Government Procurement Service ADDSS problems – what went wrong?

A few weeks ago, the Government Procurement Service had a major problem with their Applications Development, Delivery and Support Services (ADDSS) framework, worth up to £1bn over a potential four-year lifespan.

The Channel got hold of a letter to suppliers from Stephen Guy, the GPS Director of Sourcing,

“The ADDSS procurement was affected by hardware capacity issues for the servers running the eSourcing system, which affected suppliers’ ability to upload their submissions”, he said.

In response, GPS asked suppliers to re-submit documents manually but Guy said compliance checks “identified incomplete and in some cases altered tender responses from suppliers”.  We’ve learnt that not only did some suppliers submit different information in hard copy, others refused to comply or even immediately threatened legal action. So in the end GPS really had no choice but to pull the whole procurement process, affecting the 400 firms who were involved.

The Cabinet Office put out a statement, saying,

“”Due to hardware capacity issues affecting the Application Development, Delivery and Support Service (ADDSS) procurement, caused by unusually high volumes of supplier responses, the decision has been taken to pause this procurement so that we can review the current situation and consider future action. This is an important procurement and it is essential we take the time to get it right”.

“Hardware capacity”? Aren’t we in the days of unlimited cloud based capacity, all available at the flick of a switch? And how were GPS going to make sure this wouldn’t happen again? So we asked GPS a few more questions, and to their credit, they’ve given us some answers. Here is their full response.

“As you are aware, the ADDSS procurement was affected by hardware capacity issues for the servers running the eSourcing system. The servers are provided through the agreement with IBM Emptoris.

The sourcing tool is widely used across the public and private sector for procurements of all sizes, and we have successfully managed a number of large scale procurements prior to and since the ADDSS procurement. Our expectation was that it would be able to cope with the volume of activity.

We have been working very closely with IBM Emptoris to prevent a reoccurrence of these issues, resulting in:

  • Planned upgrade of available hardware that will provide an increase in system capacity as well as increased flexibility to balance the load of usage across multiple servers.  This new environment is scheduled to be delivered in mid December.
  • The development of a recovery process that will help to reduce capacity constraints in times of high volume usage.

In the meantime, IBM Emptoris is allocating additional resource to support GPS procurements where required. We (GPS) have also introduced a number of measures to smooth peak requirements and optimise access to the eSourcing system. These include the introduction of mandatory testing of procurement events prior to publication of the OJEU to optimise system configuration and the structure of the procurement event and also the inclusion of projected capacity requirements in the procurement planning pipeline.

The new eSourcing solution is part GPS’s eEnablement strategy which aims to transform the sourcing, procurement and management of centralised deals and suppliers, ensuring that all arrangements are easily accessible to customers. All elements in the strategy have been delivered, including the new procurement portal, the Government eMarketplace and the spend analysis tool which is providing a detailed view of procurement spend across Government and is available to customers across the public sector”.

Thanks to GPS for that – I do have a lot of sympathy for them here. I’m not sure I would have foreseen the problem if I’d been in their shoes, and it was worth trying the manual work-around even if it wasn’t ultimately successful.  Anyway, in part 2 tomorrow we’ll look at any general lessons we can derive from this incident.

Related Articles:

  • No related articles found

Comments

  • eSourcingSensei:

    Hi Peter

    If it involves eSourcing you can almost here my fingers tapping on the laptop keys LOL :-)

    GPS are really trying to push their “known” boundries by further investing and developing their eSourcing use and the tools they are utilising.

    But clearly there are still gaps in their usage knowledge demonstarted by this comment:
    “These include the introduction of mandatory testing of procurement events prior to publication of the OJEU to optimise system configuration and the structure of the procurement event and also the inclusion of projected capacity requirements in the procurement planning pipeline”

    I have for some years trained out the use and advised on many many events in the best ways to use and exploit these tools for a businesses advantage.

    One of the things I have always stressed as basic best practice, particularly where you are running a “Unique” event – that is a non basic event using a template framework that would have been previously used and/or tested – is the running of a test event that is an exact copy of the live event utilising dummy suppliers to test.

    This allows a “real” view of how an event looks and by acting as a number of test suppliers a buyer/team can quickly see if the responses requested make sense on supplier side and that when responses and uploads are made they “work” – once you have done that for say 6 – 10 test suppliers you can look at the data uploaded and the responses made and extrapolate that out to estimate the potential responses from the number of invited suppliers to the “live” event.

    Quickly identified is the amount of capacity needed, and you can talk to the tool supplier about making more server capacity available, unless it is behind the GPS firewall then that is an internal call by GPS IT team. You must also consider the number of events you are running across the business and allow for their response and requirements as well when considering total server space/capacity required.

    You are right Peter in that with cloud technology this should never be an issue but an organisation needs to buy that space in order to use it – it is potentially unlimited but is not free.

    This would also hugely depend (as I sort of mentioned above) if the eSourcing tool is behind the GPS firewall meaning that GPS are responsible for the server space and backroom support and they would be responsible for increasing server space and not the supplier.

    So I think the introduction of compulsory testing prior to event launch is great but really and truthfully should have been inbeded a few years ago when GPS commenced using eSourcing tool sets

    Sensei

Join the Conversation

* Required fields  [email address will not be published]

*