Industries
Restaurants & Food
Take restaurant orders, table bookings and payments in one WhatsApp chat. Auto-reply to common questions and cut support costs with ChatMitra.
Real Estate
Use WhatsApp automation to capture property leads, qualify them automatically, and follow up fast. See how real estate agents close more with ChatMitra.
Travel & Tourism
Use the WhatsApp Business API to capture travel leads, send itineraries, confirm bookings, and keep travellers updated. See how agencies grow with ChatMitra.
Banking & Finance
Run secure KYC, send account alerts and payment notifications on WhatsApp. Cut support load and stay within Meta's rules with ChatMitra.
Beauty & Cosmetics
Sell your catalog, recover carts and win repeat orders on WhatsApp. Auto-reply to product questions and send launches to opted-in customers with ChatMitra.
Education
Answer admission enquiries, fee and timetable questions, and keep parents updated — all on WhatsApp. Cut admin load and costs with ChatMitra.
EduTech
Capture course leads, qualify them and onboard learners on WhatsApp. Automate replies and cut lead-qualification cost with ChatMitra's no-code chatbot.
Events & Webinars
Handle event registrations, reminders and RSVPs in one WhatsApp chat. Cut no-shows, automate FAQs and keep attendees updated with ChatMitra.
Freelancers & Consultants
Capture leads, send proposals and chase invoices in one chat. WhatsApp automation for freelancers and consultants with ChatMitra — from ₹0.20 a chat.
Health & Wellness
Book appointments, send reminders and answer client questions in one WhatsApp chat. Auto-reply to routine queries and cut support costs with ChatMitra.
Home Decor
Share your home decor catalog, take orders and send delivery updates on WhatsApp. Automate FAQs and cut support costs with ChatMitra.
IT Services
Use the WhatsApp Business API to qualify leads, run support over chat, and wire WhatsApp into your stack with webhooks. How IT companies scale with ChatMitra.
Marketing Agencies
Use WhatsApp automation to capture leads, run campaigns and manage multiple client brands from one account. How agencies scale on WhatsApp with ChatMitra.
Offline & Retail
Turn shop footfall into a WhatsApp list. Share catalogs, add a QR-to-chat at the counter and bring buyers back — with ChatMitra from ₹0.20 per chat.
Spas & Salons
Use the WhatsApp Business API to take spa and salon bookings, cut no-shows with automated reminders, and rebook regulars. How salons grow with ChatMitra.
Automotive
Use the WhatsApp Business API for automotive dealerships to qualify buyers, book test drives and send service reminders — automated, on one chat, with ChatMitra.
All Industries
Browse every industry guide
Back to WorkflowMitra Blog
Operations, Finance & Support Workflows

Customer Success Automation for Small B2B Teams: Onboarding, Health Checks and Renewals

Nikunj Gohil
Nikunj Gohil
18 min read
On a shelf of four potted plants tagged with customer names, a customer success manager pours a thin stream of water onto the one drooping plant, tagged Kavya Logistics, under a phone card reading Amber to Red: Kavya Logistics, 8 of 20 seats active

Quick Summary

Customer success automation lets a small B2B team look after more accounts without a dedicated customer success platform. One Google Sheet holds every account and its signals: seats in use, last login, open tickets, NPS, last check-in and renewal date. Six workflows do the routine work: assigning each new customer to a CSM, scoring every account weekly with reasons, reminding CSMs of overdue check-ins, chasing detractors after an NPS survey, flagging renewals that aren't healthy, and drafting a monthly brief for at-risk accounts. The decisions and the conversations stay with people.

A software company in Bengaluru sells to about 40 business customers. Two account managers look after all of them, alongside sales calls and support escalations. Nobody tracks anything formally. The first sign that Kavya Logistics was unhappy came when its operations head asked for a cancellation quote, 50 days before renewal. Looking back, the signs had been there for weeks: only 8 of its 20 seats in use, four open tickets, an NPS of 6 that nobody replied to, and no call in six weeks.

That’s the usual shape of churn at a small B2B company. The information exists, but it’s scattered across the product, the helpdesk, a survey tool and people’s memories. Nobody joins it up until the account is already leaving.

This guide joins it up with one Google Sheet and six workflows in WorkflowMitra, powered by ChatMitra: handoff from sales, a weekly health score with reasons, check-in reminders, NPS follow-up, a renewal risk list and a monthly brief for accounts that need attention. It’s written for teams without a customer success platform, and it doesn’t cover chasing unpaid invoices. That’s finance’s job, and our guide to automating payment reminders covers it.

Product behaviour described here was checked in WorkflowMitra’s source code, and all four Code steps were run in its sandbox, on 20 October 2026.

What is customer success automation?

