Sales / account manager
Turns an enquiry into a deal: qualifying who is real, putting together a price, staying in front of the customer through a decision that takes weeks, and keeping them next time.
This is not a probability of losing your job. It combines how much of the role's task load is exposed to automation with how far adoption has actually gone — useful for comparing occupations on one consistent basis, and for nothing else.
B2B sales with named accounts, multi-touch follow-up and a quoted price — the role most SMEs staff most heavily. Retail floor selling and high-volume outbound calling are different jobs with different automation paths and are assessed separately.
What is actually changing#
The unit of analysis is the task, not the job title. A role is not replaced — its task mix shifts.
Working out who is worth your time
Automating✓ Evidence-backedReading the enquiry, the company and the person behind it, and deciding whether this is a real budget with a real date or someone collecting quotes.
Pulling together what is publicly known about a company and a contact, and drafting a first read on fit, is exactly the retrieval-and-summarise work that general models crossed the professional-usability threshold on. The judgement of whether to chase is not automated; the two hours of reading that used to precede it largely are.
Faster qualification is not more deals. It changes where the hours go, not whether the customer buys — and a rep who now qualifies three times as many leads is being measured on a bigger number with the same closing skill behind it.
Putting the price and the proposal together
Automating✓ Evidence-backedGathering supplier prices, assembling options, writing the proposal, and getting it back to the customer before they lose interest.
Reading quotes out of emails and attachments into structured fields, then assembling a document from them, is document extraction plus drafting — two capabilities that have been commercially usable for years and are now cheap enough to run on every enquiry rather than the big ones.
What automates is assembling the quote, not deciding the number. Where the margin is set by judgement about this customer and this competitor, the document gets faster and the decision does not move — and the decision is where the money is.
Chasing, on time, every time
Being augmented≈ Platform inferenceKnowing who has gone quiet, who to call today, what was promised last time, and getting back to them before the deal goes cold.
The system can rank who to contact and draft the message; a person still has to make the call and own what is said in it. This is the clearest augmenting case in the role — the work does not leave, it stops being forgotten.
Remembering is not persuading. A perfect follow-up list raises the floor for a disorganised rep and does almost nothing for a good one — so it compresses the gap between reps rather than replacing any of them, and that changes who gets hired, not how many.
The conversation where it is decided
Still human-led≈ Platform inferenceHearing the objection under the objection, knowing when to hold the price, when to walk, and being the person the customer calls when something goes wrong.
The binding constraint is not language, it is accountability: a concession has to be authorised by someone who can be held to it, and a customer who has been let down is buying reassurance from a person, not an answer. Every published rollback we carry in customer-facing automation has this shape.
Being the hardest part to automate does not make it the biggest part of the day. In most SME sales roles this is a few hours a week sitting on top of many hours of assembly and chasing — so a role can shrink substantially while its irreplaceable core stays exactly where it is.
Telling the boss what will land
Being augmented≈ Platform inferenceKeeping the pipeline honest, saying which deals will close this month, and being wrong in a way the company can plan around.
Pattern-matching past deals against the current pipeline is what these systems are good at, and it removes the rep's incentive to flatter the number. What it cannot supply is the private fact — the customer's budget moved, the champion left — which only reaches the model if the rep types it in.
A better forecast changes what management does, not what the customer does. It can also make a rep's month look worse without their work being worse, which is a management problem dressed up as a data problem.
Running the machine that now sells with you
New task≈ Platform inferenceChecking what the assistant drafted before it goes out, correcting the price it suggested, feeding back what actually closed, and catching the confident mistake before the customer sees it.
Once drafting and pricing suggestions are in the flow, someone has to own what goes out under the company's name. This is new work created by the automation — it did not exist in the role three years ago, and nobody is assigned to it by default.
New duties are not new headcount. This lands on the reps who are already there, usually without a title change and without coming off their number — which is how a role gets heavier while looking unchanged on the org chart.
Which technologies matter here#
Four separate signals. They are deliberately not added together — a job exposed to two technologies is not twice as exposed.
How it got here#
The index is not a static number. This is where it would have sat at each capability checkpoint since ChatGPT — reconstructed, and labelled as such.
● 1 verified event for this occupation, plotted at the date it happened — the parts of the line near a marker are anchored to something checkable.
Lead scoring, quote templates and follow-up reminders were in CRMs long before 2022, so the baseline is not low — sales was never un-automated, it was automated only down to the paperwork and never into the conversation. The two real steps are late 2023, when documents and tool calling let a model read quotes out of email attachments, and 2024, when the price fell far enough to run that on every enquiry rather than the large ones. It flattens at the end because what remains is gated by who may authorise a concession, and no model release moves that.
