Lead Conversion Rate Excludes the Leads You Never Touched

Why this matters

This number gets better when you work fewer leads. Not as a quirk at the edges, as the central behaviour: a shop drowning in enquiries that only ever calls back the promising ones reports a strong conversion rate while leaving most of its opportunity on the floor, and the number congratulates it every month for doing so. The busier the shop gets, the better the figure looks and the worse the business does.

What the denominator is, and what falls out of it

The usual construction: leads that reached a won state, over leads that got past the new state. Past new means somebody logged something against it, a call, a note, a status change.

A lead that arrived and was never touched has no activity, so it never leaves the new state, so it is in neither the numerator nor the denominator. It is not counted as a loss. It is not counted at all.

That is the same family as a rate computed only over the rows carrying a marker, which a sibling card owns under the name opt-in denominator: the rate improves the less you record. Here the marker is human effort, so the rate improves the less you do.

The inversion, worked

Same shop, two quarters.

Quarter one. 220 leads received. 96 worked. 34 won.

• Conversion among worked leads: 34 of 96, 35.4 percent
• Touch rate: 96 of 220, 43.6 percent
• Conversion among leads received: 34 of 220, 15.5 percent

Quarter two. The phone gets busier and the office falls behind, so the person covering it works only the leads that look best. 240 leads received. 60 worked. 27 won.

• Conversion among worked leads: 27 of 60, 45.0 percent
• Touch rate: 60 of 240, 25.0 percent
• Conversion among leads received: 27 of 240, 11.3 percent

Read the headline on its own and quarter two was the better quarter: the rate rose 9.6 points, from 35.4 to 45.0.

Read the counts and quarter two booked 7 fewer jobs, 27 against 34, from 20 more leads, 240 against 220. Conversion among leads received fell 4.2 points, from 15.5 to 11.3.

Every one of those figures is correct. Not one of them is about selling: the closest, conversion among worked leads, still has the unreachable leads sitting inside it.

Why selection looks like skill

Nothing improved in quarter two. The office picked the leads that looked most likely to buy and worked those, which raises the average quality of the worked set. A higher rate on a better-selected population is arithmetic, not performance.

It is worth being precise about the direction, because the same mechanism runs the other way and shops misread that too. A shop that starts working its marginal leads, correctly, will watch the headline rate fall while booked jobs rise. If the only number in the review is the rate, that shop will be told to stop doing the thing that is working.

The identity that ties the three numbers together

The three figures are not three independent measurements. They are one relationship:

conversion among worked leads times the touch rate equals conversion among leads received.

Quarter one: 35.4 percent of 96 worked, times 96 of 220 received, gives 34 of 220, or 15.5 percent. Quarter two: 45.0 percent times 25.0 percent gives 11.3 percent. All three terms are count shares of counts, so they multiply cleanly, and any two of them determine the third.

That is why the headline alone can never be diagnostic. It is one factor of a product, and the other factor moved further than it did.

Not everything untouched is a lost opportunity

The naive fix, count every lead received, over-corrects. Some enquiries are not leads: wrong trade, outside the service area, a supplier calling, an obvious junk submission. Ignoring those is correct, and a shop that counts them punishes itself for having a public phone number.

The distinction that matters is between disqualified and never reached, and it is a distinction only the shop can make, by marking the first group with a reason at the moment somebody looks at it.

Split quarter one's 124 untouched leads that way: 46 disqualified on sight, 78 never reached.

• Qualified leads received: 220 minus 46, or 174
• Conversion among qualified leads received: 34 of 174, 19.5 percent
• Reach rate among qualified leads: 96 of 174, 55.2 percent
• The identity still holds: 35.4 percent times 55.2 percent is 19.5 percent

That set is the honest one. It says the shop converts 35.4 percent of what it works, works 55.2 percent of what is worth working, and ends up at 19.5 percent of the real opportunity. The second number is where the money is.

Marking disqualifications costs a few seconds per lead and it is the only thing that separates a capacity problem from a lead-quality problem. Without it, a shop buying bad leads and a shop too busy to answer good ones read identically.

What counts as touched, and the rate hiding underneath it

A touch is usually one logged outbound attempt. Nobody has to have answered.

