I talked last week how the license commission structure can hurt customers. There were a LOT of great comments and insights from the product side. Would love to have customers and consultants chime in on their feelings, but I can’t have everything.
There were two comments that I wanted to address in more detail. The first was that money isn’t the only driver of behavior for people. That is very true. That said, all else being equal, ask yourself this question: Who will a sales person go talk to first? The company that will help land a commission and meet quota, or an existing customer?
The fact that money is not the sole motivator of behavior is not something to just be dismissed, but neither can the impact cannot be discounted. The very existence of financial incentives will impact behavior. Let’s face it, when those incentives aren’t earned by an employee, those details show up in a performance review.
Now that I’ve thrown some debate points out there, let’s dive into the details of this post.
Beyond the Sales Curtain
I’m not going to touch the whole Investor side of the house because I have very little exposure there. I am going to focus on how multiple vendors take the obsession with license revenue a bit too far.
As discussed, license revenue is great, but it doesn’t measure success. I can sell as much as I want, but if half of my customers are gone in two years because my product isn’t improving, or is inadequate to begin with, it doesn’t matter.
The problem comes when companies take licensing metrics and use that information to start evaluating the product teams and setting Research and Development (R&D) investments. That can lead to some poor decisions.
My AnnoDoc Flashback
Way back when, I was a product manager for a small product called AnnoDoc. It was a simple annotation tool that grew out of a consulting engagement. It worked fairly well and I was able to get a few key partners into place to help with our sales.
I remember very clearly the day that we closed a landmark deal. It was the deal that pushed our Support revenue past the cost for the entire product team, including support staff. All we had to do at that point was defend our turf and use any additional license funds to increase R&D, improve our marketing, and add to our cash reserves.
The thing is, that isn’t what happened. That money went elsewhere and I had to fight for budget to grow and enhance the product. My budget was dictated by new sales, not existing customers.
In that environment, it was very tempting to develop new features at the expense of enhancing the existing ones. I managed to do both for a while, but the newer features were slower in development because of the resources I devoted to existing customers. This approach eventually cost me my job when new management came in, saw a potential cash cow, and didn’t like my approach.
Two years and one sale later, it was widely acknowledged in the industry that the product was, at best, a legacy solution. It has been dying a slow death since.
The Real Impact
My experience is limited, but not alone. I have seen the chase for the new feature while the core is left to slowly decay. I have seen some products linger while new, add-on products are introduced.
This is a massive problem faced my many older vendors. To drive sales, R&D is funneled into new features to stay ahead of competitors. Older features, like the ability to scale and not crash, get neglected once they are “solved”. When you take a product that is over 10 years old, that can lead to large pockets of very old code.
Of course, there is always the desire to sell to existing customers. What do you get a customer that has everything? How about a new product that seems like a natural add-on to the existing set of products. Instead of a product that is steadily becoming more feature-rich and capable, customers have to buy new features.
Eventually customers move on to another vendor that is “bundled” and they repeat the cycle. The only difference is that, unlike a mortgage refinance, you don’t really get to skip a year of paying maintenance. In fact, customers usually have double payments until migrations are completed.
This entire process is aggravated when Product staff are incented by license fees. If you incent by maintenance fees, they still want to sell to grow, but they also want to work hard to keep existing customers.
The Point
It is quite simple, it is a mess. The ongoing revenue stream should measure the success, not the one-time fees that means that you got engaged. It leads to products that are less complex to purchase and companies that want to keep business even more than winning business.
There are other aspects as well, so chime in and offer your viewpoint. My next post on the topic will discuss how all of this license fee focus is preventing ECM vendors from moving forward.
Word of Pie 
What are your thoughts on subscription models (commonly, but not exclusively, used by open source vendors) rather than the traditional big up-front license fee followed by 17% – 25% M&S model each year thereafter?
LikeLike
In theory, I am positive. I haven’t seen enough of the impact to make any real assessment.
LikeLike
Very good article summarizing the dilemma -I guess- every software company and its (product) management team is in. But you should not forget that every new license means also new maintenance and maintenance is really what a company keeps going in the long run. Therefore, you have to get the right mixture everywhere in your company: hungry sales people for new sales, dedicated professional services and support engineers that keep existing customers happy and find ways not only to retain them but enhance your company’s footprint and grow the user number of your product(s). ..And turn them into references that in turn help to win new customers. And of course you have to find the right balance between new sexy features and the not so sexy tedious house-keeping. Having customer panels and user groups may help to get this balance right. And of course: It’s all about money. You can only spent one Dollar (or Euro) once. Therefore it is important to establish the right incentive model that reflects your company strategy and together with good internal communication of the company goals will help that you will achieve long term success.
I also want to comment shortly on Peter’s question. I do not think that the license model trad. vs. subscription) makes a big difference because in the long run also companies wit a traditional licensing model end up with a 50:50 distribution between license and maintenance revenue. Of course it is nice to get a huge chunk of the money as soon as you sell. This is tempting as the recent move of Alfresco shows: They push to sell 3 year subscriptions that have to paid in total upon closing the deal. No big difference anymore to the traditional model.
