How Much Time Do Sales Reps Spend Selling?
Salesforce's current data says 40% of the week. The 28% and 72% figures still circulating are from a survey fielded in 2022. Here is the full record.
About 40% of it. That is the figure in Salesforce’s State of Sales, seventh edition - 4,050 sales professionals across 22 countries, fielded in August and September 2025 and published on 3 February 2026. The widely repeated claims that reps spend “28% of their time selling” or “72% of their time not selling” are both real numbers, but they come from the fifth edition, fielded in August and September 2022. They are three years old and two editions out of date, and almost every article still quoting them presents them as current.
Key takeaways
- Current figure: 40%. Salesforce State of Sales, 7th edition - 4,050 respondents, 22 countries, fielded August–September 2025, published 3 February 2026.
- The famous 28%/72% pair is from the 5th edition - 7,775 respondents, 38 countries, fielded 24 August–30 September 2022. It was accurate when published and is not current.
- The 6th edition said 30% - 5,500 respondents, 27 countries, fielded 8 March–18 April 2024.
- “Selling” is a defined bucket, not a plain-English word. No edition defines it in words, but every edition’s own chart does: it is customer meetings plus prospecting. Research, quoting, planning and data entry all sit in the non-selling share.
- The sample is not all frontline reps. Salesforce surveys sales leaders, reps, partner reps, business development reps and sales operations staff together, so the headline n is not a count of reps answering a time-use question.
- It is self-report, not telemetry. No edition measures time by instrumentation. Treat all three figures as recalled estimates.
- Three editions, three numbers, one unanswered question: nobody has shown whether 28 → 30 → 40 is real improvement, a definitional shift, or a sample that changed shape.
What is the current number, and where does it come from?
The current answer is the seventh edition of Salesforce’s State of Sales. Its methodology is published: 4,050 sales professionals in 22 countries - Australia, Brazil, Canada, Denmark, Finland, France, Germany, India, Ireland, Italy, Japan, Mexico, the Netherlands, New Zealand, Norway, Portugal, Singapore, South Korea, Spain, Sweden, the United Kingdom and the United States - fielded across August and September 2025, published 3 February 2026. It reports that the average rep spends 40% of the week selling and 60% on non-selling work, and that Gen Z reps spend 35%, losing roughly two hours a week to manual data entry that more senior reps spend on research and relationships.
That is the number to cite in 2026. The reason so little writing uses it is that the older figures are stickier, rounder and more alarming.
How the number has moved across editions
| Edition | Fielded | Published | Sample | Countries | Selling | Not selling |
|---|---|---|---|---|---|---|
| 5th | 24 Aug – 30 Sep 2022 | © 2022 | 7,775 | 38 | 28% | 72% |
| 6th | 8 Mar – 18 Apr 2024 | 2024 | 5,500 | 27 | 30% | 70% |
| 7th | Aug – Sep 2025 | 3 Feb 2026 | 4,050 | 22 | 40% | 60% |
Three things are visible in that table that are invisible in any single quotation of it. The headline rose by twelve points in three years. The sample nearly halved. And the country count fell from 38 to 22. Any explanation of the first fact has to survive the second and third, and none of the published commentary attempts it.
What does Salesforce actually count as “selling”?
This is the question that decides whether the statistic means what people use it to mean, and no edition answers it in words. All three answer it in arithmetic.
The seventh edition itemises the week as: meeting with customers 22%, prospecting 18%, manually entering data 17%, creating quotes 16%, planning 13%, training 11%, other 3%. Those seven figures sum to 100, and the first two sum to exactly 40 - the headline selling share. The report never writes the definition down, but its own chart settles it: selling means meeting customers and prospecting. Everything else is the 60%.
The sixth edition is built the same way and itemises further:
| Activity | Share of week |
|---|---|
| Meeting in person with customers | 12% |
| Connecting virtually with customers | 9% |
| Prospecting | 8% |
| Generating quotes, proposals and approvals | 10% |
| Preparation and planning | 9% |
| Manual data entry | 9% |
| Internal meetings and training | 9% |
| Researching prospects | 9% |
| Administrative tasks | 9% |
| Prioritising leads and opportunities | 8% |
| Downtime | 8% |
Again the selling bucket is the first three lines - in-person meetings, virtual customer contact and prospecting - totalling 29 against a headline of 30.
