Towards a European Software Strategy:
|Oracle / ECIS||Cox||Karlfirstname.lastname@example.org|
|ATOS ORIGIN||de Lama||Nuriaemail@example.com|
|European Software Association||El Ramly||Siadafirstname.lastname@example.org|
|Triangle Venture Capital Group / EVCA||Geiger||Berndemail@example.com|
|Business Software Alliance||Mingorance||Francisco||FranciscoM@bsa.org|
|Open Forum Europe (OFE)||Taylor||Grahamfirstname.lastname@example.org|
|European Committee for Interoperable Systems (ECIS)||Thornby||Charlotteemail@example.com|
Few people perceive the software industry as capital intensive. Yet access to capital is one of the major critical success factors. Software funding is also very specific and difficult, given that this industry is unique and complex to evaluate, with resemblances to traditional manufacturing, intangible goods manufacturing and value added services. In Europe, observers and stakeholders agree that access to finance is even tougher.
The EU has already supported R&D in software showing some tangible results (European Technology Platforms for example). The selection of strategic areas such as embedded software, e-health or on line services is an excellent approach. Yet, few EU software projects did deliver successful products or new business models. In fact, investing in R&D is not an objective in itself. In the software industry, especially the European one, post R&D phases are even more critical. R&D is only a means to provide a competitive edge to the European industry. Strong R&D programs do not create value without a strong industry.
A more global approach of innovation is needed bearing in mind that the designed policies should ensure that only innovative companies with good commercial prospects get financing.
As in traditional manufacturing, this business consists of designing, producing and distributing products that are developed, maintained and improved thanks to the work of a large labor workforce over a number of years.
But, software is intangible. Contrary to manufacturing tangible goods, the marginal cost of developing software is almost zero. Thus, even if the initial costs are high due to investment in R&D, the cost of the production work is insignificant. The value chain of packaged software includes: - on the one hand: R&D, translation, documentation and packaging (all of which takes place before the production stage in a traditional manufacturing company)
- and on the other hand: marketing & sales, professional services (consulting and implementation) and other services (tutorials, support etc.), which represent a growing share of the costs of a software vendor when it matures and becomes “industrialized”
Like in all high value add services businesses, the costs of highly qualified personnel are by far the largest. In the packaged software industry, they represent around 70% of the value add (which is close to the IT services ratio: around 80%). Moreover, like in any services business with little intermediate consumption, this business requires a large working capital fund because: - on one hand, on average customers pay within 60 days - on the other hand, salaries and social security are paid monthly and make up the majority of a software vendor’s costs
As a result, the accounts payable do very often not offset the accounts receivables which represent a significant portion of the yearly revenue. Balance sheets are therefore very often unbalanced.
Finally, software vendors are faced, like all businesses in the world of intangibles, with the fact that investors often underestimate their intangible assets (human capital in particular) and are struggling to understand the sometimes complex economic models relating to the so critical issues of intellectual property. This complicates further the investment in software companies. Not long ago, software was not seen as being capital intensive due this intangible nature and its resemblance to the services business, where you do not need expensive raw materials or components to start companies given that human capital is the dominant expense. Investors that operate in this sector and know other sectors can easily explain that this industry is characterized with short development cycles and little basic R&D on one side and with long and cash consuming go-to market cycles on the other side.
All stages of development of software companies are critical: seed, early stage, development… Early stage is the most visible, but the development stage is even more important. A young software entrepreneur can create a company with a limited amount of capital and may find sufficient funds with venture capitalists to develop the first generation of products. It costs a lot of money to be competitive when it comes to packaging the offer and going to market. Too few European companies invest in time in marketing. Instead, they very often develop and sell custom software to fund further their packaged software development.
Once the Entrepreneur decides to develop his company he will need heavy access to capital. Without proper access to capital, how can he fund the next versions of his product? How can he fund his international expansion and face stiff price competition? How can he fund the acquisitions which will allow him to remain competitive in the market? Access to capital will directly determine how rapidly one vendor might grow. It is also critical to support the development of medium-size players which need to finance international expansion or external acquisitions. Too many successful companies miss opportunities due to the lack of capital and are acquired by global leaders once they become stagnant above € 20m sales.
As mentioned above, Europe is a less conducive place for innovation and risk than the United States. Even if some Northern European countries are much closer to the US model than the Latin countries, ecosystem clusters are still smaller and less vibrant than in the United States where innovative SMEs, universities (with massive private and public funding), venture capitalists and Business Angels work together to turn research into innovation and successful products.
Europe is also experiencing a deep structural crisis of venture capital. Here are some striking figures excerpted from the 29 September 2006 Communication of the European Commission “Implementing the Community Lisbon Program: Financing SME growth Adding European Value”57. After a strong decrease from €4.2 billion in 2001 as a result of the bursting of the technology bubble, European venture capital investment in early stage firms has stagnated at around €2 billion. The 10-year return on overall venture capital investments was 6.3% in Europe compared with 26% in the US. Business angels investments in Europe are estimated to be less than 10% of those in the United States. Again, some countries do well (UK, Sweden, Denmark) but the average performance is very low and this is not at all compensated by bank loans.
There are also software specific reasons for the weaker attractiveness of venture capital in this sector. As previously explained, the European packaged software market is quite consistent (even if there is more in house software in Europe) but the industry is very fragmented with few medium-sized enterprises, global players.
Success stories and serial entrepreneurs are rare, which leads to:
|-||a lack of visibility and attractiveness (which is very negative in a generally risk averse environment)|
|-||a lack of management skills, which is a matter of concern for investors. Investment readiness (how|
|-||to present a business plan in the correct way) has to be improved.|
|-||a lack of marketing skills, which results in low performance when turning solutions into industrial|
This insufficient anticipation and command of packaging and go to market phases generate a much less rapid
return over investment than in the United States in an industry that is supposed to deliver a rapid pay back. Venture capitalists also believe that late adoption of new technologies and innovations by the European economy has led these to higher services revenues compared to the traditional mix of revenues (1/3 licenses, 1/3 maintenance, 1/3 services) they are used to with US companies.
