Hire the Spark, Teach the Skill: The Selection Problem Hiding in the Labour Problem

The job ad was for a guest services lead at a property running a four-star rate card and a five-star ambition. It listed eight technical requirements (PMS proficiency, two specified systems named on the page), three language fluencies, a minimum tenure in branded hotels, a uniform standard that ran to a paragraph, and a list of behavioural expectations that came down to “professional, presentable, punctual, team-oriented.” Nine hundred and forty words. The word “warmth” did not appear. The word “presence” did not appear. The word “guest” appeared eleven times, never once as the subject of a sentence describing what the role’s instinct was meant to feel like. The role was described as a sequence of tasks. The person was described as a list of competencies.

The interview structure that funnel fed into ran for sixty minutes. Roughly thirty-five of those minutes were spent on technical compatibility, system experience, language testing, and scenario questions that read as if they had been copied out of an MBA textbook published in 1984. The remaining twenty-five minutes were the GM’s “feel” pass: did the candidate seem reasonable, present well, fit the team. That’s the structure the property believed was selecting the people it described, on the wall of its own lobby, as “the heart of our hospitality.”

Six months earlier this property had run its third “service culture refresh” of the year. The agency had delivered a 47-slide deck. There were posters in the staff areas. The breakfast service was efficient and unhappy in equal measure. The room felt processed.

The walk-through told me the same thing the job ad had already told me. The diagnostic was complete before I stepped on the property.

The labour problem that isn’t one

A few weeks ago I argued that the bifurcation between hotels and accommodation businesses is decided at the human service layer, and that most operators running hybrid properties can’t afford the labour layer required to convert from one to the other. The follow-on question I’ve heard most often since is some version of: how do we afford it.

The honest answer is that most operators can’t afford the labour they’re currently hiring. Not because the wages are too high. Because the selection is wrong. They’re paying market rates for a variable they aren’t actually buying.

The industry conversation about hospitality labour is locked at the wrong level. Trade press talks about wage curves, generational expectations, post-pandemic scarcity, the share of the workforce no longer willing to work split shifts. All of that is real. None of it is the problem. The problem upstream of every wage conversation is the selection mechanism in the recruitment funnel itself.

A property advertises for a guest services lead and writes a job spec that lists languages, systems, and uniform standards. Sixty applications come in. The screen filters for years in branded hotels and PMS familiarity. Twelve get to an interview. The interview tests technical fit and “professionalism.” Two get offers. They start on Monday. Three months later the GM is in a service culture meeting wondering why the floor has technical compliance and no warmth.

The floor is delivering exactly what the funnel selected for. The funnel did its job. The job description was the problem.

The inversion

Hospitality is the inverse of every other industry. You hire for the inherent care and teach the technical skills. Not the other way around.

This is not a wellness slogan. It’s the operational truth that almost every hospitality business gets backwards in its hiring funnel.

The technical skills are teachable. PMS proficiency, language polish, plate presentation, recovery scripts, the property’s particular cadence at check-in, the wine list, the room service map, the booking platform quirks. All teachable. Some of them in days. Most of them in weeks. The good ones in months. None of them in a timeframe that should determine a hiring decision.

The thing that can’t be taught in any timeframe a hiring decision can wait for is the inherent orientation to be of service. The spark. The natural inclination to read a room before being asked, to notice the person before noticing the task, to register the small difference between a guest who wants to chat and a guest who wants to be left alone. The instinct to step into a problem rather than escalate it. The reflex to remember the name that wasn’t on the briefing sheet. That orientation is either present in the candidate or it’s not. The interview is a discovery exercise, not a development exercise.

You can teach a person with the spark to run a PMS. You cannot teach a PMS operator to develop the spark.

That is the inversion. It is the entire argument. Most operators have built funnels that select hard for the second category, declare a labour shortage, and complain that the floor has no warmth.

