Information security for software delivery

ISO/IEC 27001
ISO/IEC 27001 certified
Diagnosis
Four causes, in the order they usually apply
Each is measurable from records you already have — the ticket history, the review history and the calendar.
- 01
Decisions queue
Measure the time between a question being asked and answered. Where the median is over a day, the team is idle or building on assumptions, and both show up as slow delivery that looks like a capability problem.
- 02
Practice was never written down
Definition of done, review expectations and release process live in the heads of your existing team. A new team cannot infer them, and the resulting friction is read as carelessness on both sides.
- 03
Requirements arrive as conversations
A ticket that says "make the report better" produces rework, and the rework is billed. Tickets with acceptance criteria are the cheapest productivity intervention available to a buying organisation.
- 04
The overlap is spent in meetings
Six shared hours filled with status calls leaves no time to unblock anything. The overlap is for decisions and reviews; status belongs in writing, before the day starts.
The fix
What to change in the first two weeks
Name one decision-maker
One person who can settle a question within the overlap window without convening anybody. This single change recovers more than any process. Decision latency measured in hours.
Write the definition of done
Tests, review, documentation, what "done" means before a ticket closes. One page, agreed by both sides, referenced in every review. Arguments about quality become arguments about a document.
Move status to writing
A written end-of-day state from the team and a written response before their morning. The overlap is then free for the conversations that need to be conversations. Overlap hours spent on decisions rather than on updates.
Add acceptance criteria to every ticket
Three lines: what it does, how it is verified, what is out of scope. Rework falls immediately and the effect is visible within a sprint. Less work paid for twice.
When it is the supplier
Sometimes the answer is to change, and it should be a short conversation
Not every failing engagement is fixable from the buying side. Persistent turnover, engineers who cannot explain their own code in a review, and estimates that are consistently a multiple of reality are supplier problems, and no amount of process on your side changes them.
What makes that decision cheap is having the exit terms agreed in advance: notice, knowledge transfer, repository access and a documentation standard. Where they exist, changing supplier is a planned transition. Where they do not, it is a rescue project, and the fear of that is what keeps organisations in engagements they know are not working.
Our role
What a diagnostic engagement does
- 01
Read the record, not the opinions
Ticket history, review history and decision latency measured from the tools, so the diagnosis is not a matter of whose account is more persuasive.
- 02
Fix the buying side first
Decision ownership, written practice and acceptance criteria, because they are free, fast and usually most of the gap.
- 03
Say when it is not that
Where the delivery record shows a supplier problem, we say so plainly and help plan the transition rather than selling a longer engagement.
Whether to continue with an existing supplier is your commercial decision; what we provide is the evidence to make it on something other than frustration.
Common questions
What buyers ask
How quickly should we see a difference?+
Decision latency and acceptance criteria show up within one sprint, because they change what the team can act on immediately. Practice changes — review discipline, test coverage — take a quarter to be visible in defect rates. If nothing has moved after two sprints of the first two changes, the problem is more likely to be capability than coordination.
Should we require daily video calls?+
One short call inside the overlap is useful; several are how the overlap gets consumed. The written end-of-day state does more for delivery than an extra meeting, because it can be read by whoever is available rather than requiring everybody at once.
Can you take over an engagement another supplier started?+
Yes, and the honest first step is a read of the codebase and the delivery record before committing to a timeline. Transitions fail when a new supplier accepts the previous estimate to win the work; the code decides how long it takes, and a fortnight of reading is cheaper than a schedule nobody can meet.
In this cluster
Diagnose it against the delivery record
A diagnosis read from the delivery record rather than from opinions, with the two changes that recover the most named first.
Start the conversation