```
Survey: 28% of Enterprise Software Budgets Spent on Upgrades IT Teams Never Wanted
Large enterprises are spending roughly 28% of their enterprise software budgets on upgrades that their own IT leadership believes will make matters worse, according to a survey of 2,018 heads of IT at big firms, reported by Ars Technica on 22 October. The finding is not being framed as a planning mistake or a miscommunication between IT and procurement. Ars Technica's reading of the data is blunter: enterprises keep buying upgrades because they lack the leverage to refuse them.
The distinction matters. Respondents are the people who own the budget line and carry the outcome. When a third of spend is internally regarded as counterproductive, the problem is not that engineers failed to explain their preferences to finance — it is that the vendor, not the buyer, sets the terms on which the software keeps running.
Why the leverage sits with the supplier
The mechanics are familiar to anyone who has managed an enterprise estate. Support windows close on the vendor's schedule, not the customer's. Licences are bundled so that the module you want ships with the release train you did not ask for. Migration paths narrow with each major version, and security patching is increasingly tied to moving to a newer release rather than back-porting to the one in production. In each case, the party that bears the risk of staying put is the customer, while the party that sets the timetable is the supplier.
The second-order cost: trust decay
The direct budget drag is only part of the story. Buyers who have been burned tend to respond defensively: renewals are delayed to the last possible moment, contracts grow over-specified with protective clauses, and teams maintain parallel environments to hedge against a forced migration. Each of those moves is individually rational. At portfolio scale they compound into exactly the kind of waste the survey puts at 28%.
For IT managers, the practical value of the figure is that it is quotable. "A third of our upgrade spend is wasted" is a sentence a CIO can put in front of a CFO to justify version-pinning, contractual exit clauses, modular architecture, and standards-based hedges — advice that has circulated for years without a number attached to it.
A caveat on the number: the survey's sponsoring body and full methodology were not disclosed. Treat 28% as a directional indicator, not an audited figure — and probe any statistic of this kind before restating it in a budget submission.
Open standards and open-source alternatives
One broad observation is worth flagging: a market in which switching costs are the primary source of supplier power is one where open standards and community-governed projects structurally shift bargaining position back toward the buyer. Important as that argument is, it should be stated plainly — the survey itself reports no breakdown by licensing model or by region, so no conclusion about open-source adoption can be drawn from its results.
Nor is there regional data in the study. But the general picture is easy to recognise anywhere multi-national deployments inherit decisions made at headquarters: the shorter the supplier list becomes, the less room there is to negotiate, and the more a forced upgrade propagates down to every satellite office on the same schedule.
Three things to do before renewal season
- Run a shelfware audit. Map every deployed version against its support end date and its actual utilisation. The gap between the two is your negotiating inventory.
- Trade commitment for control. Multi-year discounts can be exchanged for version-pinning rights, extended support on current releases, or modular rather than suite-wide licensing.
- Price the alternative. Whether that means a competing vendor or an open-source replacement, a credible exit is the only leverage that does not depend on goodwill.
We want to hear your war stories
Has your team been pushed onto a release it did not want? Tell us: the product, what the vendor promised, what actually happened, and how much choice you genuinely had. Please send your account to the editorial team — we will publish identifying details only with your permission.
Source: Ars Technica, "Enterprises keep buying software upgrades that IT thinks will make things worse."
```
調查：企業軟件預算 28% 花在 IT 團隊從來不想做的升級
據 Ars Technica 於 10 月 22 日報道，一項訪問了 2,018 名大型企業 IT 主管的調查顯示，大型企業約 28% 的企業軟件預算，用於其 IT 高層自己認為會令情況惡化的升級。這項發現並未被定性為規劃失誤，或是 IT 部門與採購部門之間的溝通不善。Ars Technica 對數據的解讀更為直接：企業不斷購買升級，是因為它們缺乏拒絕升級的籌碼。
這個區別至關重要。受訪者正是負責該預算項目、並為結果承擔責任的人。當三分一支出被內部視為適得其反時，問題並不在於工程師未能向財務部門說明他們的偏好——而在於設定軟件續用條件的一方是供應商，而非買方。
籌碼為何掌握在供應商手中
對於任何管理過企業 IT 環境的人來說，其中的運作方式並不陌生：支援期限按供應商的時間表結束，而非按客戶的時間表；授權被捆綁銷售，令你想要的模組隨著你從未要求的發布周期一同出貨；隨著每一個主要版本推出，遷移路徑愈來愈窄，而安全修補亦日益與升級至較新版本掛鈎，而非回移至現行生產版本。在每一種情況下，維持現狀的風險由客戶承擔，而制定時間表的則是供應商。
二階成本：信任流失
直接的預算壓力只是問題的一部分。曾受創的買家往往會採取防禦性回應：續約拖延至最後一刻才處理、合約加入大量保護條款而變得過度繁瑣，團隊並維護平行環境，以應對可能被迫遷移的風險。這些舉動逐一來看都合乎理性，但在整體組合的規模下，它們累積起來，正好構成調查所指的 28% 浪費。
對 IT 經理而言，這個數字的實用價值在於它可以直接引用。「我們三分之一的升級開支是浪費的」——這句話足以讓 CIO 擺在 CFO 面前，用以支持版本鎖定（version-pinning）、合約退出條款、模組化架構，以及基於開放標準的替代方案。這些建議流傳多年，卻一直欠缺一個數字作支撐。
關於這個數字的一點保留：該調查的贊助機構及完整方法論並未披露。應把 28% 視為方向性指標，而非經審計的數字——在預算提交中引用任何此類統計數據之前，必須先加以查證。
開放標準與開源替代方案
有一點值得指出：在一個轉換成本成為供應商權力主要來源的市場中，開放標準與由社群管理的項目，會從結構上把議價地位重新移向買方。儘管這論點十分重要，仍須清楚說明——該調查本身並未按授權模式或地區提供分項數據，因此不能從其結果得出關於開源軟件採用情況的任何結論。
該研究同樣沒有地區數據。不過，在任何跨國部署需要繼承總部所作決定的地方，整體圖景都不難理解：供應商名單愈短，議價空間愈小，而一次被迫的升級，也愈會按同一時間表傳遞至每一個附屬辦公室。
續約季節來臨前要做三件事
- 進行閒置軟件（shelfware）審計。將每個已部署版本與其支援終止日期及實際使用率逐一對照。兩者之間的差距，就是你的議價籌碼。
- 以承諾換取控制權。多年期折扣可用來換取版本鎖定權利、現行版本的延長支援，或按模組而非整套套件計費的授權。
- 為替代方案定價。無論是競爭對手的產品還是開源替代品，可信的退出選項是唯一不依賴供應商善意的籌碼。
歡迎分享你的實戰經驗
你的團隊是否曾被迫採用一個並不想採用的發布版本？歡迎告訴我們：涉及哪個產品、供應商承諾了什麼、實際發生了什麼，以及你實際上有多少選擇。請將你的經歷電郵至編輯部——除非事先獲得你的同意，我們只會刊載識別性資料。
資料來源：Ars Technica，〈Enterprises keep buying software upgrades that IT thinks will make things worse.〉