A flat line is not a forecast of safety. It says which tasks automation has reached so far — the occupations that moved least here are the ones where the constraint is physical or regulatory, and both of those can change.
Recent changes#
Wholesale and manufacturing sales representatives rank 6th by employment among the highest-exposure quintile in the paper's appendix. US private-sector payroll records from ADP, November 2022 to June 2026. The adjustment operates through reduced hiring of young workers rather than separations, and experienced workers show no comparable gap. The authors state they find no evidence of widespread, economy-wide displacement. Descriptive, not causal, and measured at occupation level rather than task level — the paper says nothing about which selling tasks moved.
Verifiable change in hiring, headcount, hours or job scope. Highest weight — but causal attribution still has to be argued, not assumed.
What this means for you#
The entry rung of this job was research and quote assembly — the two things that automate first. Expect to be hired to talk to people much earlier than reps a few years ago were, with less time spent earning that by doing the paperwork. That is a harder start, not an easier one: you get judged on the conversation before you have had two hundred of them.
What you know about this customer, this competitor and where the margin really is does not exist in any system, and that is now a larger share of what you are paid for. The risk is not being replaced; it is being asked to carry three times the pipeline on the same commission structure because the assembly work got cheap.
Your options#
Four directions, each with its real constraints and one thing you can test this week. Continuing as you are is a legitimate choice — it just has to be a chosen one.
Become the one who owns the price
Assembly automates; authority does not. The reps who stay valuable are the ones allowed to decide a margin and answer for it, not the ones who relay a number someone else set.
Pricing authority is given by managers who have been burned before, and it usually arrives after a year of being right in small amounts. It also makes your losses visible in a way a quote-relayer's never are.
Take the last five deals you quoted and write down, for each, what you would have priced it at if it were yours to decide and why. Show that page to your manager and ask where you are wrong. You will find out in one conversation whether the gap is judgement or permission.
Own the selling system, not just your accounts
Somebody has to decide what the assistant is allowed to send, which prices it may suggest, and what gets fed back when a deal closes. In most SMEs nobody owns this yet, and it is being done badly by default.
It is unglamorous, it competes with your own number, and it only becomes a real role if management agrees it is one. Doing it unpaid for a year and then asking is the common failure.
Collect ten messages or quotes your team's tools drafted this week and mark each one: sent as-is, edited, or would have embarrassed us. Bring the count, not the opinion. Three numbers make the case that a hundred sentences cannot.
Move to where the product is hard
The share of selling that survives easily is the part where the buyer cannot self-serve because the thing is complicated, regulated, or expensive to get wrong. Technical, industrial and regulated sales carry more of that per hour.
It usually means a pay cut while you learn the domain, and the first year you are worse than the person you replaced. Domain credibility takes longer to acquire than sales technique.
Find one technical salesperson in an industry you are curious about and ask them a single question: what does a customer ask you that you cannot answer without having worked here for two years? Their answer tells you what the barrier actually is.
Cross into revenue operations
Everything described above — pipeline hygiene, forecast, what the tools may do, what gets fed back — is becoming a function rather than a habit. Reps who understand both the selling and the system are the scarce input to it.
It pays in salary rather than commission, which is a cut in a good year and a rise in a bad one. And it is a support function: you will be measured on other people's numbers.
Rebuild last quarter's pipeline report from the raw records yourself, without using the dashboard. Whatever took you longest is the job — and whether you enjoyed that hour tells you more than any description of the role.
Common questions#
Not the conversation where the deal is decided — that needs someone who can authorise a concession and be held to it. But a large share of a sales day is research, quote assembly and chasing, and those are automating quickly. The realistic reading is fewer people carrying larger pipelines, not an empty sales floor.
Watch your own company, not the industry. The measurable signal is how many enquiries one rep is expected to carry this quarter against the same quarter last year. When that number rises while headcount does not, the change has already arrived in your building — and it usually shows up as a target increase before it shows up as a hiring freeze.
Qualification research and quote assembly, in that order. Both are reading-and-drafting work on information that already exists somewhere. Follow-up is being assisted rather than removed. Negotiation and being the person the customer calls when something breaks are the parts nothing on the market currently does.
Because speed is on the supply side of your own funnel, and the customer's decision is not. A faster quote wins deals where you were losing on response time and changes nothing where you were losing on price or trust. Check which of the three you actually lose to before reading a faster tool as a bigger cheque.
Under your name, only if you read it first — and the reading has to be funded. Where teams have let drafts go out unchecked, the failure is not bad grammar, it is a confident commitment nobody authorised: a date, a discount, an availability. That is the same failure mode as every customer-facing rollback in our evidence base.
Method and sources#
- Assessment date
- 2026-09-11
- Basis of the task judgements
- 2 evidence-backed · 4 platform inference · 0 not enough evidence
- Verified events
- 1