What this looks like at the recruitment touchpoint

The Total QX™ framework treats the staff experience as an equal participant in the experience design. This is not a courtesy. It is structural. The recruitment funnel is itself a touchpoint, and it’s delivering an experience to every candidate who passes through it. The candidate is reading the property’s standards before they’ve ever spoken to a manager. The way the property hires demonstrates the way the property operates. There is no separation between the two.

The first audit move, then, is to read your own job spec as if you were the person it’s meant to attract. Not “is this technically accurate.” Does this document describe the experience this person is meant to deliver, or does it describe a list of technical tools they’re meant to operate. If the words “warmth,” “presence,” “the instinct to read a room,” “the orientation to be of service” don’t appear anywhere in the document, the document has already selected against the person you say you want.

A second audit move is to count the verbs. A job spec that runs heavily on “operates,” “processes,” “manages,” “executes” is recruiting a technician. A job spec that runs on “notices,” “anticipates,” “reads,” “looks after” is recruiting a host. The verbs are the standards declaration, whether the writer of the document meant them that way or not.

A third move is to read what’s implicit. Many specs list “professional, presentable, punctual” as if those three words capture the human variable being hired for. They do not. They describe the floor of behaviour an industry should be able to assume. They are not differentiators. They are table stakes. A job spec that treats them as headline competencies has nothing left to say about what the role actually delivers.

The five-minute or fifty-five-minute interview

The interviewer who knows what to look for can read fit-or-no-fit in five minutes. Either the spark is in the room or it’s not. The conversation registers. The energy reads. The candidate either has the orientation or they don’t.

If the spark isn’t present, end the conversation. Five minutes. Politely, generously, with thanks. Don’t spend the next fifty-five minutes pretending to evaluate technical fit you’ve already decided you won’t use. You’re wasting the candidate’s day and your own. You’re also signalling, every time you do this, that the property is in the business of going through motions. That signal travels. Candidates talk.

If the spark IS in the room, invest the remaining fifty-five. Go deep. Three things to test.

The energy match. A calm property doesn’t want an electric host, however brilliant. An electric property doesn’t want a calm one. The candidate’s natural register either fits the room or fights it. The interview is the only place this surfaces honestly.

The noticing instinct. Ask for a story about a moment they paid attention to something nobody had asked them to pay attention to. The story either exists in their experience or it doesn’t. The candidate who can’t find one in two minutes doesn’t have the instinct. The candidate who has three to choose from has it everywhere they go.

The recovery reflex. Ask about a moment that went sideways for someone they were looking after. Listen for whether they describe what they did, or what they noticed the person needed. The first answer is operational. The second answer is the spark.

Most operators run sixty-minute interviews where the first thirty-five are technical and the spark questions never get asked. They’ve inverted the heuristic. They use the time backwards.

The dashboard is part of the selection mechanism

Hiring doesn’t end at the offer. It ends, if it ends at all, at the point where the new hire’s behaviour gets reinforced or punished by what the property measures.

A measurement is a selection mechanism whether the operator means it that way or not. Measure check-in speed and you select for speed. The agent who takes ninety seconds longer to look up, smile, ask the guest about the journey, read what the guest actually needs, and let the booking confirmation finish loading in the background is your highest-performing agent. Your dashboard scored them as your slowest. Three months of that, and the agent has either left for a property that recognises them, or stayed and adapted to the metric. The metric won. The hire converted.

This is the part operators tend to defend as “just how performance management works.” It is not. It is a counting exercise dressed as accountability. The KPI was built for a different era of work, when activity was a reasonable proxy for value and compliance was a reasonable proxy for performance. Neither is true any longer, and hospitality is the industry where the proxy breaks first. KPIs treat people as components in a system, and they get exactly that kind of behaviour back. Compliant. Transactional. Disengaged. The floor performs the metric and stops there. Anyone who has read a guest review describing service as “polite but cold” has read a KPI dashboard talking back through the room. The dashboard can’t see what it wasn’t designed to count. The recognition of a returning guest in the first ninety seconds of arrival. The unprompted hand-off to a colleague mid-problem. The departure that included an unscripted moment of acknowledgement. These are real. They are measurable. They are inconvenient to measure. That inconvenience is the entire reason most properties don’t measure them, and the entire reason the floor delivers what the dashboard rewards.

