Choosing help
When to bring in a consultant, and when not to
Timing beats talent. The same engagement that reshapes a project in week two is expensive wallpaper by week forty. These are the moments, across every discipline I work in, when outside help actually pays.
The short answer
Bring a consultant in at decision points, not delivery panics: before a rebuild locks in its assumptions, when a metric has stalled and the team's explanations have stopped changing, ahead of a compliance or procurement deadline, or when everyone is too close to the product to see it.
Don't bring one in to add authority to a decision you've already made, or to do work your team could do if you cleared their calendar. A good consultant will spot both on the first call and should say so.
The signals worth acting on
Across research, service design, accessibility, CRO and strategy, the profitable moments look remarkably similar.
Before a rebuild or replatform
The cheapest week to change direction is the one before anyone writes code. Research, an IA review or a UX audit at this point costs a fraction of what it costs to discover the same problems in production, where the answer becomes a change request programme.
A metric has stalled
Conversion flat across releases, sign-ups plateaued, drop-off stuck at the same step. When the team's explanations start repeating, you don't need more opinions. You need a diagnosis: behaviour analytics, a conversion review, or research with the actual users.
A compliance deadline is now real
Accessibility regulation, a public-sector audit, a client contract with WCAG requirements. Planned accessibility work is design; last-minute accessibility work is triage. The earlier the review, the more of the fix lands in normal delivery rather than emergency rework.
Before funding, procurement or a big pitch
Evidence beats conviction in front of investors and procurement panels. Validation research, a product direction review or hands-on tender support strengthens the case while it can still change, not after the document is submitted.
The team is too close to it
Internal deadlock, decisions defaulting to the loudest voice, a design everyone has stopped seeing clearly. An outside view with no stake in who was right is sometimes the entire value of the engagement, and it's a legitimate reason to buy one.
Launch happened and the line went flat
The product shipped, the numbers didn't move, and nobody quite knows why. Post-launch is a research moment: watching real users meet the real product answers in a week what internal debate won't settle in a quarter.
Why earlier is cheaper
Every discipline has its version of the same curve: the cost of changing a decision rises steeply with everything built on top of it. A finding that redirects a discovery is a conversation; the same finding after launch is a backlog. That's why diagnostic engagements are priced in the low thousands while the mistakes they catch are priced in rebuilds.
Late is still better than never, but be honest about what late help can do. Once a platform is chosen or a service pattern is live, a consultant can steer the next hundred decisions, not unwind the last hundred. The scope of what's fixable narrows; the value of an outside view doesn't.
When not to hire anyone
If the constraint is hands rather than judgement, meaning you know what to build and just need it built, you need capacity: contractors, an agency, or your own team with a cleared calendar. Paying consultant rates for production work is the wrong tool, and I'll say so on the call.
And if nobody has the authority or budget to act on what the work might find, wait. A diagnosis that can't lead to treatment is the one genuinely wasteful way to buy consultancy. Better to spend the money once someone can say yes to the findings.
Match the signal to the discipline
A stalled metric points at conversion or research; a deadline points at accessibility; a rebuild points at IA, UX or strategy. Whichever signal brought you here, there's a scoped engagement for it.
Caught it at the right moment?
Every engagement is scoped and priced to fit a decision point: reviews and audits before you build, research when the numbers stall, roadmaps before you commit. If you're not sure which fits, one conversation settles it.
Still deciding?
Frequently asked questions
Is it too late once the rebuild has already started?
No, but the shape of the help changes. Mid-build, the useful work is steering what's still undecided: reviewing designs before they're implemented, testing the riskiest journeys with users, and stopping new problems being poured on top of old ones. What a consultant can't honestly do mid-build is unwind the platform and architecture choices already made, so expect prioritised pragmatism rather than a clean-sheet vision.
How do I know the problem is big enough to justify outside help?
Price the decision, not the engagement. If the question at stake, whether that's which direction to rebuild, why customers abandon, or whether the service meets accessibility requirements, carries a five- or six-figure cost of being wrong, a diagnostic engagement priced in the low thousands is cheap insurance. If nothing much rides on the answer, it isn't a consultancy problem yet, and you shouldn't pay consultancy prices for it.
Can a consultant help before we've secured the budget?
Often that's exactly the moment. Validation research, an opportunity assessment or a direction review turns 'we believe' into 'we observed', which is what funding committees and procurement panels actually respond to. I also support pitches and tenders directly, by the day. A small evidence-gathering engagement frequently pays for itself by making the bigger budget case land.
LET'S WORK TOGETHER
Ready to improve your product?
UX, research and product leadership for teams tackling complex digital services.
Previous feedback
I had a fantastic experience working with Andy. One of his most impressive achievements during our time at NHS HEE was masterminding a deeply complex information architecture for a new platform that brought together a large number of legacy websites.
Will Parkhouse
Senior Content Designer