Professional practice and the HCI profession

Doing good work and getting it adopted are different problems, and the second one is where capable graduates most often lose their footing. A design can be correct in the abstract and still fail. Nobody agreed what question it answered, or the person who would fund it never heard a reason to, or the researcher lost access to participants three days before the deadline. Professional practice is the set of skills that stop good work being wasted.

The contested question in this area is whether the value of design can be argued on the terms a business already trusts. One position is that designers must speak in money, and it points to commercial studies that link strong design to financial performance. Another is that those studies show association rather than cause, and that treating a correlation as proof corrupts the argument it was meant to win. The evidence supports a middle position, and holding it well is harder than choosing a side.

The industry is not one industry, and each part rewards something different

An in-house product team rewards a designer who can hold a long relationship with one product and the people who use it. A consultancy rewards someone who can enter an unfamiliar domain quickly and leave a written asset behind. An agency rewards pace and craft inside another organisation's brand. A research organisation rewards a method that will survive scrutiny years later. The public sector rewards work that meets a legal standard and a public duty, not only a commercial target. The same person can move between these settings, but the definition of a good week changes with each move.

Cennydd Bowles and James Box wrote for the common situation in which the organisation has not bought into user experience at all. Their book is frank about fitting research, prototyping and testing into the time, budget and culture a team actually has (Bowles and Box, 2011). Russ Unger and Carolyn Chandler describe the same reality as a project ecosystem, in which the designer is one role among several and each role carries its own measure of success (Unger and Chandler, 2024). Reading the setting correctly, before proposing anything, is the first professional skill.

Scoping decides whether the work is possible at all

Feasibility rests on a short list of dull questions. Who will use this, and how will we reach them? What decision will the work change? What has already been written that we could read instead of asking? What date is the answer needed by, and which parts are fixed and which are negotiable? A brief that asks for a usability test often conceals a scoping problem, because there may be nothing built yet to test.

The client's real question is often which of two directions to take, and no usability test settles that. Unger and Chandler put the project's objective and approach ahead of the method for exactly this reason (Unger and Chandler, 2024). Naming the decision the work will inform is the most useful sentence in a proposal. It tells the client what they will do differently once the work is finished. A proposal that cannot name that decision is a description of activity, not a plan.

Honest estimates come from the record, not from the plan

Estimates go wrong in a predictable direction. Roger Buehler, Dale Griffin and Michael Ross found that people underestimate their own completion times but not other people's (Buehler, Griffin and Ross, 1994). They also found that people build an estimate from an imagined plan rather than from how long comparable work took before. Across five studies with 465 undergraduates, predictions were too optimistic for academic and everyday tasks alike. The authors report that the bias shrank when participants were told to connect relevant past experience with the prediction. That instruction is the practical lesson: estimate from the record of finished work, not from the story of how this job might go.

The evidence has limits. The studies used students on short tasks, so the size of the effect transfers to a design project only by analogy. The mechanism, an inside view that ignores how similar work has gone before, is well supported. A team that logs estimated against actual effort, and reads that log before quoting the next job, is applying the finding the study actually supports.

A proposal holds assumptions, and a contract makes them binding

A proposal is a set of assumptions written down before the work begins. It fixes the deliverables, the number of research sessions, the formats, the dependencies the client must supply, and the points at which either side may stop. The contract turns some of those assumptions into obligations. The style of the writing matters far less than whether the assumptions are visible, because a hidden assumption is where the argument starts later.

Intellectual property is the term designers most often skim. The Intellectual Property Office states the default plainly: for designs, a commissioned design belongs to the designer and not to the commissioner, unless a contract says otherwise (Intellectual Property Office, 2014). A designer who signs a client's template without reading the ownership clause may give away rights that were never discussed. The clause is not hostile. It is where ownership and rights of use are settled, and settling it before the work is far cheaper than settling it afterwards.