One important segment has fewer difficulties to attract investors, and this is the application software segment. This segment is less exposed to global competition; it is seen as more in line with stable vertical markets and is integrated in a clear value chain. Technical software products (Infrastructure and tools software) are, on the contrary, exposed to global competition as soon as they are on the market. This is a very cash consuming business to be competitive.
The market failure regarding access to capital for innovative SMEs has been identified by the Commission that called for action.58 The European Investment Bank did the same during the 2008 Small Business Act Consultation59. The decision in September 2006 to establish the Competitiveness and Innovation Framework Program (2007-2013)60 acknowledged that “Community financing has a role to play in leveraging support and providing complementary funding in order to tackle situations of market failure”. As we will see, a new promising instrument has been launched: the “High Growth and Innovative Facility”. Generally speaking, objectives and instruments of the new Competitiveness and Innovation Framework Program are complementary to European R&D funding schemes. But, it is of course too young to be evaluated.
European R&D funding schemes remain the main existing tool. But the share of research and innovation in European budgets are still weak despite recent progress. Moreover, as stated by the 2008 Aho report61 our member and described by companies and venture capitalists surveyed is the general feeling funding schemes are often “too far from the market” and miss the “high growth companies” (gazelles) target.
The European Commission initiative for a strategic use of pre-commercial R&D public procurement62 is also to be mentioned as it can be considered as an alternative way to bridge the gap in the funding of marketable research. In fact, despite the success of such programs as STTR and SBIR63 in the United States, “European public authorities do not fully utilize their considerable purchasing power to foster innovation through procurement of innovative services and technologies” as stated by the 2008 Aho report. As expressed by Venture Capitalists and the European Private Equity and Venture Capital Association (EVCA) in its response to the European Commission consultation on a Small Business Act in Europe, we believe that “facilitating SME’s access to public procurement is crucial to promote competitiveness, job creation and innovation (also benefiting public authorities). Public procurement can be an important driver for supporting R&D activities and encouraging innovation.” The “pre-commercial R&D public procurement” initiative should be encouraged provided it is based on fair and open competition with pricing at market conditions.
The Commission works to bridge the gap between capital markets and innovative SMEs through proposal. But so far, no specific action has been launched to address the specificities of software as a strategic sector.
Generally speaking, consensus has been built in recent years to “move up a gear” in Europe for innovation and access to capital.
In the software arena, a trend is raising awareness: consolidation has accelerated at the top of the market and some of the rare worldwide European success stories in software are already over. As a result, fewer success stories are as likely to lead to a lower amount of venture capital invested in the sector.
Of course, the current “credit crunch” is weakening European software players even more, the overwhelming majority of which are SMEs.
The main reason why Europe should act now has been stated by Commissioner Reding in her speech “Towards a European Software Strategy”: there is a window of opportunity for European software. But, the window is small and it will soon be closed if we don’t act.” The era of computers is gradually scaling down and paving the way to the predominance of the network. Europe failed to take the right approach to computers a few decades ago. But it has invested in networks and on line services research to be competitive in this new era. As previously seen, Software as a Service will probably be a major contributor to the industrial revolution that has just begun. It may enable today’s European challengers to gain position and become top leaders in tomorrow’s software market. But one has to bear in mind that European software companies are traditionally technology and R&D centric solution providers. Many of them have already added the SaaS delivery model to their offering, but it will require them heavy investment to compete with more business- and process-oriented global players once the market booms. Besides the potential establishment of ongoing revenue stream attracts investors. But at the same time, SaaS subscription model also complicates the recognition of revenue and the valorization of the created software. From now on, software players need more cash than ever to be “SaaS ready”.
The following recommendations are not all software specific but they can impact the software industry in a very positive way.
RECOMMENDATION 1: Benchmark and promote national public-private funding mechanisms adapted or specific to software within the framework of the European Software Strategy
It appears that joint initiatives between Government and private financial institutions (banks, insurance companies...) are the most efficient way to leverage public financing. At the European level, industry-led Joint Technology Initiatives pool public and private (at least 50%) investment in areas where existing funding mechanisms cannot deliver the scale and speed needed. However, money can be more easily found and put together at the member states level.
International benchmarking (both in Europe and outside Europe) can provide the European Commission with examples of initiatives which have been successful and should be promoted across member states to boost public-private funding in strategic innovative sectors for Europe, such as software.
In Scotland, several investment funds65 have been established to help businesses at different stages of growth. For example, the Scottish Venture Fund (SVF) has been set up to invest £500,000 to £2 million, alongside private sector partners, in company finance deals of between £2 million and £10 million. The SVF supports a broad range of industry sectors, typically high growth companies.
The France Investissement initiative66 is another interesting example of supporting innovative companies through a combination of public (€2 billion between 2007 and 2012) and private funding (€1 billion within the same period). After the first year, more than €800 M had been invested by France Investissement: €537 M in 44 ventures and development capital funds to reduce the capital gap between innovative companies and capital markets and €350 M in 802 companies (almost 40% of them being ICT companies including software vendors). Inspired by the Israeli Yozma67 program model, the share of private funding is expected to rise until 2012 up to six times as high as the public share.
Due to the strategic importance of software and the specific difficulties to bridge the gap between finance and industry, the launch of national specific funds dedicated to the software industry should be contemplated. It could be one of the objectives of the European software strategy. Some of these funds could support key European strategic objectives, for instance embedded software or “Internet of things”.
RECOMMENDATION 2: Create a European Software Fund within the existing instrument run by the European Investment Fund
The Competitiveness and Innovation Program (CIP) has increased the focus on the urgent need to strengthen innovative SMEs in risk/venture capital. The whole program is supposed to provide around one billion Euros through its financial instruments, which are expected to leverage around 30 billion Euros of new finance for SMEs. It added a new risk capital instrument to the existing instrument fostering SME start-ups. This non-grant-based instrument is aimed specifically at innovative and high-growth SMEs, which need capital at their crucial growth phase.