That consistency matters, because the 72% figure is almost always deployed as though it describes 72% administrative waste. It does not. Writing the proposal, researching the prospect and preparing for the call are all skilled, deal-advancing work, and all of them sit in the non-selling bucket. The honest version of the claim is narrower: in the sixth edition roughly a fifth of the week went to data entry and administration specifically, and in the seventh, 17% went to manual data entry alone. That is still a real problem. It is simply not the problem the headline is used to sell.
Is the rise from 28% to 40% real?
Unresolved, and anyone who tells you otherwise is guessing. Three explanations compete, and the published data does not separate them.
- Genuine change. Sales tooling absorbed a great deal of manual work between 2022 and 2025. If data entry fell, selling share rises arithmetically without anyone selling more hours.
- Composition. The sample fell from 7,775 to 4,050 and the country count from 38 to 22. A smaller, differently distributed sample can move a mean on its own. Salesforce does not publish a like-for-like reweighting.
- Measurement. Every edition is self-report. People are poor estimators of their own time use, and the direction of that error is not stable across years or seniority levels.
There is a fourth consideration that cuts across all three: the respondent base is mixed. Salesforce surveys sales leaders, frontline reps, partner reps, business development reps and sales operations professionals under the single label “sales professionals.” The headline sample size is therefore not a count of quota-carrying reps reporting their own week, and no edition publishes the frontline-only subgroup for this question.
None of this makes the figure useless. It makes it a self-reported estimate from a mixed professional sample, which is what it has always been, and which is how it should be quoted.
Does any independent source agree?
Partly, and the comparison is useful.
HubSpot’s 2024 Sales Trends Report - 1,400+ sales professionals across the United States, United Kingdom, Japan, Canada, Australia, France and Germany, fielded August 2023 - puts actual selling at about two hours a day, with roughly an hour a day on administrative tasks. Two hours of an eight-hour day is 25%; the report separately quotes a 33% figure attributed to outside research rather than to its own sample, which is worth knowing before citing it.
So two independent survey programmes, fielded roughly a year apart, land in the same region: somewhere between a quarter and a third of the week going to customer-facing selling in 2022–2023, rising on Salesforce’s own measure by 2025. What neither programme offers is instrumentation. There is no widely published time-and-motion or CRM-telemetry study of rep hours at comparable scale, which is the single biggest gap in the evidence base and the reason the surveys deserve hedged language rather than confident restatement.
Who is citing this wrong, and how
The pattern is consistent enough to describe precisely.
- Undated restatement. The 28% or 72% figure appears in present tense - “reps spend 72% of their time not selling” - with no edition, no field date, and often no year. This is the most common error and the one that makes the number folklore rather than evidence.
- Edition drift. The 30% figure from the sixth edition and the 28% from the fifth are used interchangeably, sometimes in the same article, as though they were one statistic with rounding.
- Bucket inflation. The non-selling share is described as administrative work, when the source itself puts proposal generation, prospect research and planning in that bucket.
- Secondary citation. The number is credited to whichever blog the writer found it on, not to the report. Chains three and four links long are easy to find, and the edition is lost at the first link.
- Sample laundering. The headline n is quoted as a count of sales reps. It is a count of sales professionals - the fifth edition’s own methodology note describes a mix of reps, sales operations professionals, leaders and managers, all drawn from third-party panels.
- Unit substitution. A percentage of people is restated as a percentage of time. “X% of reps say admin slows them down” and “reps spend X% of the week on admin” are different claims, and they get swapped.
The simplest fix is also the best test of any article you read on this subject: if it does not name the edition and the field dates, it does not know how old its number is.
What the number actually tells a rep
Set the citation argument aside for a moment, because the underlying finding has survived three editions and two publishers, and it is the part that matters if you carry a quota.
