What the award records reveal about Denmark’s public procurement market

The award records contain 5,008 separate supplier records across 7,085 procurements. Of these, 3,350 appear in only one procurement. The other 1,658 appear in more than one and account for 74% of all occasions where a supplier is named on an awarded contract package.

Why it matters: suppliers and business networks can use the wider group to find less familiar competitors or potential partners. Contracting authorities can use it to broaden early market research beyond the names they already know. One recorded appearance does not mean that a company is small, new or unsuccessful elsewhere.

A procurement can be divided into many contract packages, often called lots. Counting packages helps explain what was bought, but it can make one large framework agreement look like many separate market events. For questions about market presence, the number of distinct procurements is usually the better starting point.

SoftwareONE Netherlands is named on 118 awarded packages in one cloud procurement. COWI is named on 142 awarded packages spread across 128 procurements. Their award totals look similar, but their history in the market is very different.

The Open Clouds for Research and Education procurement (OCRE) is an unusually large case: one procurement with 326 awarded packages and recorded supplier addresses in 19 countries. Because it creates 306 of the 862 awarded IT-service packages, it has a large effect on any chart that counts packages.

At procurement level the picture is much steadier. With OCRE included, 92 of 514 IT-service procurements (17.9%) name at least one supplier address outside Denmark. Without it, the result is 91 of 513 (17.7%). The apparent swing is therefore mainly about how one procurement is structured, not a sudden change in the supplier market.

The sector differences are still meaningful. A supplier address outside Denmark appears in 33.8% of medical-goods procurements, 21.6% in finance and insurance, 17.9% in IT services, 3.7% in construction and 2.4% in environmental services. These figures describe where the named suppliers are registered; they do not measure the Danish share of contract value.

One procurement may report a separate bid count for each contract package. The archive therefore contains 20,008 package-level results, rather than 20,008 separate procurements. Of the 15,777 results with a usable count, 3,558 (22.6%) report one bid. Another 4,231 results (21.1% of all results) have no usable count.

This makes the single-bid filter useful for finding procurements that deserve a closer look. It is not enough on its own to rank a buyer, compare sectors or conclude that competition was weak. Procedure type, the number of packages and the original notice all matter.

A reported zero also needs context. In this report's original frozen projection we checked 111 result lots with both a selected winner and zero bids; 95 used a negotiated procedure without a prior public call. Zero is not proof of a deserted competition or of no supplier interest. The source audit dated 30 September 2026 additionally found 13 lots with conflicting repeated tender counts: refreshed dashboards treat these as unknown, not last-value wins. Their new approach, reason and no-response filters distinguish direct procurement, unsuccessful competitions and failed sibling lots. The figures above remain the original dated report, not the revised classification totals.

Dated classification supplement · 30 September 2026: the stricter, source-grounded rules identify 102 wholly deserted procedures, 198 with some no-response lots and terminal outcomes, and eight with some no-response lots but incomplete or review-required outcomes. The earlier 315-procedure screening count did not mean 315 wholly deserted tenders. Partial award means only that some lots did not receive an award; it does not describe delivery.

The ten largest notice values contribute 27.7% of the dashboard’s reported DKK total. This concentration means that one unusual entry can materially change the overall picture. It also makes a source check of the largest values essential before the total is used for market sizing.

The history view contains 13,271 procurements. A usable result is available for 9,590 of them. The separate supplier analysis covers 7,085 procurements after additional checks to identify winners. These are two views of the archive with different rules; they should not be read as stages in a funnel.