The new “High Growth and Innovative SMEs Facility” (GIF) created within the CIP’s Entrepreneurship and Innovation subprogram consists notably in funding venture funds that in turn invest in innovative SMEs. It is run by the European Investment Fund and is too young to be evaluated. Yet, interviews conducted with venture capitalists and member companies demonstrate that there is still very little known and not enough targeted on strategic sectors such as software.
Due to the very high needs, it should be made a strong priority to quickly increase visibility and, later, budget of the GIF 1 and GIF 2 programs. Critical mass is important to attract investors and increase visibility of this instrument. As the software sector is strategic for Europe and suffers from a bigger venture capital shortage, public policy should do a massive effort to correct the market failure. The way to do this is creating a specific program for software. Furthermore, a targeted sub-program will increase visibility of this instrument.
This European Software Fund could also be, together with the proposed ESEN (European Software Expertise Network), the appropriate framework to favor a better understanding between venture capitalists and software entrepreneurs. We welcome the 2008 Aho report recommendation69 to create such a platform between venture capitalist and all companies involved in FP7. As it appears that European software entrepreneurs very often lack management and business skills to present their project to potential investors in the most convincing way, the European Software Fund should include “do’s and don’ts” for entrepreneurs who decide to seek venture capital. As it was suggested in a recent workshop organized by the Commission70, professionally-run “investment readiness” programs could also be part of this measure.
RECOMMENDATION 3: Adapt European R&D funding schemes to software companies
Software companies experience the same difficulties to access to R&D funding schemes as SMEs in general (lack of information, too much red tape generated by calls, submission process, negotiation process, project management and reporting rules…). The existing programs are still too complex (although improved since 2007) and are still too often only accessible to the “happy few” companies that have the information and the know how to apply or to big companies that can afford relying on specialized fund raising cabinets. In that perspective, we support the proposal of the two-step approach of the 2008 Aho report71 consisting of giving a chance and a small amount of seed money to applicants passing an initial light evaluation.
We would like to draw attention on difficulties that are more specific to software companies and propose a concrete program to address them.
As previously seen, the innovation cycle in software is faster with less massive R&D than in other high-tech sectors. It hardly matches the pace of traditional European R&D projects. Very often, software companies already developed the technology and would like to focus on pilots, demonstrators and go-to-market. Furthermore, in the software industry, most of the early investments are not counted as “assets”: personal skills and knowledge pieces of code, algorithms. Experts and venture capitalists involved in the evaluation process often do not know the sector and are reluctant to fund companies with balance sheets that appear very weak.
The top-down approach of the preparation of Strategic Research Agenda, which defines top EU research priorities, is a hurdle for software SMEs that cannot be well represented in such a process. One of the interesting things about software is that there can be ideas of innovative software applications on every area of the economy and society that do not necessarily imply new technologies. These projects should be given a chance as they very often meet demand and can generate rapid growth and jobs.
IP issues are also particularly sensitive. When contemplating whether to enter into European projects, software SMEs do not see how they will have a satisfying return on projects, even in small and medium sized collaborative projects, as protective measures imply a full and complex process.
To tackle these issues, we propose to focus on the following “innovation oriented” priorities.
All companies fitting in the new definition of “Innovative enterprises” should be eligible to receive R&D funding.
Then, a new balance has to be found between R&D and innovation in the evaluation criteria and the eligible scope of expenses to be covered. Progress has already been made with the Eurostar program that is more oriented around business issues. Unfortunately, it suffers from different rules and coverage rates in 31 countries.
Furthermore, access to the European market as a potential “second domestic” market should be one of the objectives of this policy. As previously seen, the costs to cope with the complexity of the European market are high. All measures reducing these costs are more than welcome. Software is in essence a complex product that includes “Friendliness”, technical ability and multiple dimensions that can be summarized under the now widely used “User experience “term. In Europe, due to all the different languages and cultures, this “User experience” includes multiple translation and packaging issues that are a major investment to carry out and to learn how to carry out.
We suggest that the decisive later stages of the innovation cycle (validation-pilot, go-to-market both locally and beyond domestic borders within the EU) should be covered by grants. Once this is done, the weighting in the evaluation criteria of the “impact” of a project (dissemination and use of project results”: standards, publications, recruiting distribution partners in other EU countries etc.) should be raised.
In a step-by-step approach, we propose that a new ambitious program called “European Software Innovation Program” (EUROSOFT) should be tested on a small scale for software “innovative enterprises” with:
|-||a focus on delivery of new products or new services for existing software products that have proved|
|to be locally successful|
|-||grants to cover all later stages of the innovation cycle: validation-pilot, go-to-market|
|-||a high rate of coverage for these non R&D expenses: establishment of a business plan, translation|
|costs, packaging and product marketing costs (cultural adjustment costs etc.)|
|-||100% coverage of IP costs|
|-||a bottom-up approach (no compliance to the Strategic Research Agenda and no guidelines on how to|
|build the consortium, only an obligation to share the benefits)|
|-||a two-step approach in the evaluation process (as proposed by the recent Aho report)|
|-||an involvement of successful entrepreneurs and managers (notably via NTAs) in the evaluation|
|process (at least one per evaluation) to balance the technical point of view of experts. Other hands-on|
|stakeholders such as venture capitalists, Business Angels or national and regional innovation support|
|agencies should also be involved|
|-||information and visibility provided by NTAs|
|-||a “EUROSOFT” label to be awarded software companies participating to the program and meeting|
|criteria such as usability of products in different European languages and cultures|
|-||an incentive for venture capitalists to fund the “EUROSOFT” enterprises|
We do believe that such a program and a label would help generate successes throughout Europe.
