The education and research sector has a set of special requirements when it comes to cloud services. First we want the service to be compatible with SAML 2.0 based federated single sign-on to access the services. Secondly we run our own network. We don’t want to have network traffic/data transfer charged to us or have it as part of the cost structure. When it makes sense we pick-up the traffic at your datacentre ourselves. This results in a cheaper service with higher performance. Thirdly we want the best possible conditions. We believe that negotiation on a global scale will give us more leverage with suppliers to achieve this.

So far the general idea. The execution will prove to be somewhat complicated. First hurdle we have to take is to find out how can we accommodate country specific contract requirements with regards to privacy and other important clauses. Second big hurdle led to a big debate at the Terena conference. How will the global negotiated deal fit the various national public procurement laws.

Technology moves faster then laws. As a consequence we have to find a way to adopt old laws to new online services.

We all know that public cloud services are driven by consumerization. Even in the business to business market it is more and more at department level that service selection is done without consulting IT. According to Gartner 35% of cloud purchases are done without informing or consulting the IT department.

This leads to a couple of interesting challenges. Suppose 10 departments decide to buy the same service for 20k Euro each. Suddenly the company is bound by public procurement laws for that purchase. On a central level nobody knows that this has happened. There is no governance.

What will happen if you actually do publish a tender? My guess is that Innovative new services emerging in the cloud do not look at the TED publications. Their way of selling a service is customer self service via a web interface. By tendering you will probably not get a response from these companies. I even doubt that more established leading Iaas providers are not watching TED.
So if you are tendering Iaas, will you get the responses you want?

Another interesting point is the pay per use dimension of cloud service.
As many cloud providers say: adoption is the new ROI. This means that if they have a customer, but the customer is not using the service, the cloud provider doesn’t earn a single Euro. How different is this from traditional IT! The solution is to make an estimate. However difficult to predict usage of a service, we need to make an estimate of potential three years spend. If you end up spending more the 200k you should have seen it coming!

Educational institutions have called in SURFmarket to help them arrange calls for tenders. This is done by means of the in-house procurement model. By signing the Accession Agreement, the institutions have transferred their obligation regarding calls for tenders for ICT materials to SURFmarket (if this is included in their core package)

SURFmarket uses the Dynamic Procurement System (DPS) which is an online ease-to-use ICT marketplace for the Netherlands. Vendors can register any time they want during a certain time (four years). SURFmarket offers high market penetration, single point of contact, central payment on behalf of all institutions and technical delivery of cloud services.
This is how it works: vendors make an indicative offer which is assessed against the requirements by SURFmarket. Next step is admitting the vendors to the DPS. Final arrangements are recorded in an agreement for intermediary services. After closing the deal, users of higher education have on line access to the agreement for intermediary services en are allowed to order. When the orders do exceed the limit amount for European tendering a tenderprocedure is needed. But the procedure is a so called light version, compared to the traditional procedures.
SURFmarket supports various European tender procedures and also intends to gain experience with Best Value Procurement (BVP). BVP (developed by Dean Kashiwagi and his staff at Arizona State University) increases the importance of vendors to show dominant value using performance measurements of their key personnel and processes. It does not require the buyer to identify what is being procured at the beginning of the procurement and does not require the selection committee to have technical expertise. All submittals are non-technical in nature, and technical questions are not asked or discussed until after the best value