Customer success automation means letting workflows do the routine work of looking after existing customers: assigning new accounts, scoring account health from usage, support and survey signals, reminding the team about overdue check-ins, following up on low survey scores and flagging renewals at risk. People keep the conversations and the decisions; the workflows make sure nothing is forgotten.

For a small team, the value isn’t sophistication. It’s that every account gets looked at every week, by a rule that doesn’t get busy.

What a small team should automate, and what it shouldn’t

Customer success is a relationship job. Automating the wrong part makes customers feel processed. A useful dividing line:

AutomateKeep human
Assigning a new account to a CSM and sending the kickoff emailThe kickoff call and the onboarding plan for that customer
Collecting signals and scoring every account each weekDeciding what a falling score means for this customer
Telling each CSM which check-ins are overdueThe check-in itself
Alerting the right person about a low NPS scoreCalling the customer to find out why
Listing renewals that aren't healthy, months aheadThe renewal conversation and any commercial offer
Drafting a brief for at-risk accountsAnything sent to the customer

Everything in the left column produces an internal message to a named person. None of the six workflows below sends anything to a customer except the kickoff email, which a person writes once.

The six workflows

#WorkflowRunsOutput
1Handoff and kickoffWhen sales adds a won customerCSM assigned, account row created, kickoff email sent, CSM told
2Weekly health scoreMonday morningScore, band and reasons per account; history logged; changes posted to the team
3Check-in remindersEvery working morningOne email per CSM listing overdue and upcoming check-ins
4NPS follow-upWhen a survey response arrivesAccount updated; detractors routed to their CSM within minutes
5Renewal riskMonday morningRenewals in the next 90 days that aren't Green, soonest first
6Monthly account briefFirst working day of the monthAn AI-drafted internal brief per at-risk account, for the CSM

They all share one spreadsheet, so start there.

The account sheet

Create a Google Sheet with three tabs.

Accounts, one row per customer, with these columns: Account, CSM, CSM email, Tier, Contact name, Contact email, Seats, Active users, Last login, Open tickets, Last NPS, Last contact, Onboarding done, Start date, Renewal date, Champion left, Score, Band, Reasons.

  • Tier is High, Mid or Low: how much personal attention the account gets. It sets the check-in cadence: 30, 60 or 90 days in the examples.
  • Seats, Active users, Last login and Open tickets come from your product and helpdesk. If your product has a database, a nightly workflow can copy them in with a MySQL or Postgres step; otherwise paste a weekly export. Our guide to moving data from a database to Google Sheets shows the nightly version.
  • Last contact is the date of the last real conversation: a call, a meeting or a reply to an email the customer started. CSMs update it themselves. It’s the one manual column, and the check-in reminders depend on it.
  • Champion left is Yes when your main contact has moved on.
  • Score, Band and Reasons are written by workflow 2. Don’t type in them.

Keep dates as real dates in the sheet (format the columns as dates), and Account names unique: several workflows look accounts up by name.

Won deals is where sales adds a row for each new customer: Account, Contact name, Contact email, Seats, Tier, Start date, Renewal date.

Health log gets one row per account per week: Week, Account, Score, Band, Reasons. It’s what lets you check later whether the score predicted anything.

You’ll also need WorkflowMitra’s Google Sheets integration, a Slack incoming webhook for a #customer-success channel (or email, or a Telegram group), and a mailbox WorkflowMitra can send from.

Workflow 1: handoff and kickoff

The riskiest moment in a new account is the gap between “sales closed it” and “someone in CS owns it”. This workflow closes that gap the minute sales adds the row.

  1. Trigger: Google Sheets, row added, on Won deals, checking every few minutes. Ask sales to add rows at the bottom and never sort the tab.
  2. Assign to the next person, named Assign. Under Who should this go to?, list your CSMs with their email addresses. Each run picks the next person in the list and starts again at the top, so new accounts are shared evenly. The step outputs assignedTo and assignedName.
  3. Google Sheets, Add row(s), on Accounts: Account, Contact name, Contact email, Seats, Tier, Start date and Renewal date from {{ $trigger["Account"] }} and the other trigger fields; CSM {{ assignedName }}; CSM email {{ assignedTo }}; Onboarding done No; Last contact empty.
  4. Email to {{ $trigger["Contact email"] }}: a short welcome that names the CSM, says what happens in the first week and offers two times for a kickoff call. Write it once, as the CSM would.
  5. Slack to #customer-success: New account {{ $trigger["Account"] }} ({{ $trigger["Tier"] }} touch, {{ $trigger["Seats"] }} seats) is yours, {{ $node["Assign"].assignedName }}. Kickoff email sent to {{ $trigger["Contact name"] }}.

The Slack step reads the Assign step by name, because a plain {{ assignedName }} after the Add row(s) and Email steps would read the Email step’s output instead.