BSA believes that software innovation is a driving force for economic, social and technological progress. To ensure continued innovation and a diverse software marketplace, products should be forced to compete on their merits, and to be developed in response to consumer demands rather than government mandates. ICT procurement policies should reinforce this diversity by fully respecting the principles of technological neutrality and non-discrimination which underlie Community procurement rules.
Issue: How can the EU best ensure that both it and its Member States and trading partners conduct technologically neutral, non-discriminatory software procurement, in order to promote a competitive, diverse software market?
|-||BSA has long believed that when software is procured on the merits, it encourages firms to compete and|
|innovate; in contrast, preferences for one solution or business model over another can chill innovation and|
|undermine diversity.||Fortunately, Community procurement rules are founded on the principle of non|
|discrimination, and guarantee neutrality as to technology, vendor and business model.||The Commission|
|should ensure that these core values of Community procurement law are robustly reflected in the|
|-||We also encourage the Commission to avoid duplicating any other initiatives underway in this area (such|
|as DG Enterprise’s ICT standards initiatives or the IDABC EIF process) -- any future software strategy that|
|addresses procurement matters should take a fresh perspective.|
We believe there are several significant trends that the Commission should carefully analyze when considering what recommendations, if any, should be made concerning software procurement:
-Europe’s software market is increasingly characterized by hybrid business models and solutions. In order to meet customer needs, software companies now routinely deploy products from a range of development and business models (proprietary, open source, implementing open and/or proprietary standards, etc.). Competing firms frequently work together in the pursuit of new innovation, disclosing information, licensing technologies and building on one another’s solutions to develop and market complementary product offerings. When distributing their innovations, technology providers often employ hybrid business and licensing models that incorporate aspects of various development and distribution regimes.1
-There is an increasing willingness in Europe to utilise procurement to promote particular business or standards development models. Most recently, EU initiatives -- in particular the EIF -- have sought to encourage Member States to prefer products that implement narrowly-defined open standards, regardless of whether those products in fact provide better functionality, greater interoperability or are otherwise stronger on the merits or better value for money. Individual Member States have followed the EU’s lead, adopting formal policies in favour of such technologies. This trend threatens to exclude many popular solutions based on widely-recognised open standards from public procurement in Europe, in particular by
1 To give only a few examples: mySQL AB, a venture capital funded software enterprise, provides a dual license (open source and proprietary) for its product; and some open-source technologies can be found in mainstream proprietary software applications developed by companies like Adobe, Real Networks and McAfee among others; Microsoft is engaged in many direct collaborative efforts with open source software companies, has provided open access to the resources of Microsoft's Open Source Software Lab to enhance interoperability with OSS products, and has undertaken a variety of initiatives to meet customer needs for source code access.
excluding technologies implementing RAND-based standards. It is also potentially inconsistent with the fundamental principles that underlie European procurement rules.
-EU polices are influencing public procurement around the world. As a result of the EU’s promotion of IPR free standards in public procurement, third countries have adopted policies that will disadvantage many leading firms. For example, India’s recent draft procurement policy has been directly inspired by the Commission’s draft policy in EIF v2.2 Furthermore, at the recent Internet Governance Forum (IGF) in Hyderabad, representatives of the South African and Brazilian governments announced that they would adopt the Commission’s approach in their ICT procurements. 3
Community procurement law works well in practice, ensuring that companies can compete on the merits for government contracts -- in a technologically neutral and non-discriminatory manner -- and that procuring authorities can obtain the products that best meet their needs. In doing so, procurement policy supports the diversity that is the principle strength of the European software market. To the extent that barriers exist to technological neutrality and non-discrimination in procurement, they are largely policy-related, and are not a result of problems with the legal framework.
-As described above, efforts to use procurement to promote particular business or standards models has been among the primary barriers to ensuring technologically neutral and non-discriminatory procurement. In practice, this may lead to the exclusion of the hybrid solutions that are at the forefront of software development today. It may also lead to the exclusion of proprietary technologies that meet the needs of the procuring authority in an “equivalent” manner, infringing one of the key principles of Community procurement law and harming diversity in the software sector.
-Furthermore, some policymakers continue to promote an overly narrow definition of open standards, and to focus more on whether a particular standard satisfies that definition than on whether it achieves interoperability or otherwise offers better value for money. The vast majority of ICT standards incorporate essential technologies subject to IPRs. Government policies that seek to exclude such standards from public procurement will exclude many innovative European software products. This discrimination will harm growth and innovation in the European software market, in addition to being inconsistent with Community procurement law.
-Finally, some policymakers also appear to believe that uses of multiple standards in procurement are anti-competitive and hinder interoperability. In fact, the opposite is true. Public entities should be able to procure the software that best meets their needs -- based on functionality, performance, security, value, and cost of ownership. This is facilitated by avoiding preferences for a specific standard or any categorical preferences for open source software, commercial software, free software or other business models.
Addressing the identified barriers will, we believe, have the following positive effects:
-Reinforce technological neutrality and non-discrimination in ICT procurement. Application of these principles helps maintain diversity in the European software sector, which is the industry’s key strength.
-Promote diversity in the software sector, which in turn will drive innovation in Europe. Allowing multiple software business and licensing models to compete on the merits is the best way to promote software
innovation and ensure that customers -- both public and private -- have a range of choices in software procurement decisions.
If the Commission decides to include a procurement section in the Software Strategy, we would encourage the following action:
-The impact of EU policies on European businesses operating in foreign markets should be assessed. As noted above, EU initiatives on procurement in the ICT sector are influencing the policies of other governments. But such policies could ultimately prove harmful to European businesses competing in those markets. Careful consideration should be given to the effect of current EU policy on the legislative landscape of other countries and the environment in which European businesses must compete.
There is an ongoing debate in the sector on the role of "open standards". While many actors support "open standards" in principle, there is a disagreement on the proper definition of the term open standards.
