of competitors within a specific vertical industry.

In the “old days” we would have called these things portals but it’s really an understatement to reference any of these products that way now. Portals relied on a single vendor or approved partners to supply functionality that was unavoidably focused around a single vendor’s products and was also typically transactional data focused. With the emergence of unstructured content and social interactions being the bigger drivers of user focus, portals were poorly equipped to deal with this and it opened the door for a menu of competitive products to emerge, Jive being one of the more successful offerings.

A further data point that underscores the point above is that the technical specifications for what constitutes a portal component are less of an issue today, and as Jive and Socialtext both demonstrate, an OpenSocial widget is just as accepted as a native component. The evolution of widgets demands that they move beyond content and creative to social awareness, in other words, how the widget or component interacts with other components is of equal importance to what the widget or component itself does.

This is a pretty competitive sector and there are firm lines that are developing. Microsoft and IBM offer the biggest footprint enterprise social software stacks and as can be expected they are expensive and timely to implement but on the other hand they offer a lot of functionality and demonstrable ability to scale to very large user numbers while also offering strong integration options to other important enterprise products. Other vendors have emerged that challenge Microsoft and IBM, such as Jive, while another class is extending the big enterprise offerings (most significantly what NewsGator is doing on Microsoft Sharepoint). With a flight to quality as a consequence of current economic conditions, the large vendors will continue to dominate while challengers like Jive with extensive customer lists and mature product offerings will close the window for new startups to establish a foothold.

Today the focus in on what users are doing rather than what companies want them to do and Jive’s SBS is well positioned to take advantage of that with a compelling user experience, strong social functionality, a “marketplace” for third party components and federated community sites, and lastly, advanced functionality (e.g. analytics) that grow in importance as usage grows.

Ubikwiti – DIY Business Processes

Ubikwiti is, despite a really unfortunate name, a very cool service. In a sentence, this is what I’ve always wanted to do with business mashups, combine off the shelf componentry at the business user level to achieve highly personalized business process models that still accommodated the need for master data and workflow integrity.

This is a huge challenge for any company because this is not how business users are accustomed to interacting with business applications. It’s also not clear that an significant number of business users want to do this but IT definitely wants more control over business apps in a manner that is abstracted from writing code, so maybe IT is really the front door in to business users.

The company is correct to assert that big enterprise apps don’t scale down well and large vendors have economic issues that obstruct a shift to pay-as-you-go pricing. The a la carte selection of business processes is something traditional enterprise companies will never do on their own but considering the waste that is in modern ERP systems from seats and functionality bought but not implemented, this is a good move aligned with the best interests of customers. Having said that, it’s also clear that there is a big functional gap between the top and bottom of the market so comparison to traditional enterprise software isn’t appropriate all the time.

Ubikwiti pricing is not pure a la carte though, it’s all you can eat for $4.95 a month per user. This is roughly half of the price point for Quickbooks Online but I am not in a position to evaluate the functional footprint relative to QB Online so I’ll reserve any judgement on their pricing model, but suffice to say, it does make it easy to get on to the system.

This is an intriguing service, I will be watching to see how they progress.

Open Source Companies to Watch

This is a good list, the three that I very much like are Untangle, SnapLogic (fyi, I met SnapLogic last year, wrote about it here) and Kickfire.

It’s easy to focus on SaaS companies in the enterprise space, relegating on premise to a wheezing and gasping dinosaur in it’s final days, but the fact remains that on premise software remains the overwhelming majority of spend for enterprise IT and that’s not likely to change for a few more generations.

Open source has been a bigger disrupter to enterprise IT than on demand SaaS but because it tends to focus on IT needs rather than business user needs, it gets much less attention. While open source for enterprise business applications has moved up the food chain more slowly than I predicted, it is happening and in the coming years we will see traditional proprietary code vendors come under assault once again by the prospect of open source, this time resulting in more significant progress than in years past.

Other open source companies I like:

Zenoss: Network and system monitoring software.

Enomalism: Build your own private elastic compute cloud.

rPath: Virtual appliances.

Qumranet: Red Hat just acquired this company, more virtualization technology (hypervisor).

OpenAir: Project management software for professional service organizations.

Piling on the VMW Mess Today

Lot’s of commentary today about VMware missing their number and CEO Greene out of the top job. For the record, when I was at SAP Ventures we looked at this deal but passed because of the husband/wife team (generally a big red flag for venture deals). It worked out for Diane and Mendel, and it would be stating the obvious to say that I regretted we passed on that deal but not just for the financial implications, Diane is the kind of entrepreneur you want to deal with, sincere and genuine but also really really smart.

Having said that, I think comments like this are misplaced and reactive:

Er, VMware may have tweaked its revenue forecast for 2008 to be “modestly below” previous guidance of 50 per cent growth. Few executives of multi-billion dollar companies usually get fired for 49 per cent growth, especially with an imploding worldwide economy in the background.

[From EMC CEO's ego has cost investors billions | The Register]

VMW has been at the center of a lot of confusion about their revenue forecasts and actual performance. Following the catastrophic Jan 28th conference call with analysts there was a lively discussion on the Enterprise Irregulars message board about their call and the details. In Q4’07 the company put up $412m against a $417m forecast… which being off $5m doesn’t seem disastrous but the fact that they came up $5m short while sitting on $550m in deferred revenues made no sense.

On that Jan 28th call they forecast that revenue would grow from $1.3b to $1.9b in 2008, 50% growth which again does not look bad. What that meant is that the company would need to add about $650m in 2008 revenue to meet their growth target of 50%, but when you consider that $550m in deferred revenue on the books it means they were really treading water for 2008. Basically they weren’t growing 50% at all but rather just covering their bases… and this from a company that had reliably put up 80% revenue growth.

At the time there was a line of thought that the company was resetting Wall St. expectations but today we can comfortably suggest that the company has some serious competitive and market challenges facing them. Blame Wall St. for being overly reactive but what investors were challenging is that apparent deceleration in VMW’s business and no substantive answers for what was going on.

More on this topic (What's this?)
VMware Inc Approaches The 50-Day Moving Average
VMware Worn Out
VMW Inc Is Trading Higher, But For How Long?
Read more on VMware Inc. at Wikinvest

Are we headed for a nuclear winter?

Dennis puts up a good post on the intersection of web 2.0 and the enterprise. O’Reilly actually talked about this in his keynote, that the enterprise is one of the 3 big themes he is focusing on, but I have to say that this segment will be a rude awakening for the 99.9% of companies who have been dealing with individual consumer and business users. Success in the enterprise requires, as Dennis points out, a deft touch when it comes to change management and bringing about change within the power structures that exist.

When it comes to the enterprise there will be a lot of roadkill among startups who thought they had a better mousetrap, which makes it all the more essential to build around you a good team of people who understand those markets but at the same time aren’t just going to say “we need to hire more $300k a year account execs and become a Gartner client”.

And therein lies the real crunch. While onlookers may bemoan the party bills at Web 2.0 events, the real problems for enterprise spend lay elsewhere. When I look at the margins vendors like Oracle make on legacy maintenance, it’s easy to see where fat can be cut and oxygen released for the kinds of innovation that drive value. Or what about the mad dash for governance, risk and compliance consulting projects at premium rates? Then there is the whole problem of delivering value from social networking applications.

[From Are we headed for a nuclear winter? | Irregular Enterprise | ZDNet.com]