The arguments that recur, and what settles them

Disputes between a designer, a product manager and an engineer recur because each role is optimising something different. The product manager is accountable for shipping something that moves a number. The engineer is accountable for a system that stays correct and maintainable. The designer is accountable for the person using it. When a designer asks for a change late in a build, the engineer hears rework and the product manager hears delay, and neither is being unreasonable.

What settles these disputes is evidence that both sides accepted in advance as decisive. A recorded session in which three of five participants fail the same step persuades where an opinion does not. Agreeing before the work what result would change the decision turns an argument about authority into a test. Bowles and Box approach the same problem from a weaker position, where the researcher has no mandate and must fit the work to the constraints the organisation will allow (Bowles and Box, 2011).

Making the financial case without overclaiming

The strongest commercial argument for design carries a weak spot, and being honest about it is itself a professional skill. McKinsey tracked the design practices of 300 publicly listed companies over five years (McKinsey and Company, 2018). It found that companies in the top quartile of its design index grew revenue and shareholder returns substantially faster than their industry peers. The report states plainly that this is a correlation. Its method, an index built from survey, interview and financial data, cannot show that good design caused the growth. Well-run companies may simply be good at both.

A designer who presents the gap as proof invites one informed question and loses the room. A more defensible move is to price the problem being solved inside the client's own numbers: the support calls, the abandoned applications, the rework. That estimate is local and uncertain, but it belongs to the organisation and the client can check it. The same report notes that over 40 per cent of the companies it surveyed were not talking to their end users during development at all (McKinsey and Company, 2018). That is a fact about practice rather than a claim about cause, and it is safer ground.

Research operations is the infrastructure that makes research repeatable

Research becomes fragile when it depends on one person's goodwill and a spreadsheet. Research operations is the layer that makes it survive staff changes. It includes a participant panel that can be reached and re-contacted lawfully, a repository where past findings can still be found, a recorded consent process, and ethics arrangements that do not have to be rebuilt for every study. Kate Towsey describes eight elements of research operations and argues that ethics and privacy have to be more than a list of prohibitions (Towsey, 2024).

The professional standards are specific. The Market Research Society's Code of Conduct requires informed consent before data collection and sets out what participants must be told (Market Research Society, 2023). It requires that a participant's right to withdraw is respected, and it forbids passing participant details to a third party without consent. A designer who cannot say where last year's interview recordings are stored, or who consented to what, is one subject access request away from a serious problem.

Leading a small team means reviewing the work, not the person

Leading a small team is mostly the skill of critiquing work without damaging the person who made it. The failure modes are symmetrical. A lead who softens every comment leaves a colleague unprepared for a client's first hard question. A lead who delivers blunt judgement without separating work from person produces defensiveness and silence. The BCS Code of Conduct asks members to respect and value alternative viewpoints, and to seek, accept and offer honest criticism of work (BCS, 2022). Honest and respectful at once is the standard.

A workable habit is to review the artefact against the goal it was made for rather than against personal taste. Ask the author what they were trying to achieve before saying whether it worked. That is the distinction the code draws between criticising the work and injuring the colleague. In a team of four or five, the habit is the difference between a shared standard and a set of private preferences.

Professional duties, chartership, and what comes next

Professional standing is not a certificate earned once and kept. The BCS Code of Conduct obliges members to work within their competence and to keep their knowledge current (BCS, 2022). It also obliges them to promote equal access to the benefits of information technology. Chartered IT Professional status asks for evidence of technical depth and breadth against a skills framework (BCS, n.d.). It also asks for demonstrated leadership, decision making and problem solving, and for recorded continuing professional development. Equality, inclusion and accessibility sit inside those duties rather than beside them.

The law gives accessibility a hard edge. The Public Sector Bodies (Websites and Mobile Applications) Accessibility Regulations 2018 require public sector bodies to make their websites and apps perceivable, operable, understandable and robust, and to publish an accessibility statement (Public Sector Bodies, 2018). A claimed disproportionate burden must itself be assessed and justified. A professional who treats accessibility as a late checklist is exposed both ethically and legally.

The next stage of the programme turns these skills into practice, in a sustained team engagement with an external client where the proposal, the estimate and the handover all have to hold.