Image credit: US Navy Archives Personnel, Bess Glenn in foreground, August 14, 1942. (catalog.archives.gov/id/3493249)
Why is it so hard to find the right public sector tender opportunity for your business? We’ve been looking at some of the reasons.
We estimate that there are around 2,500 tenders published a week by UK public sector; working on a rolling eight week programme, means that there will be 20,000 live tenders at any given moment. That is a non-trivial amount of data to process without specific, organised search tools.
Processing 20,000 records is hard enough on its own, but it is virtually impossible when the data is spread over a multitude of different procurement portals. Despite the efforts of the Contracts Finder team, over 450 unique tender portals have been used by the public sector since March 2014. Contracts Finder is gathering a lot of this data, but buyers in the wider public sector are still not using Contracts Finder, despite a change in the law requiring them to use it.
It would be much easier if suppliers knew when a buyer’s contracts were due to end, so that they could prepare in advance of a tender being published. Public bodies are supposed to publish details of all of their contracts, but this data is very poor, most bodies don’t publish all of their data and what is published isn’t very good.
Tenders tend to use the Common Procurement Vocabulary (CPV) classification system, a list of just under 10,000 descriptions for goods or services that can be bought via a tender. Choosing the correct codes can be a chore and buyers frequently publish incorrect codes or codes that are too broad to be useful. Today it is possible to auto-classify tenders based on their titles and descriptions, where there is sufficient data to auto-classify a tender (see below) perhaps we could help buyers by suggesting the codes that best fit the data they’ve entered?
When it comes to tender adverts, there’s a mismatch between the priorities of publishers and consumers. There’s a legal requirement for public bodies to publish tender notices, which often means that the information is published only to meet this need. Conversely, suppliers need to make a decision about whether to bid or not, so their needs are for nuanced, contextualised descriptions that precisely guide them to the opportunities they want. Despite this, more than 3% tenders published in 2015 contain descriptions with less than 20 characters. A consulting tender with the description "Consulting contract" and no more is unlikely to provide suppliers with the information they need, amongst a sea of other tenders, prospective bidders need to secure the additional documents before they can bid. A crucial piece of information is the expected price, but over the last seven years, 94% of tenders give no indication of the expected price range or budget for a tender.
So you can’t search by a particular sector of Government (e.g. Justice), nor can you restrict searches to particular types of organisation (e.g. Mental Health Trusts). Data linking is more than just knowing more about the buyer, linked data is also about the whole life of a contract, starting with linking contract awards to tenders and then linking spending to the contract. This is what linked data should look like in procurement:
The Spend Network teapot: Linking procurement data
Over the coming months, the team at Spend Network will be working to solve some of these problems for our customers, we’ll be bringing you lots of new contract and tender data, as well as new tools to help you find the best tenders for your needs.
In the meanwhile, we’d love to hear about what you’d like to see that could help you when you’re looking for tenders. Either comment below or get in touch directly.