LikeLike
The right mix is important. However, you can account for new sales, which are important, by measuring growth of that Maintenance revenue. It won’t grow without new sales and solid retention. 1% growth is not enough, but 10%+ might be when you consider that there will always be some loss of maintenance revenue. The key for measuring it effectively is determining what amount of loss is “normal” and then factor new sales on top. That requires numbers I just don’t have.
As for the 3-year subscription model. It seems like a good deal, but could lead to neglect of customers for 2 years. That said, the savvy account rep will realize that getting that year 4 money will be easier if they visit and chat throughout the 3 years.
LikeLike
First, thanks so much for one of the most interesting and thought provoking blog series I’ve read in a long time. I CAN’T wait for the next one based on the topic.
Any way, I know you said you were not going to get into the investor side of things so I’ll try and keep this short. When talking about how a company’s growth is measured we simply cannot ignore the different rules public and private companies are forced to play by. Right or wrong, most financial analysts look at new license growth as the speedometer for a company’s overall growth…and so do most tech saavy investors. Hence the focus, by public enterprise software vendors, on new license sales and the inevitable clusterf**k of R&D investment you describe so accurately. Until the markets change the way they evaluate a publicly traded independent software vendor it will continue to be a challenge to change their operating model.
LikeLike
Bring on the investor stuff. It is interesting. I just don’t feel I know enough to speak on it with any authority. More than willing to discuss it.
I feel the fixation on license revenue is brought-on by the software companies themselves. They spent years educating the market analysts that it was the best way to measure growth, but it isn’t. How much license revenue does Google have? It is going to take these Internet-based software vendors to drive the market to look at alternative ways of evaluating growth. Then the other companies might be able to make the transition.
The alternative is for a company to focus all of their incentives around sustained income and see what happens. License revenue may slip, but profit would likely increase over time, as would market penetration. After a few years, people would realize that it was everywhere and that maybe they are the ones to own, license revenue be damned.
LikeLike
Rainer, I think you’ve misunderstood how subscription models typically work. Commonly, there is *zero* up-front license fee – the purchaser simply starts paying M&S, and they continue to pay that same fee as long as they’re satisfied with the products and/or services they receive for that fee (assuming they don’t expand or reduce their usage of the product and/or services – something that also happens, but is unrelated to the mechanics of the subscription model itself). Here, this is commonly referred to as “every year is a voting year”.
Contrast this with the legacy model, where there’s a large up-front license fee, followed by something like 15% to 25% M&S each year thereafter. With such a large investment, customers will typically work harder to make software purchased under the legacy model work, even when it has become clear that it is not a good fit for their use case and/or simply doesn’t work. Yes this investment is a sunk cost, and theoretically should be ignored for the purposes of future decision making, but as behavioural economists have shown, that’s not actually how decisions are made in practice.
Having worked inside vendors with both pricing models, what I can tell you from direct experience is that this *massively* changes the dynamics within the sales team. The legacy model basically mandates that “new business is king” (since that’s where the big $$$ are) and as a result there is very little, if any, natural incentive to keep existing customers happy. After all, at 20% M&S it takes 5 years to equal the up-front license fee, and how many systems these days have a life time much greater than 5 years (particularly in such fast-developing areas as WCM!)?
The subscription model flips this balance of power on its head, since a renewal brings in the same amount of $$$ for the sales rep as new business, but (if things are running smoothly) at significantly lower (even zero!) cost-of-sale. If things are not running smoothly, then the rep has a vested interest in fixing whatever is broken, since the renewal is worth as much to them as new business (more even, since if they can get things moving smoothly they “win” that revenue year after year, with little to no effort).
Also, your comment about Alfresco and 3 year contracts is quite inaccurate (disclaimer: I work for Alfresco). While 3 year contracts are offered to prospective Alfresco purchasers, Alfresco does not (in fact cannot) recognise that revenue any faster than it would for an initial new-business subscription followed up by 2 renewals.
In fact all it really does is allow us (for 2 years only) to reverse the default “opt-in” for renewals (i.e. each customer’s default position each year is to not renew) to an “opt-out” model (i.e. the customer has to explicitly opt-out of renewing if that’s what they wish to do). This is a win-win, in that the customer gets a cheaper subscription (multi-year contracts are typically discounted) and Alfresco saves on cost-of-sale for the 2 years of renewals. In fact arguably the customer gets more out of it than Alfresco does, since often they can negotiate to retain their existing subscription price beyond the 3 years, even though our benefit ceases at that point.
LikeLike
Laurence, great series of posts. From reading the comments it’s tempting to think that the open source/subscription based model only differentiates from the license fee model on the basis of when exactly the vendor receives their cashflow, but I think it has more substantial implications than that.
At it’s worst the license fee model as used by the big vendors encourages the huge multi-month sales effort focussed entirely on ‘the big sale’. Often this is not what purchasers want. They don’t always know exactly what their solution is going to be and are not sure exactly which mix of products is right for them. They’d like to dip their toe in the water and try out a few. But some sales efforts are focussed on turning the small evaluation into a full scale pilot in the knowledge that if you get the customer to commit enough resources they will feel that they can’t back down as all the effort on pilot/proof of concept will be wasted.
The idea that subscription based, open-source models are more aligned with what customers want is , I believe, the message that Cheryl McKinnon of Nuxeo would like to encourage. Like you I don’t have enough experience with open source models to confirm that but it’s certainly a message I want to be true.
LikeLike