How to test an adviser’s diagnosis, independence and first assignment before buying a programme.
Read the articleJNAGA insights
What is worth changing?
What should you leave alone?
Articles on the decisions behind digital work: when to hire advice, buy or build software, automate a process, trust a report and revisit an investment after launch. Each takes a position you can test against your own situation.
Explore the articles
Advisory & consulting
A first advisory assignment should end with a decision brief: the issue, evidence, options, trade-offs and owner.
Read the articleCompare the cost of the underlying problem, not the volume or visibility of each technology request.
Read the articleCommission independent diagnosis when you have not yet established what should be built or bought.
Read the articlePause when the case for continuing, the decision owner or the conditions for success have materially changed.
Read the articleAI & decision-making
Define the task, permitted information, error cost and human authority before selecting an AI tool.
Read the articleChoose a task with frequent examples, accessible evidence, a clear reviewer and a safe failure path.
Read the articlePut human approval at consequential actions, and give reviewers enough evidence and authority to exercise it.
Read the articleUse explicit rules for stable, auditable conditions; consider a model only where interpretation adds real value.
Read the articleWorkflows & automation
Automate stable repetition; repair conflicting rules and unresolved exceptions before encoding them.
Read the articleFollow recent cases and the people who handled them; the unofficial route often carries essential decisions.
Read the articleReserve approval for consequential exceptions; let routine cases pass under clear rules and audit.
Read the articleBespoke systems
When a product is enough, when to connect tools and when a custom layer earns its upkeep.
Read the articleReplace a system when its core no longer serves the work; integrate when the real failure is at the boundary.
Read the articleA good brief includes real cases, exceptions, decision rights, information sources and observable success.
Read the articleConfigure common work; customise only where the difference matters enough to maintain through upgrades.
Read the articleThe largest unpriced obligation is often the decision-making and maintenance required after delivery.
Read the articleAgree which system may create, change and correct each fact before building a connection.
Read the articleInformation & decisions
Choose the facts needed for a decision, then make their source, owner and update path dependable.
Read the articleAuthority belongs to an agreed process for creating and correcting a fact, not automatically to the largest database.
Read the articleDisagreement usually begins in definitions, timing or corrections; visual polish will not settle it.
Read the articleDigital experiences
A new page will not fix an enquiry that the organisation cannot recognise, route or answer.
Read the articleOrganise a complex offer around the visitor's decision, not an internal catalogue of capabilities.
Read the articleA long form can filter out serious enquiries when the firm has not earned the right to ask detailed questions.
Read the articlePrioritise the decision, essential information and next action—not merely a layout that fits a narrow screen.
Read the articleGovernance & delivery
Assign ownership of the work, the rules and the exceptions—not just the technical system.
Read the articleCompare the original problem before and after launch, including work displaced into new exceptions.
Read the articleA pilot may prove a feature works under close support; routine use also needs ownership, exceptions and capacity.
Read the articleOngoing stewardship
The spreadsheet or inbox that survives launch may tell you exactly what the new system missed.
Read the articleRevisit the original case, the work people now do and the continuing cost before extending the investment.
Read the articleCommon questions
Where might
we begin?
Do we need a specification before contacting JNAGA?
No. Bring one example of the problem and the decision you need to make. We can establish whether a brief, an independent review or a delivery proposal is the right first step.
Can advice stand apart from implementation?
Yes. You can commission a defined assessment and use its findings with JNAGA or another delivery partner. The role and outputs are agreed at the outset.
When would you advise against automation?
When the rule is disputed, the exception has no owner or a simpler process would remove the work altogether. Automating those conditions makes the problem harder to correct.
How are fees and responsibilities decided?
After we understand the assignment, we agree scope, responsibilities, timing and fees in writing. There is no rate card presented as a substitute for that conversation.