EICTA, the European ICT Trade Association representing more than 10.000 ICT businesses in Europe, has adopted in its White Paper on Interoperability the following definition of open standards:
"Control: the evolution of the specification should be set in a transparent process open to all interested contributors Completeness: the technical requirements of the solution should be specified completely enough to guarantee full interoperability Compliance: there is a substantial standard-compliant offering promoted by proponents of the standard Cost: fair reasonable and non-discriminatory access is provided to intellectual property unavoidably used in implementation of the standard"
While these criteria encompass the full range of issues relevant for an open standard, some specific aspects of the process of open standards development can further be emphasized:
ECIS has long supported a definition of "open standard" which includes the following characteristics.
In ECIS' view, open standards are characterized by :
o collaborative and democratic development and management processes; o transparent evolution and management processes open to all interested parties; o approval through due process arriving at consensus among participants; o implementations which interoperate among each other; o platform-independence, vendor-neutrality, and unrestricted numbers of competing implementations; o open and complete publication of specifications and documentation sufficient for fully independent implementations; and o royalty-free or FRAND licensing terms that do not discriminate against the open source software development or licensing model. • While the EICTA definition addresses a number of these issues, the important point of non-discrimination against the OSS development or licensing model are not included in the EICTA definition.
The prime focus of this submission is Public Procurement. Financing Innovative Software has not been a prime focus of OFE so any comments made would be superficial and not evidenced. Nevertheless(!) a few points have been noted at the end, outside the main text.
Public procurement is probably the most important aspect in terms of any analysis being undertaken across any work group. Not only is procurement where policy meets practice, and where 'fine words' have to be delivered, but for the supply industry (of all size) it will have the greatest single impact and will have the greatest influence on their own direction and commitment. Public sector in Europe is by far the largest purchaser of solutions in Europe. Equally SMEs account for 90% of the European Software Industry. These two aspects cannot be ignored in any recommendations.
OFE broadly supports the four themes identified in the earlier calls, and has focused its comments on each of these – firstly in summary:
We would observe that currently the European software and services industry is heavily weighted in favor of small SME organizations, yet current government procurement practice is equally heavily weighted towards awarding contracts to large consortia and/or with heavy financial penalties. Such an approach, whilst understandable in the past does need review because of new business models, the move to software as a service, and the impact it is having on European suppliers.
Ref: OFE Brief No 2 (09.02.09) SMEs
• Ensuring fair procurement practices which avoid preferences for specific development models or individual vendors.
We support that government procurement is based on a level playing field so that no business model is disadvantaged without reason. However, whilst opponents may well plead the opposite, the current position is stacked heavily in favor of incumbent large proprietary suppliers, and lock-in has occurred to a significant extent precluding new suppliers or alternative models. OFE past analysis indicated this may be true in 90% of current public sector organizations. Having a level playing field is more than just having fine policies and strategies, it is the procurement phase where it matters and can be realized. OFE, together with partners, has developed a scheme under which such lock-in can be identified and avoided in new procurements and developments. This scheme (Certified Open® ) is currently being reviewed and trialed in partnership with a European national government. Certified Open is independent of business model.
There are occasions where an open source based development or procurement can be justified in preference to one based on proprietary models – these have been well identified in the current consultation paper for EIF Version 2, and we will not try to rewrite that text or repeat those arguments here. For the purposes of this input, however, OFE suggests the focus should be on the avoidance of lock-in to any one supplier, as described in Certified Open.
Equally there is concern that even when current legislation in Europe exists it is not being followed, to the detriment of European (small) suppliers. Notably in 2008 OFE undertook a survey of public procurements to measure the level of observance to existing legislation which except in specific circumstances prohibits government from mentioning trade marks in tenders. The OFE study indicated some 25% of all tenders broke such rules, and this may be just the tip of the iceberg. The net result is that existing suppliers are favored and new, potentially innovative, more effective or lower cost solutions are excluded.
Refs: Certified Open website
OFE Procurement Study October 2008
• Promoting open standards based interoperability of e-government applications
Unsurprisingly(!) OFE is fully supportive of the use of open standards in the area of interoperability, which is a fundamental area in both avoiding lock-in, and in encouraging open, competitive choice. Open standards are as relevant to proprietary solutions as they are to open source solutions. We would observe, however, that it is not just a factor of interoperability between e-government applications but equally so with citizens, and business. It also is one of the building blocks that allows localization of integrated solutions, including those consisting of both proprietary AND open source solutions (the majority). Again we would point to the work in IDABC in the development of EIF version 2. This includes the controversial area of definition for an open standard. OFE does not support the definition proposed by EICTA, but does support both the definition currently proposed by IDABC in EIF, AND the recognition of the openness continuum explained in EIF V2 consultation. It would be unusual if one part of the Commission suggested one definition whilst another suggested something else? A key factor for OFE is that the definition must allow all business models without discrimination. The EICTA proposal does not appear to achieve this. This Working Group (or any other) is unlikely to resolve this difference, so we would suggest the differences of view are noted but no attempt is made to resolve.
Ref: OFE Briefing Paper No 1 “The Importance of Standards to Interoperability”
• An 'open procurement ' model allowing open, competitive choice of supplier both initially and downstream to take advantage of new innovation, lower cost is unchallengeable in theory and potentially in practice.
OFE does not feel competent to offer specific comment and notes only the following broad points:
|OFE Procurement Study||www.openforumeurope.org/press-room/press-|
|OFE Brief No 1 Standards||www.openforumeurope.org/library/onepage-briefs/ofe-open-|
OFE Brief No 2 SMEs www.openforumeurope.org/library/onepage-briefs/ofe-sme-2009- final.pdf/view
In WG7 for the European Software Strategy, the issue of procurement and discrimination against Free Software (Open Source) came up. Mr Anglard, the chair of WG7 suggested sharing these parts with you so that you could include them into the work of your group.
Please find the corresponding Issue:
Experience suggests that lack of interoperability consumes around 30-40% of IT budgets in both the private and public sector. Since procurement calculation generally does not account for “decommissioning” or “exit” costs from a particular solution, a procurement decision for a specific solution often establishes a strong bias in favor of the vendor of the first solution for all consecutive tenders. This violates European legislation which mandates vendor neutrality based on transparency and non-discrimination. 