Even at the most favourable reading available - the current 40% - the majority of your working week is spent on activity that is not selling, inside a system you do not control, on accounts you do not own. The percentage has moved. The structure has not. You can become more efficient inside it, and reps do; what you cannot do is convert that efficiency into an asset, because the efficiency belongs to the territory and the territory belongs to the company. That is the same arithmetic as the sales rep career ceiling, approached from the direction of the calendar instead of the compensation plan, and it is why what you actually own at your sales job is the harder question than how your week is divided.
The honest conclusion is not that 60% of your week is wasted. Much of it is skilled work. The conclusion is that all of it - the 40% and the 60% - accrues to someone else’s balance sheet.
That is the problem Rep2Owner exists to solve: Prep to Launch ™ is the incubator where a rep validates and launches a business they own, and Build to Exit ™ is the accelerator that takes it to a sale or a raise. If you are weighing the move while still carrying a number, start with can you start a business while carrying a sales quota, or read the full sales rep to business owner guide. When you are ready to see whether you qualify, the Prep to Launch ™ page is the next step.
Frequently asked questions
How much time do sales reps spend selling in 2026? About 40% of the working week, according to Salesforce’s State of Sales, seventh edition - 4,050 sales professionals across 22 countries, fielded August–September 2025 and published 3 February 2026. The other 60% goes to non-selling work including data entry, research, proposal generation and internal meetings.
Where does the “sales reps spend only 28% of their time selling” statistic come from? Salesforce’s State of Sales, fifth edition - 7,775 respondents across 38 countries, fielded 24 August–30 September 2022. It is accurate as a 2022 measurement. It is not the current figure, and articles presenting it in the present tense are three years and two editions behind.
Is the “72% of time not selling” figure accurate? It is the arithmetic complement of the fifth edition’s 28%, so it is accurate for 2022 and outdated now - the current complement is 60%. It is also routinely misdescribed: the source counts proposal writing, prospect research and planning as non-selling, so the 72% was never 72% of administrative busywork.
What counts as “selling” in these surveys? Salesforce never defines it in words, but the arithmetic of its own charts does. In the seventh edition, meeting with customers (22%) plus prospecting (18%) equals exactly the 40% headline. In the sixth, in-person meetings (12%), virtual customer contact (9%) and prospecting (8%) sit against a 30% headline. Research, quoting, planning, internal meetings, data entry and administration are all counted as non-selling.
Why did the number jump from 28% to 40%? Nobody has shown. Three explanations are live - real reduction in manual work through tooling, a changed sample (7,775 respondents in 38 countries down to 4,050 in 22), and self-report error - and the published data does not separate them. Any source stating a confident cause is going beyond the evidence.
Are these figures measured or self-reported? Self-reported. No edition of either survey instruments rep time directly, and no comparable large-scale time-and-motion or CRM-telemetry study is publicly available. The figures are recalled estimates and should be quoted as such.
Sources
- Salesforce, State of Sales, seventh edition - 4,050 sales professionals across 22 countries, fielded August–September 2025, published 3 February 2026; reps spend 40% of the working week selling and 60% on non-selling work, Gen Z reps 35%. Respondent base includes sales leaders, sales representatives, partner representatives, sales and business development representatives, and sales operations professionals.
- Salesforce, State of Sales, sixth edition - 5,500 sales professionals across 27 countries, fielded 8 March–18 April 2024; 30% of the week selling, with the itemised activity breakdown reproduced above.
- Salesforce, State of Sales, fifth edition (© 2022) - 7,775 respondents across 38 countries in North America, Latin America, Asia-Pacific and Europe, fielded 24 August–30 September 2022; 28% of the week selling, 72% on other tasks. The report states all respondents were third-party panelists and describes the base as a mix of sales reps, sales operations professionals, leaders and managers.
- HubSpot, 2024 Sales Trends Report - 1,400+ sales professionals across seven countries, fielded August 2023; approximately two hours a day actually selling and about one hour a day on administrative tasks. The report’s separately stated 33% figure is attributed to outside research rather than to its own sample.
Figures are drawn from the published editions named above. Where a cause for a change is not established in the source, this article says so rather than supplying one.