Seller selling time: only 40% of the week, survey finds

A B2B seller’s selling time sits close to 40% of the week, according to the seventh edition of State of Sales, a Salesforce report published in 2026. It is a CRM vendor survey of 4,050 sales professionals in 22 countries, 250 of them in Brazil, run as an anonymous questionnaire between August and September 2025. Within those 40%, the report adds up meetings with customers (22%) and prospecting (18%). The other 60% goes to creating quotes, planning, entering data, training and other tasks. Bain, in a 2025 report, arrives at a lower figure of about 25% of time spent selling to customers. I read these numbers as a picture of capacity: the company hires a seller and, in practice, gets two days of selling per week, and the cheapest way to grow sales capacity is usually to give the team back the hours the operation itself consumes.

At Iasbeck&Co we implement CRM and AI agents in B2B operations, and the complaint that sellers “live filling in the system” shows up in almost every assessment. The board then wants to know whether automation and AI solve it, and the answer depends on where the hours are going and on what leadership does with the hours left over. I wrote this for the board, sales and revenue operations leaders, and HR.

How much time do sellers spend selling, according to research?

The available measurements range from a quarter to two fifths of the week, and the gap comes more from the definition of “selling” than from any real change in the work. Salesforce measures by self-report and counts prospecting inside the selling share. Bain states that sellers “may spend only about 25% of their time actually selling to customers” and does not publish the sample or method behind the number on the page, which appears as the firm’s own estimate in a chart credited to Bain. Both sources agree on direction: most of the week goes to activities the customer never sees.

The Salesforce series helps show how much the figure moves between editions. In the sixth edition, with 5,500 professionals in 27 countries surveyed between March and April 2024, sellers reported spending 70% of their time on non-selling tasks, such as administrative work and meeting preparation. In the seventh edition, the number fell to 60%. I would not read that drop as proof that AI gave sellers ten points of time back, because the sample size changed, the country list changed and so did the way activities were grouped.

The most useful data in the Salesforce survey is the activity breakdown. Sellers spend 17% of the week creating quotes, more than the 13% they spend entering data into the system, and 16% planning. Leaders usually blame the CRM for lost time, and the data shows quoting weighs more than record keeping.

Source	Year	Sample and method	Main finding
Salesforce, State of Sales 7th edition (vendor survey)	2026	4,050 sales professionals in 22 countries (250 in Brazil), anonymous questionnaire with third-party panelists, Aug. to Sept. 2025	Sellers spend 40% of the week selling (22% with customers and 18% prospecting); 42% feel overwhelmed by too many tools
Salesforce, State of Sales 6th edition (vendor survey)	2024	5,500 sales professionals in 27 countries, Mar. to Apr. 2024	Sellers report 70% of their time on non-selling tasks
Bain, Technology Report	2025	Firm’s estimate, no sample published on the page	Sellers may spend only about 25% of their time selling; AI could double that share
McKinsey	2020	McKinsey Global Institute research; top-down assessment and interviews	About a third of sales and sales operations tasks can be automated; only one in four companies had automated at least one sales process
Gartner	2023	501 sellers, Mar. to Apr. 2023, on their relationship with technology	77% say they struggle to complete their assigned tasks efficiently
Gartner	2026	Survey of 210 CSOs and senior sales leaders, Jan. to Feb. 2026; the release does not detail how hours were measured	AI saves sellers an average of 4.8 hours per week; 72% of organizations reinvest little of that time in high-value activities
Tarafdar, Pullins and Ragu-Nathan, Information Systems Journal	2015	Questionnaire with 237 institutional sales professionals	Technostress creators are negatively associated with seller performance
Delpechitre, Black and Farrish, Journal of Business & Industrial Marketing	2019	Over 200 salespeople at a firm serving small and medium businesses, structural equation modeling	Technology overload has linear or curved relationships with role stress, effort and performance

Where does a seller’s week go?

I converted the Salesforce breakdown into hours of a 40-hour week. The conversion is mine, it illustrates the order of magnitude and assumes a working week the report does not state. The last column is a working hypothesis, based on what I see in the operations I serve.

Activity (Salesforce, 2026)	% of the week	Hours in a 40 h week	What operations can do
Meetings with customers	22%	8.8 h	Protect the calendar and cut internal meetings that fall in selling hours
Prospecting	18%	7.2 h	Account research and first draft by an agent, with cadence defined in the CRM
Creating quotes	17%	6.8 h	Product and price catalog in the CRM, proposal templates and discount approval by authority level
Planning	16%	6.4 h	Weekly priority list prepared by the system from the pipeline
Manual data entry	13%	5.2 h	Automatic logging of email, calendar and meeting summaries, reviewed by the seller
Training	11%	4.4 h	Short training, in the company’s own CRM, tied to the process that changed
Other	3%	1.2 h	No specific action

Training shows up as non-selling time, and I would not treat those 4.4 hours as waste, because a seller who masters the price catalog wins back part of the quoting time. Planning becomes a problem when the seller builds by hand, every Monday, a list the CRM could deliver ready-made.