Examples for this practice were highlighted by a recent study  of Open Forum Europe (OFE), which scanned 136 tenders for trademarked names. 25% of these tenders were specifically requesting trademarked products, violating the principle of vendor neutrality. A particularly drastic example is a 2007 tender for 34 million EUR by the Bulgarian government , which attracted public criticism  because the per-license cost of the proprietary software involved obtained through the tender was above the regular sales price available in stores in Bulgaria. The exact terms of the resulting contract cannot be assessed due to lack of disclosure caused by a claim for trade secrets by the vendor supplying the licenses.
Experience suggests that many such cases remain undetected due to lack of translations, which themselves constitute a violation of EU procurement rules. The above example from Bulgaria only became public due to the watchfulness of a native speaker.
 http://www.osor.eu/news/hidden-cost-of-proprietary-standards-may-lead-to-illegal-tenders http://www.osor.eu/news/it-open-source-group-protests-proprietary-software-deals
http://ted.europa.eu/Exec?DataFlow=N_one_doc_access.dfl&Template=TED/N_one_result_detail_curr.htm&d ocnumber=257996-2007&docId=257996-2007&StatLang=EN, the tender is archived at http://www.pressboard.info/preview.aspx?articleid=92793
and the corresponding Action:
Procurement policy review
There is currently no reliable way to assess decommissioning or exit costs from an existing proprietary solution. Being able to calculate these costs would bring transparency and allow a truly non-discriminatory evaluation of the competing offerings.
Providing guidance and metrics on how to assess this cost in combination with recommendations on how to reflect strategic goals for the IT infrastructure in tenders would help reduce one of the largest barriers to OSS adoption.
It seems that the calculation of exit cost - while beneficial from the perspective of Free Software companies - would also be helpful for other market participants, especially SMEs.
Bernd Geiger (EVCA)
OFE says: "The VC community is still split between those who insist on IPR based models, and those who have moved on - both to open source based models, and service based models."
That insinuates that some VCs are behind the times whereas others understand and do the right thing. For the VC industry, nothing is more natural than to follow the money. There are no ideological hurdles to favor one or the other model (although in reality the models are mixed anyway). If there were more money to earn with the "open" model, all VCs would probably "move on" quickly. Information about how the VC process works, especially in light of "non-classical" business model (inquired by Rigo Wenning) has been given to the WG3. Please find the contents of the email below.
STATISTICS OF VENTURE CAPITAL COVERAGE OF OSS COMPANIES
US: 3.4% (232 OSS out of 6,801 total; trend from 1.2% to 5.5% in ‘08) EU: 1.1% (39 OSS out of 3,435 total; trend from 0.7% to 1.0% in ‘08)
You might change the statistical approach, but it would not get you in the double-digit range for OSS companies having received venture capital.
THE VENTURE CAPITAL PROCESS / INVESTMENT MODEL
Second, some general remarks about the nature of VC investors: VCs want to invest in something unique which leverages their investments over time - almost always through entrepreneurial activities. The multitude of "unique things" to invest in and "entrepreneurial approaches" already show that there are many different strategies for making VC investments. There is one common denominator among all approaches: the unique thing VCs want to invest in. VCs are extremely keen that it stays unique and does not dissipate (because that's what VCs put their money in - it MAY also explain why the "classical" business model is prevailing). Taking into account that the VC process consists of these two components of "uniqueness" and "entrepreneurial activity", it also becomes clear that there is no single divide between classical and non-classical business models - rather, the success of a venture depends on a combination of different contributing business components.
NON-CLASSICAL BUSINESS MODELS WHICH ATTRACT VENTURE CAPITALISTS
What does that mean in practice? In the deal sourcing process VCs look for these unique things and, depending on their strategy, the search might be focused or very open (opportunistic). An example: a provider of special information services has pitched to a VC whose investment strategy is general IT. The information service hasn't been seen before (the "unique thing"), the information can be produced cheaply (e.g., by the use of various OSS components), and the VC invests so that the service improves, attracts and serves more customers. The ultimate goal eventually is to be acquired by a much larger company who wasn't aware of such a service until the startup made it apparent through the ever-increasing user community. These strategies, however, have some immanent disadvantages (though the VCs take them into account because of the extraordinary expected gains). After some time, the entrepreneurial component battles with the dissipation effect (e.g., key people leaving to start a clone, etc.). One good example of this effect is Facebook and all its competitors. Subsequently, the stochastic leads to the winner-takes-all effect. A prominent example here is eBay - which also shows that the winner-takes-all effect may not last but that by that point the VC typically cashed out long before (as they are builders, not maintainers). On the way to the tip of the pyramid, smart VCs (who won't make it) typically sell to the expected "winner-takes-all" (e.g., Alando); investments made by the not-so-smart VC go bust.
Ironically enough, most of the heavily disputed patents evolve out of aforementioned strategy in order to prevent too easy dissipation, unique or not, so unique business models with a less technical basis are patented to increase the chance becoming the winner-takes-all.
I'm stating the obvious when I say that internet communication increases the potential number of new business models enormously. But the days of easy exploitation of unique business models seem to be over. Innovative, intelligent (non-commodity) software that bears some uniqueness is necessary to offer added value in almost all "non-classical" business models in order to be unique enough to be attractive to VCs.
NON-CLASSICAL BUSINESS MODELS WHICH DO NOT EASILY ATTRACT VENTURE CAPITALISTS, AND EXCEPTIONS
There are many other non-classical business models, e.g., those who provide services instead of proprietary software packages. The problem inherent in non-automated services is that they are not scalable (not to be confused with SaaS which has more an architectural character). Thus, VCs typically avoid them, as their entrepreneurial activity cannot be leveraged.