That means the worked denominator holds leads nobody actually spoke to, and those convert at zero by construction, so conversion among worked leads is dragged down by a reachability problem and reported as a selling problem.

Split quarter one's 96 worked leads again: 61 reached a two-way conversation, 35 were attempted and never answered.

• Contact rate among worked leads: 61 of 96, 63.5 percent
• Conversion among contacted leads: 34 of 61, 55.7 percent

The same shop, the same quarter, the same 34 jobs, reads 35.4 percent or 55.7 percent depending only on where the denominator stops. Both are real. The 20.3-point gap between them is not selling ability, it is the 35 leads that never picked up.

Stack the whole chain and it multiplies out, which is the thing worth keeping:

qualified received, times reach rate, times contact rate, times conversion among contacted, equals jobs won.

For quarter one: 174 qualified, 55.2 percent reached, 63.5 percent of those contacted, 55.7 percent of those won. Multiply the three rates and you get 19.5 percent, which is 34 of 174. Each term is a count share of counts over the population named immediately to its left, so the chain is legitimate arithmetic rather than a metaphor, and every term has a different owner and a different fix.

Is the never-reached pile worth staffing

Before hiring or reassigning anyone, find out what is in the 78.

Take a random sample of 25 of them, not the 25 that look best, or you have reproduced the selection bias you are trying to measure. Call them. Record only what they were: real work in your area that you could have done, or not.

Size the trigger before you dial, from your own numbers. If 4 of the 25 turn out to have been real and reachable, that is 16 percent, so about 12.5 of the 78 were real work. At this shop's own worked conversion of 35.4 percent, those 12.5 opportunities are about 4.4 more wins, which is 13 percent more booked jobs than the 34 it managed. Set the trigger where clearing it means at least a tenth more booked jobs, and for this shop that is 4 of 25.

Two honest limits on that test. Four of twenty-five is thin evidence, and a single unusual week produces it, so read a clear result as a reason to sample another 25 rather than as a finding. And 25 is an audit sample for a yes-or-no question, which is a different thing from the 30-record floor a rate needs before it can be ranked against another rate.

If the sample comes back with almost nothing in it, the untouched pile is genuinely junk, the lead source is the problem, and no amount of staffing fixes it.

Two ways to book the same seven jobs

Quarter two lost 7 jobs against quarter one. There are exactly two levers, and sizing both is what makes the choice obvious.

Lever one: close better. Hold the worked count at 60 and lift conversion among worked leads until it produces 34 wins. That needs 34 of 60, or 56.7 percent, against the current 45.0 percent. An 11.7-point lift in closing rate inside one quarter is not something anybody reliably delivers, and attempting it usually means discounting.

Lever two: work more. Raise the worked count until it produces 34 wins, at the rate a broader set actually converts at rather than the rate a hand-picked one does. The 60 already worked are the best-looking ones and hold 45.0 percent, giving 27 wins; the leads added back convert nearer quarter one's blended 35.4 percent, so 7 more wins takes at least 20 more leads. That is 80 worked against the current 60, about 1.5 more leads worked per working week.

Both land on 34 booked jobs. One is a skill transformation; the other is a lead and a half more worked per week, which is a few more calls. That comparison is only visible when the touch rate is printed beside the conversion rate, which is the whole argument for printing all three.

Put the chain on one line in the review, always: leads received, disqualified, worked, contacted, won, with the rates between them. A conversion rate on its own is a number nobody can act on and everybody can misread, and each term in the chain has a different owner. Reach is a staffing and hours question. Contact is a timing question, mostly about when you call back rather than how often. Conversion among contacted is the only one that belongs to whoever is doing the selling, and it is the only one worth putting in front of them.

References

• See related: First Response Time and the Window That Decides the Job - the reason a never-reached lead is usually gone rather than waiting
• See related: Average Sales Cycle Describes Winners Only - the companion exclusion on the other side of the funnel
• See related: Pipeline Value Is Not a Forecast Until You Weight It - what the never-reached leads are doing to the pipeline figure meanwhile
• See related: The Close Rate Improved Because They Stopped Quoting - the same inversion one stage later, at the estimate rather than the lead