How and why consulting goes wrong

Barriers to great advice

Consultants are experts in consulting

The core competence of a consulting firm is selling consulting engagements. Solving the client's problem is a priority, but it's often priority number two. It's a matter of incentives. Consulting firms make money when they sell engagements to clients — and whether or not their advice actually helps the client is almost irrelevant to their internal measures of success.

The conclusion arrives with the proposal

To win the work, consulting firms usually have to submit a proposal. The only way for this proposal to be engaging or persuasive is for it to speak to the client's problems and how they can be resolved. So the firm sketches the broad brushstrokes of its conclusion in the pitch — before the data, the analysis, or the synthesis.

Advisory work is open, abstract, and hard

Advisory engagements are typically open and often abstract. They involve addressing very broad problems, with a near-infinite set of potential moves, in a shifting environment, with multiple stakeholders. They're hard problems. It isn't surprising that consultants often fail to maintain their intellectual discipline, fall victim to lazy thinking and cognitive biases, or lose perspective.

Reporting on the work isn't doing the work

Many engagements have heavy process layers. The consultant team has to manage a charter, terms of reference, the steering committee, project control board, approval gates, weekly progress meetings, steerco packs, status decks, weekly reports, sync sessions, and more. Participating in the meetings and producing the artefacts about the work can be mistaken for actually doing the work. The real thinking is often squeezed into whatever time is left over.

Consultants have never built anything

Few consultants have ever built a business, managed a large organisation, or shipped a product. They spend their careers making boardroom presentations and moving on to the next client long before anybody discovers whether their recommendations worked. They're insulated from reality and lack a feedback loop to teach them what actually works in the real world.

The partner who sells the engagement is barely involved

A partner wins the work on the strength of their experience and judgement, then hands it off to a team of analysts a decade or two younger. The juniors are bright and they learn fast, but they aren't seasoned operators. They're also learning on the client's problem at the client's expense.

Consultants want to please the sponsor

The client sponsor commissioned the work, set the context, framed the problem, provided the data, and controlled access to the key people. It's not surprising that consultants often arrive at a conclusion similar to the sponsor's. They're also usually close to the sponsor, feel allied with or beholden to them, and don't want to disappoint them. Advice bends towards the sponsor.

Sponsors can't afford a failure

The bias runs in the other direction too. Having proposed the engagement, justified the expenditure, and fronted the whole thing internally, the sponsor now needs it to succeed. A failure would reflect poorly on them. Even mildly disappointing engagements are declared successes and unmitigated failures are reframed as valuable lessons.

Few can tell good advice from bad

Even a client determined to assess the work objectively would struggle. The counterfactual isn't available, the results arrive years later, and by then execution, market conditions, and a hundred other decisions have muddied the outcome beyond attribution. So quality gets judged on proxies: the polish of the deck, the confidence of the delivery, the seniority in the room, and the name on the cover. An industry that can't measure success will optimise the proxies instead.

Advice anti-patterns

There are recurrent pathologies and dysfunctions in the bad advice that consultants provide. I group these anti-patterns into three buckets — non-advice, weak advice, and biased advice. Non-advice contains no real recommendation at all. Weak advice contains recommendations, but they're really not worth paying for. Finally, biased advice contains real recommendations, but they've been skewed by something other than the client's problem.

Non-advice

Problem as solution

The problem is flipped and reworded as a solution. A client describes their business as disjointed, so the recommendation is an integrated operating model. A client is concerned by rising costs, so the recommendation is to reduce them. Restating the problem in the imperative doesn't convert it into an answer — even when it's dressed up in sophisticated language on polished slides.

Decomposition as solution

The problem is broken into components, the components into sub-components, and the resulting structure is presented as the output. Decomposition is a genuinely useful step; it just isn't a destination. A tidy tree demonstrates that the problem has been organised, not solved. It just takes the client from one unsolved problem to a dozen unsolved sub-problems.

Vocab as insight

Familiar problems and obvious approaches are relabelled with catchy buzzwords. Translating things into business-speak is mistaken for actually resolving them. Polished jargon can obscure the fact that the consultant's advice is just empty words.