Sometimes the uniqueness is a perceived trend, where VCs do not yet fully understand the economics behind it, but believe in the avalanche effect or hype. The sheer belief of all the participants in the hype moves enough money that there will always be some winners (and many losers).
HURDLES FOR NON-CLASSICAL BUSINESS MODELS
For all the non-classical business models holds, the entrepreneurial component in the VC process is more important than for classical business models (again, because the unique component does dissipate easily). Europe, with all its regulation and state interventions, is not the most fertile ground for growing entrepreneurs. If an OSS-based company struggles with an unjustified granted patent (in its opinion) then this is a cost issue (to find a different solution). No VC would bother with it because when in doubt, the entrepreneurial process drives the success (the "engine" behind it - the software might be replaceable; that also goes along with VCs conventional wisdom that execution and thus management is the key to success). A sufficiently large number of entrepreneurs to select from and the right environmental conditions without too many restrictions to conduct the business will determine whether a good idea flies or crawls. In other words, to increase the number of "gazelle"-type companies, the political imperative must be tuned towards an entrepreneurial society.
* DowJones VentureSource Database **Number of companies which received at least one VC investment in the particular year ***Number of companies which refer to OSS in their self-portrayal
Bernd Geiger - EVCA
In contrast to manufacturing tangible goods, the marginal cost of developing software is almost zero. Thus, even if the initial costs are high due to investment in R&D, the cost of the production work is insignificant.
-> The use of "developing" might misleading, as in the same sentence the acronym R&D is used with different attribution.
Not long ago, software was not seen as being capital intensive due this intangible nature and its resemblance to the services business in which you do not need expensive raw materials or components to start companies given that human capital is the dominant expense. Investors that operate in this sector and know other sectors can easily explain that this industry is characterized by short development cycles and little basic R&D on one hand and long and cash-consuming go-to-market cycles on the other.
-> That generalization might not hold true for the very early innovative software sectors when all stakeholders always knew that software is either over-engineered or needs too much engineering to get it to work.
-> A very important point has not been mentioned in the comparison between EU and US. Until SOX (Sarbanes-Oxley regulation after Enron), the US/Canada always had a very vibrant public stock market with risk-taking investors - investment funds and asset managers who invested on behalf of others, e.g., pension funds and private individuals. This was the main cash supply for the expansion phase of software companies.
-> Furthermore, restrictions of the VCs vis-ą-vis software companies are not only influenced by the ongoing consolidation (as stated in 1.5) but much more so by the missing creation of new buyers on the dried-up stock market due to SOX (no new, large software companies could be created to buy smaller software companies at an earlier stage which is the main exit route for VCs - IPOs typically accounted for less than 25% of all exits). Adoption of similar capital markets in Europe never really came to life - markets like the "Neue Markt" weren't build up again because of a risk-averse public and lack of available institutional money to be invested.
Europe is also experiencing a deep structural crisis of venture capital.
-> Some industry practitioners believe that this is because of the first cycle of venture capital in Europe (compared to the 4th+ cycle in US): Similar to the credit crunch, cheap liquidity and excessive state intervention tried to shorten the learning curve of EU VCs in the first cycle which lead to the said structural crisis.
"(how to present a business plan in the correct way)"
-> This may be relevant to banks, but not at all for VCs - what Europe lacks is the right number of entrepreneurs who present their ideas (this may even mislead politicians to spent money to teach people present a flashy business plan - that is not a solution).
One important segment has fewer difficulties attracting investors, and that is the application software segment. This segment is less exposed to global competition; it is seen as being more in line with stable vertical markets and is integrated in a clear value chain. Technical software products (infrastructure and tools software) are, by contrast, exposed to global competition as soon as they are on the market. This is a very cash-consuming business to be competitive.
-> The numbers draw a different picture, and with them one can deduce that investors aim for global markets. No one wants to be stuck in a regional niche - there is no upside! The facts support that: VC coverage of the last 5 years (all IT, no hardware):US: 6,7341; EU: 3,695, and for application/vertical type of IT markets: US: 1,257; EU: 708.
-> It might be far beyond the scope of the WG to dispute if there is something like market failure, some believe this is always the effect of too much intervention. However, it might be necessary to point out what the root causes are for a less developed European software landscape. This was mentioned in the bullets of 1.3 and the following sentence: "Venture capitalists also believe that late adoption of new technologies and innovations by the European economy has led to higher services revenues compared with the traditional mix of revenues (1/3 licenses, 1/3 maintenance, 1/3 services) they are used to with US companies." The latter can be clearly improved with a progressive public procurement strategy (without ideological preferences) as stated.
But at the same time, the SaaS subscription model also complicates the recognition of revenue and the valorization of the created software. From now on, software players need more cash than ever to be "SaaS ready".
-> That depends also on scale - the SaaS model is for many VC financed startups an easy option to monetize, especially to overcome excessive hurdles to market entry (i.e., sell it over the internet) - but many larger buyers opt for a one-off payment rather than a (in the extreme case) "per transaction fee".
It appears that joint initiatives between government and private financial institutions (banks, insurance companies, etc.) are the most efficient way to leverage public financing.
-> Banks and insurance companies are the least likely candidates to play a role. As already stated about the US model, private equity and venture capital firms run by privates are the most likely players to fill the gap (maybe in conjunction with government initiatives).
RECOMMENDATION 2: Create a European Software Fund within the existing instrument run by the European Investment Fund.
-> In the end, every direct subsidy skews the market. Indirect subsidies strengthen the demand side through public procurement and money matching of private investments may contribute to the growth of the economic subsector. In the US, nascent entrepreneurs become real entrepreneurs by knocking on the doors of the many private institutions that potentially supply them with funds. In Europe, nascent entrepreneurs become a kind of civil servant by going through all the submission processes. The conception of "public policy should do a massive effort to correct the market failure", e.g., by telling the entrepreneurs the "do's and don'ts..." to "seeking venture capital" is not easy to follow from a venture capital perspective. That might be one of the reasons why investors in VC funds tend to shy away from Europe - it stifles entrepreneurism than to sprouting it.
