Sales Productivity: Four Different Ratios Wear the Same Name

Four people sit in a quarterly review and agree that sales productivity is the problem. The VP means revenue per head is down. The operations lead means reps are spending their week in the CRM. The enablement manager means ramp is too slow. The founder means the team costs more than it did last year and books the same number of meetings. Everyone nods, a tooling budget gets approved, and the four problems continue independently.

That meeting is the reliable consequence of a term that names at least four different ratios. Sales productivity has no settled arithmetic, and the version somebody has in mind decides what they will do next.

Four ratios, one name

Reading across the pages that rank for this term, the definitions offered are not variations on one formula. They are different formulas, with different inputs, pointing at different fixes.

Output divided by input is the most common framing, usually illustrated with deals closed over hours worked. Revenue per rep is the version leadership tends to hold, because it is the one that appears in a board pack. Share of time spent selling is the version that produces the admin argument. Handle time, meaning how long a rep takes to respond to a lead or work through an interaction, is the version that produces the automation argument.

Each is defensible. None of them is convertible into the others, because they do not share a denominator. A team can improve share of time selling while revenue per rep falls, if the freed hours go into the same unproductive accounts. A team can improve revenue per rep while handle time gets worse, if the reps stop chasing everything and spend longer on the deals that were always going to close.

Revenue per repThe board's version

• Denominator is headcount
• Improves by cutting heads or raising deal size
• Says nothing about how the week is spent
• Moves on segment mix as much as on effort

Share of time sellingThe admin version

• Denominator is hours in the week
• Improves by removing non-selling work
• Requires a definition of what counts as selling
• Can rise with no revenue effect at all

Output per inputThe efficiency version

• Deals or pipeline over hours or cost
• Improves by better targeting as much as by speed
• Sensitive to which deals were winnable anyway
• The only one that can be gamed by taking easier accounts

The figure that circulates, and the page it is attributed to

There is one number attached to this subject more than any other. The claim, attributed on most pages that rank for this term to Salesforce, is that reps spend roughly 30 percent of their time actually selling, with the rest going to admin, data entry, internal meetings and quoting.

That attribution is worth handling carefully. Salesforce's own page on sales productivity was fetched for this piece and searched for the claim, and the figure is not in it. What that page does carry is different material: it states that 67 percent of sellers say they do not expect to meet their annual quotas, and that 81 percent of sales reps say team selling helps them close while roughly the same share find aligning with other sellers at least somewhat challenging.

The honest statement is narrow. This is one page, and a figure absent from one page can be published on another, so nothing here establishes that the 30 percent claim has no source. What it does establish is that the page most commonly pointed at for it does not carry it, which is enough reason not to repeat the number here with a citation nobody has checked.

The general habit is worth more than the specific finding. A statistic that every page in a category quotes and none of them sources is the normal condition of a category, not an anomaly, and the cost of repeating one is that your page becomes the next unchecked citation for somebody else.

Which ratio you pick decides what the team does

The choice is not academic, because each ratio contains an instruction.

Pick revenue per rep and the available levers are headcount and deal size, so the conversation becomes about who to cut and which segment to move up into. Pick share of time selling and the levers are process and tooling, so the conversation becomes about what to automate. Pick output per input and the lever is targeting, because the cheapest way to raise deals per hour is to spend the hours on accounts more likely to buy.

The last of those is the one most often skipped, and it is usually the largest. A rep working a list where a meaningful share of accounts cannot buy is unproductive by every one of the four definitions at once, and no amount of freed admin time changes it. Our guide to sales rep coaching makes the same point from the individual side: when a rep is flat, the three layers underneath are the list, the message and the live conversation, and only the third is coachable in the ordinary sense. A quarter of flat performance that traces to the first two gets absorbed as a personal development problem.

The equivalent at team level is that a productivity number carrying a targeting failure inside it reads as an effort failure, and effort is the lever people reach for.

The admin argument, and the part it usually gets wrong

The share-of-time-selling framing has a genuine finding underneath it, which is that expensive people spend a lot of their week on work that does not require expensive people. Our own piece on restructuring a sales team around high-leverage work sets out the case: the old model hands every rep the whole chain, from list building through enrichment and sequence writing to selling, and only the last step is what they were hired for. When every rep builds their own list, multiple reps build overlapping lists, and that is what the structure produces rather than a discipline problem.

