The BBC will not be able to deliver value for money if performance and expected benefits of its projects are not defined from the beginning, a report claims.
According to the National Audit Office (NAO), the broadcaster has improved its portfolio of critical projects in response to serious shortcomings identified in the delivery of its Digital Media Initiative (DMI) project.
But the report also found that documentation from the BBC did not make it clear what the overall accountability for the End-to-End Digital - which included a number of sub projects - encompassed.
The report, Management Of The BBC's Critical Projects, looked at eight ventures the broadcaster identified as "most strategically important, complex and high risk".
The corporation forecasts that seven of the eight critical projects will cost a total of £885.1 million, and expects all eight will achieve £1.9 billion in financial benefits.
It rated achievement of planned outcomes as "probable" or "highly likely" for seven projects and "feasible" for one - E20, the rebuilding of the EastEnders set.
The report is released two days days before Culture Secretary John Whittingdale could publish the White Paper on BBC Charter renewal.
The proposed legislation would guarantee the BBC's independence in the management of its own affairs including all matters of output, editorial and creative decisions.
Under the peers' plans the licence fee would rises at least in line with inflation and be for the exclusive use of the BBC, members of a new BBC board would be appointed through an independent committee and the Royal Charter on a statutory footing with votes in both Houses of Parliament to reduce ministerial interference.
The NAO found a delay in five of the eight projects, in comparison with the timetable approved in the business plan. It said this suggested optimism bias at the business plan stage.
Nick Prettejohn, chairman of the BBC Trust's Value for Money Committee, said: "We welcome the National Audit Office report, which recognises that the BBC has strengthened its oversight of critical projects."
He added: "We will continue to press the BBC Executive to make sure these benefits are delivered on time and on budget, as well as monitor the implementation of the NAO's recommendations."
Amyas Morse, of the NAO said that while the broadcaster has learnt from its report of the failure of the DMI, "further concerted action is needed".
He continued: "The BBC needs to do more to manage its critical projects as a coherent portfolio if it is to achieve value for money from its assurance arrangements."
According to the report the BBC's assurance arrangements cannot give confidence that projects will achieve value for money if performance and expected benefits are not defined from the start.
It reveals that three projects (Aurora, E20 and Smart) have each been delayed by 22 months or more.
Only in the case of Smart were delays associated with significant cost increases, from £39 million to £55.7 million, during implementation.
In response to the failure of the DMI project, the BBC was found to have increased the frequency of reporting on project performance to the Executive Board from quarterly to monthly.
It also reduced the time taken to get performance information to the board, from 66 days in 2013 to 27 days in 2015.
A report by the National Audit Office (NAO), published in January 2014, was critical of the DMI scheme and said the BBC Executive "did not have sufficient grip" on the IT project and did not properly assess the system to see whether it was "technically sound".
DMI was meant to allow staff to handle all aspects of video and audio content from their desks, but after years of difficulties - during which £125.9 million was ploughed into it - the plug was pulled last year, leaving a net cost of £98.4 million.
The NAO considers that there is scope for all members of the Executive Board, including non-executives, to take a more active role in deciding which projects are included in the portfolio.
It recommends that the BBC should make it clear who is accountable overall for projects, ensure that reporting on project performance is always clear and comprehensive and to set out at an early stage the expected benefits and outcomes.