Input from CompTIA to WG4 on support to maintaining European neutral approaches to
European principles of software procurement for government infrastructures
Procure software on its merits, not through categorical preferences
All software products offer varying benefits and costs. Public entities should procure the software that best meets their needs and should avoid any categorical preferences for open source software, commercial software, free software, or other software development models. Governments are best served when they can select software from a broad range of products based on such considerations as value, total cost of ownership, feature set, performance and security. Governments should let the marketplace continue to encourage innovation in software development and should avoid intervening through preferences or procurement requirements that would discriminate in favour of one model over another.
Promote broad availability of government funded research, for all kind of software development models
For many years, governments have made important contributions to technology by funding basic software research. When public funds are used to support software research and development, the innovations that result from this work should be licensed in ways that take into account both the desirability of broadly sharing those advances as well as the desirability of applying those advances to commercialized products. The dissemination of results broadly in this manner contributes to a sustainable cycle of innovation in which government funding for basic research advances the set of knowledge available to the public while helping spur advances in commercial products. These products in turn create the jobs, profits and tax revenue necessary to fund future rounds of public research.
Promote interoperability through platform-neutral standards, without any preference for one or the other software development model
Voluntary, industry-led standard setting is the most effective way to develop platform-neutral and market-based standards. When these standards are open and available to all through reasonable and nondiscriminatory licensing they help developers to create products that can interoperate with each other. It is important that government policy recognize that open standards - which are available to any software developers - are not synonymous with, and do not require, open source software either for their adoption or utility. Developers of commercial software that may not typically publish their source code often contribute technology and intellectual property needed to develop new standards. Governmental policy on software standards should not discriminate in favor of or against any particular software development model.
Maintain a choice of strong intellectual property protections, as widely used by all kind of software companies, including SMEs
Policymakers should not make rigid intellectual property licensing choices a precondition for eligibility for procurement, nor should they discriminate between developers that choose to license their intellectual property on commercial terms, and developers that choose not to charge licensing fees. Commercial and community-based software developers both typically rely on intellectual property rights, though some seek compensation for their exercise of intellectual property rights while others refrain from charging fees. Allowing rights holders to offer a range of intellectual property licenses promotes choice and furthers innovation.
3. Public procurement contracts at national and European levels should be accessible to SMEs.
The members of EICTA welcome the opportunity to partner with public authorities in EU funded contracts such as the Instrument for Pre-Accession Assistance. This letter is to seek your assistance in correcting problems arising from inappropriate model contract terms and conditions imposed by the European Commission on certain EU funded projects. Such conditions can be contrary to the interests of contracting authorities in the public sector and vendors alike. We believe and are concerned that the current contract terms used for procuring information and communication technology (ICT) services with EU funding are in many cases counter-productive in that they discourage a number of potential suppliers from bidding and ultimately result in more expensive and less productive projects for the contracting authorities. Compared to market standards the contract terms in question (General Conditions for Service Contracts for European Community External Actions and, to a lesser extent, General Conditions for Supply Contracts Financed by the European Community) are unreasonably burdensome on the suppliers and not commensurate with the nature of modern ICT projects, which generally require a high degree of participation and involvement on the customer side. Examples of such burdensome terms include the assignment of all intellectual property rights to the customer, unlimited liquidated damages (in the Service Contract), subjective determinations by the customer project manager and other onerous terms (please see examples overleaf). The fact that these contract terms are used on a non-negotiable basis further compounds the above challenges by excluding any contractual dialogue/negotiation which could otherwise alleviate some of the above concerns and ensure that the final contract is balanced and fit for the purpose at hand. In our view the contract terms used for externally funded projects should be reasonable and balanced, thus motivating as many suppliers as possible to bid for the projects and thereby supporting the overall success of the funding programme. We would therefore welcome the opportunity to discuss the above with you in more detail and are of course prepared to help in reviewing the current contract terms with a view to identifying the problematic areas and suggesting appropriate amendments. I will be contacting your office in the coming days to coordinate a meeting with you, should you be interested in discussing further the issues raised in this letter. Yours sincerely, Leo Baumann Director – Public affairs EICTA
"It is thus important to ensure that these incentives are created (for the deployment of open standards for interoperability), and in this regard the public sector has an important role to play. The public sector is not only one of the most significant users of ICT, they are also in a position to influence the purchasing of private sector customers who need to interact with public authorities. Thus, the public authorities have a duty to follow a procurement strategy that encourages innovation and competition. They can do so by ensuring that the products they purchase are interoperable with products from other vendors, thereby ensuring that (i) the authority is not locked into a single vendor’s product, and, more importantly, that (ii) consumers and companies dealing with the authority are not forced to use the same software as the authority. This duty to promote interoperability can be discharged by the public authorities by insisting that their suppliers implement open standards in their products. By so doing, the authorities will create incentives even for dominant companies to become more open, as a decision not to do so would cost the company the significant revenues it gains from the public sector. After the dominant companies adopt open standards, competition will be reinvigorated as it becomes possible to gain market share and revenue by offering customers innovative new products that are differentiated from those of the competition.
Once defined, public authorities in Europe should be asked to require such open standards to be supported in the products they acquire, regardless of what type of ICT products they are purchasing, be it office applications, server computers, accounting software or web application subscriptions"
Here are a few points which ECIS would suggest for inclusion under the points listed in your note :
Find below some explanations on what we mean by issues, trends, barriers, etc.
*Issues:*/ Meant to capture the main problems /
*Trends:*/ Meant to capture the current and foreseeable trends (within and outside of Europe) /
• moves toward a shared services model is increasingly appealing to public administrations seeking to reduce costs
*Barriers:*/ Meant to capture the barriers at the source of the highlighted issues.
*Benefits:*/ Meant to capture the benefits achievable if barriers are removed and issues fixed or at least improved.
*Actions:*/ Meant to capture the proposed actions in order to remove or lower the barriers/