In operations with many items and discounts, the seller builds the proposal in a spreadsheet, asks for approval by email, waits and then copies the values into the CRM. Each step takes time and creates data that can diverge from the system, and fixing that flow gives hours back and improves the database. The design of the proposal itself is covered in B2B sales proposal.

Does too much technology hurt sellers?

Academic research on technology overload in sales says yes, past a certain point. The study by Tarafdar, Pullins and Ragu-Nathan (2015), with 237 institutional sales professionals, found a negative association between technostress creators and seller performance. The authors show that building technology competence reduces part of the effect, and that offsetting the drop in overall performance takes more: the seller’s confidence in their own ability to use technology, information systems literacy and involvement in systems initiatives. A screen tutorial teaches technical competence and leaves the rest out.

The study by Delpechitre, Black and Farrish (2019), with over 200 salespeople at a firm serving small and medium businesses, split overload into three dimensions: too many features, too much information and too much communication. The relationships with stress, effort and performance came out linear in some cases and curved in others, and the authors conclude that the benefit of more technology does not grow in a straight line. They recommend that managers control the pace of change and the amount of information and functionality they hand to the team. Both studies measure association at a single point in time, in specific samples, and do not prove that the tool caused the drop in performance.

Market surveys show the same symptom from another angle. In the 2026 Salesforce edition, only a third of sales teams use a single platform, the rest combine an average of eight tools per team, and 42% of sellers say they feel overwhelmed by too many tools. In a 2023 Gartner survey of 501 sellers, designed to understand their reactions to technology, 77% said they struggle to complete their assigned tasks efficiently. The tool inventory and the decision on what stays are covered in RevOps stack.

What changes with automation and AI?

The most quoted estimate of automation potential in sales comes from McKinsey, in 2020: about a third of sales and sales operations tasks could be automated with the technology of the time. The text itself warns, in a footnote, that the number is directional, based on a top-down assessment and interviews, and that the exact figure depends on each company’s starting point. The same piece includes an observation I use often with boards: according to McKinsey, high-performing sellers spend on average 20% to 25% more time with customers than lower performers. It is a correlation, and it can run both ways, because the seller who closes more also earns more meetings.

Bain, in 2025, estimates that AI could double selling time by taking on the low-value work around selling, and reports early gains of 30% or more in win rates. The text does not say how many cases support that number.

The figure I consider most important in this series comes from Gartner, in May 2026. According to the firm, AI tools save sellers an average of 4.8 hours per week, and 72% of sales organizations reinvest little of that time in high-value activities. In the survey of 210 CSOs and senior sales leaders, organizations that achieved moderate to large time savings and reinvested those hours in high-impact sales activities were 2.2 times more likely to exceed customer growth goals. Saved time dissolves into the routine if nobody decides where it goes, and sales leadership makes that call.

How do you give hours back to sellers without losing data?

The risk in every productivity initiative is cutting the record along with the bureaucracy. If the company removes required fields to spare the seller, the sales forecast and any AI agent that reads the CRM end up working with less information. The approach that works replaces typing with capture and keeps the seller as reviewer of what the system recorded.

Capture instead of typing

Email, calendar and calls can flow into the CRM automatically, associated with the contact and the deal. The meeting summary, generated from the transcript, becomes a note with next step and date, and the seller checks and approves it in a minute, with a more complete record than they would have typed. The condition is clear rules for association and duplicates, the subject of CRM data quality.

Require fewer fields, at the right moment

A required field at deal creation produces false entries, because the seller does not know the answer yet. Spread the requirements across funnel stages, with few fields per stage and each one tied to a decision someone makes with that data. Sellers accept logging when they see it used in the pipeline meeting, and the reasons they work around the system are in CRM adoption.

Take quoting out of the spreadsheet

A product and price catalog inside the CRM, proposal templates and discount approval levels inside the system cut the back-and-forth by email. The time saved comes with value and discount data that finance can audit.

Prepare planning and prospecting

An agent can build the week’s priority list, with stalled deals, proposals without a reply and accounts showing buying signals, and can research the company before first contact and write a draft. The manager gets the same kind of support preparing the pipeline meeting and the 1:1, the subject of AI for sales management.

Cut internal meetings

The surveys cited here do not isolate internal meetings as a category, so I treat this point as field observation. A pipeline meeting run inside the CRM, with an exception list prepared the day before, runs shorter than one that walks through the whole funnel, and the weekly spreadsheet report loses its reason to exist when the dashboard answers the same questions.