Where the standard version goes wrong is the conclusion. The usual recommendation on this SERP is to buy a platform that removes the admin, and that is an incomplete answer for a reason the same piece names. Tools do not change who does the work. A team where every rep owns the whole chain will use a better tool to do the same low-leverage tasks slightly faster.

The change that moves the number is an allocation decision: the low-leverage half of the chain goes to one owner serving the whole team, and the reps keep the half that needs judgement. The software makes that possible and does not make it happen. That is also why the honest version of the cost comparison includes an operator rather than comparing headcount against a licence, which is worked through in more detail alongside the stack itself.

Before you call it a productivity problem

• ✓Yes: You have named which of the four ratios you mean, in writing
• ✓Yes: The rate carries its denominator wherever it is reported
• ✓Yes: You have checked whether the list can buy before checking whether the reps are working
• ✓Yes: Ramping reps are shown separately from fully ramped ones
• ~Depends: The comparison period has the same segment mix
• ✕No: The plan is to buy a tool before the allocation question is answered
• ✕No: Activity volume is on the scoreboard as the productivity measure

Two arithmetic traps that make the number lie

Both are avoidable and both are common enough to be worth naming.

The first is the missing denominator. A rate computed over an already-filtered population describes the survivors, so a number that looks excellent can be measuring a small remainder of what went in. Our piece on sales operations KPIs works this through with an illustrative batch and states the habit that prevents it: print the input count beside the accepted count beside the flagged count, and treat any rate whose denominator is unnamed as a failed check rather than a passing one.

The second is tenure. A team at full quota coverage on paper, where a third of the quota sits with people still inside their ramp interval, has a capacity problem that the coverage figure hides. Productivity per rep computed across a team with several new starters is measuring the hiring plan as much as the team, and the fix is to show ramping and ramped separately rather than to explain the blend afterwards. Where the headcount a target implies gets worked out properly, including the interval that keeps getting left out, is sales capacity planning.

The following arithmetic is invented purely to show the shape of the second trap, and no figure in it is a measurement of any team.

8Reps on the team

Invented example, not a measured figure

3Still inside their ramp interval

Producing at a fraction of the steady-state rate

37%Share of headcount carrying the average down

Arithmetic on the invented numbers above

0Change in what any individual rep did

The average moved on hiring, not on effort

What to measure instead

The useful set is small, and its usefulness comes from being paired rather than from being long.

Name one ratio as the headline and write down its definition and its denominator. Report it split by tenure. Beside it, report the thing it can be improved at the expense of, because each of the four ratios has one: revenue per rep can improve by cutting heads, share of time selling can improve with no revenue effect, output per input can improve by taking easier accounts. A number with its companion beside it is honest under pressure, and the same number alone is flattering.

Then read the input side before the output side. Whether the accounts on the list can buy, and whether enough of them are being reached at all, explains more variance in every one of these ratios than anything a rep does in an hour. What belongs on an individual's review once those are separated is set out in the sales rep scorecard, which draws the same line between what the rep controls and what the plan decided for them.

Where we sit

RevenueFlow runs cold email and LinkedIn outbound for clients, paid per qualified meeting against criteria agreed in writing before anything sends. That is our documented policy rather than a claim about anybody's productivity, and it is relevant here for one structural reason: it moves the list and message layers off the rep's week entirely, which is the allocation change rather than a tooling one.

Each campaign carries a single message, with no bumps and no thread replies, so a later approach to the same audience is a separate campaign with its own reason for contact. The productivity consequence is that there is no chasing activity to measure, and the input side of the arithmetic is a list decision rather than a touch count.

The short version

Sales productivity names at least four ratios: revenue per rep, share of time spent selling, output over input, and handle time. They do not share a denominator and improving one can move another the wrong way, so the first useful step is writing down which one you mean.

The 30 percent selling-time figure that anchors most writing on this subject is not on the Salesforce page it is usually attributed to, which is a reason to leave it out rather than proof it has no source. The admin finding underneath it is real, and the fix is an allocation decision rather than a purchase, because tools do not change who does the work.

Print the denominator, split by tenure, pair every headline number with the thing it can be improved at the expense of, and check whether the list can buy before concluding that the team is not working.

If the constraint turns out to be the number of qualified conversations reaching the team, that is the half we run: see what a first campaign produces.

The Salesforce figures quoted above are statements on Salesforce's own sales productivity page, fetched 22 August 2026, with a dated snapshot retained. The absence of the 30 percent claim is reported for that page only and is not a claim about Salesforce's other surfaces.