Discussion not recommendations

The consultants address the question and its context thoroughly. Options are laid out, considerations are weighed, trade-offs are set side by side — and then it stops. Assembling the options is the easy half. The value was always in saying which one the client should choose.

Unactionable recommendations

Deepen the customer relationship, reduce costs, improve collaboration between departments — these are all outcomes, not actions. Good advice tells the client what they should do. Launch a loyalty program, negotiate bulk discounts with key suppliers, standardise all departments on the same core software — these are actions that the client can take on Monday morning.

Deferral to the next phase

The engagement concludes by recommending another engagement — a pilot, a deep dive, a detailed design. It's most galling when the proposed new work bears a striking resemblance to exactly what this engagement was supposed to achieve.

Weak advice

Reflection not insight

Dozens of interviews are conducted, the themes are collated, the analysis is performed, and the final recommendations are just the interviewees' own views played back to them. It tells the client nothing new. A synthesis of what the organisation already believes is a reasonable place to start the analysis, but it's routinely presented as the result of it.

Overly conservative advice

Advice that cannot fail conspicuously cannot achieve much either. It also tends to be recycled — the same tightening of governance, the same clarification of accountabilities, the same phased roadmap that would suit almost any client in almost any industry. Nothing in it is wrong. It just tends to be low-value, generic, one-size-fits-all advice.

Compromise

Recommendations are negotiated before they're delivered. Each stakeholder's position is accommodated, each objection is softened, and what emerges is acceptable to everyone and useful to no one. Worse is the strategic split: two coherent options averaged into a single incoherent one that inherits the costs of both directions and the benefits of neither.

Best practice is regression to the mean

Best practice is, by definition, what other companies are already doing. Adopting it moves a client towards the industry average — upwards if they were behind it, downwards if they were ahead. It's defensible, it's occasionally right, and it's sometimes exactly what a laggard needs. But it's worth being clear about what's on offer: an average, not an advantage.

Biased advice

Solution looking for a problem

Every firm has capability it wants to deploy — a new service offering, a practice with people on the bench, or a vendor partnership. When a firm has an excellent cloud migration practice, it's remarkable how many of its clients turn out to urgently need a cloud migration. The diagnosis conveniently matches the cures the consultants happen to have in stock.

Cost bias and tech bias

Recommendations are skewed towards cost reductions and technology implementations. They can be justified, implemented, measured, and pointed at afterwards. Less tangible priorities such as capability, culture, quality, and customer trust are slow to change, diffuse, hard to quantify, and awkward to claim credit for. Advice favours the deliverable, measurable, and attributable.

Hypothesis confirmation

The hypothesis-led method is sound in principle: form a view early, then attack it. In practice the attack often softens. Evidence that fits is pursued, evidence that doesn't is noted and set aside, and the hypothesis survives on the strength of never having been seriously tested.

The way forward

None of this dysfunction requires a villain. Every barrier described is a natural response to the incentives consultants and clients face, and the advice anti-patterns are their natural results. But even when these patterns are well understood, well-intentioned people find them difficult to avoid.

Not all consulting firms or clients want to avoid these patterns. Many consulting firms are perfectly comfortable providing uninspired, overly conservative, one-size-fits-all advice. It's a profitable business model. And it's what many clients have come to expect. Clients may even prefer advice that gently leads them towards perfectly average “best practice” over genuinely insightful recommendations that would be far more impactful for the organisation. After all, more impact usually involves more change and may be harder to justify to a board, require the acknowledgement of earlier missteps, or bruise the egos of key stakeholders.

I recommend clients and consultants do two things:

Match the consultants to the engagement. If the client truly wants transformational change, is open to new ideas, isn't wedded to an assumed solution, and isn't overly concerned with compromise or ego preservation, then not every firm will suit. If all the consultants can talk about is supposed “best practice” or the management fad of the day, they're probably not the right fit. A client who wants independent thought and genuine insight has to select for it.

Acknowledge how consulting can go wrong and set expectations up front. Both sides should have an open and honest conversation before the engagement is booked. It's in the interest of both parties to acknowledge the ways engagements can drift off track and the substandard outcomes that drift leads to, and to agree the level of impact and change the client is prepared for. Ideally these expectations would be written into the brief and revisited throughout the engagement.