What this data does not say

Time allocation numbers come from self-report. Sellers estimate their own week from memory, and the estimate carries their fatigue with the system. None of the sources in this text measured time through activity logs or observation.

The definition of selling changes between sources and even within the same source. In the 2026 Salesforce report, the chart puts prospecting inside the 40% of selling, while the text on the same page talks about non-selling work “like data entry and prospecting.” Bain does not say what counts as selling. Comparing 25% with 40% without looking at the definition leads to the wrong conclusion.

Vendor surveys have a stake in the result. Whoever sells CRM and AI agents benefits when the market believes sellers waste time and that technology gives it back. The numbers come with published sample and method, and the reading needs to separate the data from the recommendation.

The percentages are global averages. The Salesforce sample has 250 respondents in Brazil, and the report does not publish a breakdown by country. A technical sale with a six-month cycle and a transactional phone sale have very different weeks, and the average hides both.

No source proves causation. The McKinsey figure on high performers and the Gartner figure on reinvested hours show association. Better-managed companies may reinvest the time and grow more without one causing the other.

What leadership does with these numbers

Market numbers open the conversation, and the decision on where to act depends on your own team’s week. The plan below fits in a quarter.

1. Measure your team’s week for two weeks. Ask a sample of sellers to log their time in 30-minute blocks, using the same categories as the table above, and cross-check with CRM activity records. The gap between what is reported and what is logged is already a finding.
2. Pick the two biggest losses that depend on operations. Quoting, data entry and planning usually top the list, and each has a different owner: pricing with finance, logging with RevOps, priorities with the manager.
3. Define the data that cannot be lost. Before cutting any field or step, list what the forecast, the revenue metrics and the AI agents need to read, and make sure automatic capture covers that minimum.
4. Automate with review. Turn on email and calendar capture, meeting summaries and proposal drafts, with the seller approving what goes into the CRM.
5. Decide where the hours go. Write down, before the pilot, how much of the freed time becomes customer meetings, prospecting or a larger book, and track it in the CRM.
6. Repeat the measurement at 90 days. Compare the new week with the first, along with conversion and meeting volume, in the RevOps metrics the board already follows.

What HR and L&D need to consider

Research on technostress in sales points to L&D. Tarafdar and colleagues show that technical competence alone does not offset the drop in performance, and that sellers need confidence in their use of technology, information systems literacy and involvement in technology initiatives. Training works better in role-based groups, inside the company’s own CRM, on the process that changed and with room for sellers to point out what gets in the way. HR spots the need in paid tools with little use, overload complaints in engagement surveys and experienced sellers going back to spreadsheets. To measure, compare reported hours and CRM records before and after the training, in the same group.

Giving selling time back to sellers requires an agreement between teams on what each one records, when and for which decision, and a shared yardstick to measure the effect. That is the work of the in-company Revenue Operations training, built on the company’s own process and data, with groups for the board, leaders and operations. It delivers the SLA between marketing, sales and CS, with what passes from one team to the next and with which data, and the RevOps metrics matrix leadership uses to track sales capacity and results.

Frequently asked questions

How much time do sellers spend selling?

In the most recent Salesforce survey, from 2026, with 4,050 sales professionals in 22 countries, sellers spend 40% of the week in customer meetings and prospecting. Bain, in 2025, estimates about 25% of time actually selling. The difference comes mostly from the definition of selling and from each source’s method.

What consumes most of a seller’s non-selling time?

In the Salesforce breakdown, creating quotes takes 17% of the week, planning 16%, data entry 13% and training 11%. Quoting weighs more than data entry, which puts pricing, catalog and discount approval at the center of any productivity project.

Does AI increase seller selling time?

Gartner calculates that AI tools save sellers an average of 4.8 hours per week, and found 72% of organizations reinvesting little of that time in high-value activities. AI frees hours, and leadership decides whether they become customer meetings or dissolve into the routine.

Does cutting CRM fields improve sales productivity?

It reduces typing time and can weaken the forecast, the metrics and the agents that read the database. The safer path replaces typing with automatic capture of email, calendar and meeting summaries, with few fields per stage and the seller reviewing the record.

How do I measure my team’s selling time?

Ask a sample of sellers to log the week in 30-minute blocks for two weeks, using the same categories as the surveys, and cross-check with CRM activities. Repeat after 90 days of change, along with conversion and meeting volume.

How do I prepare the team to work with fewer administrative tasks?

Train by role, in the company’s own CRM, showing what the system now records on its own and what the seller needs to review and approve. Pair the training with leadership’s decision on where the freed hours go, because sellers only change their routine when they know what is expected of them in the time left over.