Automate the admin in your sales process: CRM updates after calls, follow-ups on open deals, proposals, order processing and handoff to delivery.
You automate a sales process by mapping every step a deal goes through after the first conversation, marking the steps that are admin rather than selling, and handing those to software with a person approving anything a customer sees. In most teams the admin is the same five jobs: logging the call in the CRM, writing the proposal, following up on open deals, processing the order, and handing the customer to delivery. Automating those gives reps back hours each week and stops deals from stalling between steps. This guide gives you a framework to find and automate that admin, from first meeting to signed order. Prospecting and outreach are out of scope.
What sales process automation means
Sales process automation is the use of software to run the repetitive, rule-based steps of moving a deal from first meeting to signed order, so salespeople spend their time on conversations and judgment rather than data entry and document assembly.
It is not the same as sales engagement or lead generation tools. Those find and contact new people. Sales process automation starts once a real opportunity exists.
Before you start
- Your sales stages as they are in the CRM (for example: discovery, demo, proposal, negotiation, closed).
- One or two reps willing to time their admin for a week.
- Your rules for discounts, approvals and who signs what.
How to automate a sales process: step by step
The framework: map the stages, find the admin in each, score it, automate the top steps with approval, then measure deal speed.
Step 1: Map the steps between first meeting and signed order
List every action that happens on a typical deal after the first call. Keep it concrete:
Stage | Typical actions |
|---|---|
After discovery call | Write notes, update CRM fields, send recap, book next meeting |
Demo | Prepare tailored deck, send recording and resources |
Proposal | Build quote, write proposal, get discount approval, send |
Negotiation | Track redlines, answer security questionnaire, chase signatures |
Closed won | Create order, set up billing, hand over to onboarding or delivery |
You should see: 20-40 actions per deal. Most teams are surprised how many are copy-paste.
Step 2: Separate selling from admin
Mark each action as selling (needs a person’s judgment or relationship) or admin (rule-based, repeatable). Asking the buyer about budget is selling; typing their answer into three CRM fields is admin. Negotiating the discount is selling; building the quote after it is agreed is admin.
Step 3: Score the admin steps
For each admin step, note how often it happens, how long it takes, and what goes wrong when it is late or wrong. Prioritise by hours per month multiplied by the cost of delay.
Steps that almost always score highest:
- CRM update after calls. Every call, 10-15 minutes each, and the source of bad forecasts.
- Follow-up on open deals. The most common reason deals stall.
- Proposal and quote assembly. Long, error-prone, and directly on the path to revenue.
- Order processing and handoff to delivery. Where customers first notice internal friction.
Step 4: Automate CRM updates after calls
Feed call notes or transcripts (from Gong, Fireflies, Zoom or Teams) to the automation and have it update deal stage, next step, close date, stakeholders and risks in HubSpot or Salesforce, then post a two-line recap in the deal channel. Keep a person reviewing changes to amount and close date.
Step 5: Automate follow-up on open deals
Every morning, list deals with no activity in seven days, and draft a short, specific follow-up for each based on the last conversation. The rep approves or edits and sends. For the routine right after a meeting, see same-day follow-up.
Common mistake: generic “just checking in” emails. A useful follow-up references the buyer’s own words and a concrete next step.
Step 6: Automate proposals and quotes
Generate the quote from CRM deal fields (products, seats, term, agreed discount) in HubSpot Quotes, Salesforce CPQ or PandaDoc. Draft the proposal from your template, filled with the buyer’s goals from discovery notes. Route discounts above your threshold to the approver automatically. A person reviews before it goes out.
Step 7: Automate order processing and handoff
When the deal is marked won and the order form is signed:
- Attach the signed document to the deal.
- Create the onboarding project in Asana, Linear or Notion from a template.
- Post a handoff summary to the delivery team: what was sold, promises made, key contacts, start date.
- Trigger billing setup. The steps from here to cash are covered in revenue automation.
Step 8: Measure deal speed
Track before and after:
Metric | What good looks like |
|---|---|
Rep hours on admin per week | Down by half or more |
Days between stages | Shorter, especially proposal to signature |
Deals with no next step | Near zero |
Time from signature to kickoff | Under three business days |
Which tools automate the sales process
Tool | Best for | Limit |
|---|---|---|
AI employee (Viktor) | All admin steps across CRM, call recorder, quoting, docs and project tools, briefed in plain language in Slack or Microsoft Teams, with approval before customer-facing actions | Needs clear approval rules |
CRM workflows (HubSpot, Salesforce) | Stage-based tasks and field updates inside the CRM | Limited outside the CRM |
Conversation intelligence (Gong, Fireflies) | Call notes and deal risk signals | Do not build proposals or handoffs |
Proposal and CPQ (PandaDoc, Salesforce CPQ) | Quotes and e-signature | Start only once deal data is right |
Workflow automation (Zapier, Make, n8n) | Single triggers such as “deal won, create project” | Each flow is built by hand |
An AI employee ties these together. Viktor connects to 9,800+ tools including HubSpot, Salesforce, Gong, PandaDoc and Asana. A rep can say in Slack: “Update the Acme deal from today’s call, draft the proposal with the 3-year option, and remind me Thursday if they haven’t replied.” He does the steps and asks before anything is sent. Read what is an AI employee for how it works, or AI for sales for the wider sales picture.
Troubleshooting
- Reps don’t trust the CRM updates. Start with suggestions the rep approves; switch to direct updates after two clean weeks.
- Proposals come out generic. Discovery notes are thin. Add a short discovery template the automation fills in.
- Handoffs still miss details. Add a required “promises made” field to the won stage.
FAQ
What is sales process automation?
Sales process automation is software running the repetitive steps of a deal after the first meeting, such as CRM updates, follow-ups, proposals, order processing and handoff to delivery, so reps focus on selling.
Which sales tasks should be automated first?
CRM updates after calls and follow-up on open deals. They happen on every deal, take the most time and cause most stalled pipelines.
Will automating the sales process make it feel impersonal?
Not if a person approves customer-facing messages and the drafts use the buyer’s own context. Automation removes typing, not the relationship.
What is the difference between sales process automation and revenue automation?
Sales process automation covers the deal until it is signed. Revenue automation covers what follows: billing setup, invoicing, collections and renewals.
Do I need a RevOps team to automate sales?
No, an AI employee like Viktor is briefed in plain language and works with your existing CRM and tools without code.
Give your reps their selling time back
Viktor works in Slack and Microsoft Teams, connects to your CRM and 9,800+ other tools, and asks before anything reaches a buyer. Start free and hand him this week’s stalled deals.
Ask ChatGPT
Ask Claude