Value Impact Goals (VIGs™) are the framework most operators haven’t heard of. The framework isn’t new. I developed it more than three decades ago and have deployed it inside private equity portfolio companies and client companies ever since, hospitality among them. The logic inverts the KPI from the inside out. Anchor in Value (the purpose behind the work, the why the rate is being charged in the first place). Drive through Impact (the difference the work actually makes, tangible and intangible together). Let the Goals emerge as the byproduct of the alignment between those two. A KPI measures whether the agent processed the booking in seventy seconds. A VIG measures whether the guest who walked out three days later is the kind of guest who comes back, brings someone, and mentions the property by name without being asked. A KPI is rear-view. A VIG is live. A KPI counts what happened. A VIG measures whether anything is actually happening at the level the property charges its rate to deliver.

How to run a VIG framework belongs in another article. What belongs in this one is the recognition that the instrument exists. The inability of a KPI to capture the spark isn’t the limit of what’s measurable. It is the limit of the wrong instrument.

The interview is mutual

There’s a related point worth naming here, because it operates in the opposite direction and most operators forget it exists.

The spark candidate has options. They always have options. The hospitality labour market isn’t short of warm bodies; it’s short of candidates with the spark, and those candidates choose where they work. They’re interviewing the property at least as hard as the property is interviewing them. They’re reading the GM, the duty manager, the front office lead, the people who will be their direct reports and their leadership for as long as they choose to stay. Good people in hospitality don’t stay in one role at one site forever, and they shouldn’t. They want to move, to travel, to grow into the next thing. The sophisticated operator plans for that across multiple sites within the group and keeps a good person in the family. The unsophisticated operator loses them at the resignation conversation and watches them walk to the competitor down the road.

The interview is mutual. The candidate who has the spark is choosing a boss as much as choosing a role. The property that doesn’t show that candidate a leader worth working for is losing them to a property that does. Most operators don’t realise their senior team is being audited every time they sit across the table from a candidate worth hiring.

This is upstream of everything else. The recruitment touchpoint is also a leadership audit, and the audit happens whether the property knows it or not.

The standards installation has nothing to install onto

Hiring is half of the work. The other half is what Built To Own™ calls the standards installation: the discipline of writing, demonstrating, and reinforcing what “great” looks like until it’s the norm rather than the miracle.

The reason this loops back into the hiring conversation is that the standards work has nothing to install onto if the hire is wrong. You can’t run a recognition standard against a hire who isn’t oriented to notice the people in front of them. You can’t run a recovery standard against a hire who escalates every problem out of the building. The standards work assumes the underlying orientation. Hiring is where that orientation gets selected for or selected out.

This is also where the conversation about “service culture” usually goes wrong. Culture is not the lever. Standards are the lever. Hires are the substrate. Most properties are running culture initiatives on a substrate that was selected against the very behaviour the culture is trying to produce. The agency arrives, the deck gets built, the posters go up, the breakfast service remains efficient and unhappy. The substrate does not have the spark. The culture has nothing to grow on.

The labour problem is a selection problem

The labour problem most operators are trying to solve at the wage layer isn’t a wage problem. It is a selection problem. The funnel is producing the floor they’re now complaining about, and the funnel was designed to produce exactly that floor. Until the job spec, the interview structure, and the goal framework underneath them are rebuilt to select FOR the spark and teach TO the skill (instead of the inverse), the property will keep buying competence and keep delivering process. The wage will keep going up. The retention will keep going down. The service culture day will keep getting rescheduled.

Every operator complaining about the labour layer is right about one thing. The cost is real. The conversion is real. The economics are unforgiving. But the conversion doesn’t start at the wage line. It starts at the job spec, the interview question, and the goal framework underneath them.

The floor you walk past every morning is the floor you hired. The wage line is the bill for not having hired better.