What to measure, and what measuring the wrong thing does
Every measure of a sales team is also an instruction, and the instruction is frequently not the one intended. Close rate per appointment, the most common single metric in home services, tells a rep that the appointment where the honest answer is a small repair is a failure. Do that for a year and you have built a team that recommends replacements, and the reviews will arrive about eighteen months after the policy did. What you measure is a design decision about behaviour, and most businesses make it by default rather than deliberately.
What close rate on its own actually incentivises
Consider the appointment in module 3.3, where the truthful answer is a four-hundred-pound repair rather than a nine-thousand-pound replacement. Under close rate per appointment that visit is neutral at best. Under revenue per appointment it is a disaster. Under both, the rep who tells the truth is punished relative to the rep who does not.
Reps do not respond to this by becoming dishonest. They respond by drifting: emphasising the worst finding, mentioning the repair option briefly, letting the customer's own worry do the work. Everything said remains true and the recommendation bends, and no individual conversation looks wrong.
The cost arrives late and lands somewhere else: second opinions, cancellations, reviews mentioning being oversold, and a referral rate that slowly falls while the sales numbers look fine. By the time it is visible in revenue, the incentive has been running for two years.
None of this is an argument against measuring close rate. It is an argument against measuring only close rate, because a single metric on a team that has real discretion will always be gamed in the direction the metric points.
Measure the process as well as the result
Outcome measures tell you what happened and nothing about why, so they cannot be coached. Close rate went from twenty-two to eighteen percent tells an owner there is a problem and nothing about where it is.
Process measures are the ones that can be acted on. Was the follow-up sent the same day. Was the lead source recorded correctly. Was permission to record asked. Were the promises to the crew captured. Every one of those is checkable, every one is within the rep's control, and every one connects to a lesson someone can be coached on.
The combination is what works: outcomes to know whether the business is healthy, process to know what to do about it. Businesses that track only outcomes end up coaching by exhortation, which is where "we need to close harder" comes from.
And the specific thing worth adding that almost nobody has: **honest no-sale as a recorded, non-punished outcome.** A rep who correctly told a customer they did not need the work should be able to log that as a good appointment. Otherwise you are asking for integrity while paying for the opposite, and the pay wins.
A measurement set that does not backfire
Outcome and process together, and the loss reasons honest enough to be useful.
- Close rate, but segmented by lead source
- An unsegmented close rate mixes emergency leads and cold canvass and produces a number that means nothing. Segmented, it tells you which channels produce work rather than appointments, which is a different and much more useful question. We have written that up separately, linked below.
- Average ticket, alongside close rate, never instead of it
- Either one alone is gameable in an obvious direction. Together they constrain each other, and a rep whose average ticket is well above the team's is worth listening to and also worth checking.
- Same-day follow-up rate
- The single most valuable process metric available, because it is binary, entirely within the rep's control, and directly connected to outcomes. It is also the one most often not measured at all.
- Honest no-sale, logged and not penalised
- The appointment where the answer was a small repair or nothing. Track it, name it, and make sure the rep who did it is not worse off. This is the metric that decides whether your other metrics produce overselling.
- Real loss reasons, with a trigger and a date
- Not "price" for everything, which is what most systems contain and which hides lead-time problems, competitor problems and individual-rep problems behind one word. Module 8.2 covers what a useful loss record looks like.
What it sounds like
Two owners looking at the same drop in close rate, from 24% to 19% over a quarter.
What to notice. Both owners saw the same number. Owner A had only an outcome, so the only available action was exhortation and training, and nothing changed. Owner B had process measures alongside it and could see that the follow-up collapse coincided with a system change, which is a fixable operational problem rather than a sales-skill problem. Note the second finding, buried in the loss reasons: a lead-time issue that no amount of objection-handling training would have touched.
The mistake: one number, and a leaderboard
A single headline metric on a wall is the most common measurement design in the trade and it is the one most reliably gamed. Whatever it is, the team will optimise for it precisely, including in ways nobody intended, and the unintended part will not be visible in the metric by definition.
Leaderboards add a second problem, which is that they reward the reps with the best leads. A rep working referrals and emergency calls will beat a rep working canvass leads at identical skill, and a public ranking teaches everyone that lead allocation matters more than technique, which is true and demoralising. If you rank anything, rank it within lead source.
The third mistake is measuring what the system happens to record rather than what matters. CRM completion rates get tracked because they are easy, and a team optimising for CRM completion produces populated fields containing nothing, which module 8.2 and the empty-notes problem both describe.
And the quietest one: measuring reps and not the business. When close rate falls, the cause is at least as likely to be lead quality, pricing, lead times or a new competitor as it is to be rep skill. A measurement set that can only produce conclusions about reps will always produce conclusions about reps.
Auditing your own metrics
- If a rep honestly told a customer they did not need the work, does any number reward that?
- Is close rate segmented by lead source, or aggregated into a meaningless average?
- Do we measure anything that is a process rather than an outcome?
- Do we know our same-day follow-up rate?
- What percentage of our loss reasons say "price", and do we actually believe it?
- Does any loss record carry a trigger and a revisit date?
- If close rate fell next quarter, could our numbers distinguish a rep problem from a lead-quality or lead-time problem?
- Is anything on a leaderboard that is mostly determined by lead allocation?
Read next
- Close rate by lead source The join most businesses cannot make, and why it is the number that changes marketing decisions.
- When it is not as bad as they think The appointment your metrics either reward or punish.
- The no that is not final Why "lost on price" hides every real problem a business has.
Is close rate a bad metric?
It is a good metric and a bad only-metric. On its own, on a team with discretion over what to recommend, it pays for overselling and the bill arrives late as reviews and lost referrals. Paired with average ticket, a process measure and honest loss reasons, it is useful.
How do you measure an honest no-sale without it being abused?
The same way any qualitative outcome is handled: it is logged with a reason and it is reviewable. If someone has recordings, it is checkable directly. The risk of a rep over-claiming it is real and much smaller than the risk of a team that is never allowed to claim it at all.
We are three reps. Is this overkill?
At three reps you probably do not need dashboards, and you do need the incentive question answered, because it applies at any size. Knowing your same-day follow-up rate and not accepting "price" as a universal loss reason are both free.
What is the single most useful number to start with?
Same-day follow-up rate, if you are not already tracking it. It is binary, fully within the rep's control, connected to outcomes, and in most businesses it is materially lower than the owner assumes.