Round-robin isn’t always right. If one CSM looks after a vertical or a region, put an IF or Switch before the Assign step that routes those accounts to a fixed person. Note that every run moves the rotation on, test runs included, so a few test rows will skip people in the first real round.

The onboarding sequence itself, with reminders until the customer finishes setup and a handoff when they get stuck, is a workflow of its own. Our guide to customer onboarding automation builds it in detail. When onboarding is complete, set Onboarding done to Yes; the health score below takes points off accounts still onboarding after 30 days.

Workflow 2: a weekly health score with reasons

A health score is only useful if a CSM can read it and know what to do. “62” tells them nothing; “62: 12 of 20 seats active; last check-in 41 days ago” tells them to call and ask about adoption. So this score is built from rules you can see, and every point lost comes with a reason.

A customer success manager listens with a stethoscope to an account folder labelled Kavya Logistics, with tags fanning out for seats active, open tickets, NPS and last check-in, each with the points lost, ending in a badge reading 35, Red
Every point an account loses comes with a reason a CSM can act on.

The scoring rules

SignalRule in the examplePoints off
Seat useUnder 50% of seats active / 50–70%25 / 10
Last login14 days or more ago20
Open tickets3 or more10
Last NPS0–6 (detractor) / 7–8 (passive)20 / 5
Last check-inOlder than the tier's cadence, or never10
OnboardingNot done 30 days after the start date15
Champion leftMain contact has moved on20, and the score can't go above 70

Scores of 75 and above are Green, 50 to 74 Amber and below 50 Red. These numbers are a starting point, not a standard. The NPS bands follow Bain’s definition, where promoters score 9 or 10, passives 7 or 8 and detractors 0 to 6; everything else should be tuned to your customers.

Two design choices matter more than the weights:

  • A red flag caps the score. In testing, an account whose main contact had left, renewing in 20 days, still scored 75 under plain addition, because everything else looked fine. Losing your champion before a renewal is never “Green”, so that rule limits the score to 70 on its own.
  • Missing data isn’t counted as good or bad. A blank NPS or ticket count adds “no data: NPS” to the reasons instead of taking points off. An account with no signals at all shouldn’t look perfectly healthy, and it shouldn’t look broken either; it should look unknown, so someone fills the gap.

Build it

  1. Trigger: Schedule, Mondays at 9 a.m. (cron 0 9 * * 1), time zone Asia/Kolkata.
  2. Google Sheets, Read rows, on Accounts. Set Read values to Raw values (numbers stay numbers), so dates arrive as sheet serial numbers that the Code step can subtract. With the default, “As displayed”, they arrive as text in your sheet’s date format.
  3. Code step named Account review:
// Points taken off a score of 100. Tune these to your own churned accounts.
var CONTACT_DAYS = { High: 30, Mid: 60, Low: 90 }; // check-in cadence by tier

var today = Math.floor((Date.now() + 5.5 * 3600000) / 86400000) + 25569; // sheet serial, India time
function num(v) { var n = Number(v); return v === "" || v === null || v === undefined || isNaN(n) ? null : n; }
function yes(v) { return String(v || "").trim().toLowerCase() === "yes"; }

var changed = [], log = [], expansion = [];
var week = new Date((today - 25569) * 86400000).toISOString().slice(0, 10);

(input.rows || []).forEach(function (r) {
  if (!String(r["Account"] || "").trim()) return;
  var score = 100, why = [], gaps = [];
  var seats = num(r["Seats"]), active = num(r["Active users"]);
  var login = num(r["Last login"]), contact = num(r["Last contact"]);
  var tickets = num(r["Open tickets"]), nps = num(r["Last NPS"]);
  var start = num(r["Start date"]);

  if (seats && active !== null) {
    var use = active / seats;
    if (use < 0.5) { score -= 25; why.push("only " + active + " of " + seats + " seats active"); }
    else if (use < 0.7) { score -= 10; why.push(active + " of " + seats + " seats active"); }
    if (use >= 0.9) expansion.push(r["Account"] + " (" + active + " of " + seats + " seats)");
  } else gaps.push("usage");

  if (login !== null) {
    if (today - login >= 14) { score -= 20; why.push("no login for " + (today - login) + " days"); }
  } else gaps.push("last login");

  if (tickets !== null) {
    if (tickets >= 3) { score -= 10; why.push(tickets + " open tickets"); }
  } else gaps.push("tickets");

  if (nps !== null) {
    if (nps <= 6) { score -= 20; why.push("NPS " + nps + " (detractor)"); }
    else if (nps <= 8) { score -= 5; why.push("NPS " + nps + " (passive)"); }
  } else gaps.push("NPS");

  var cadence = CONTACT_DAYS[String(r["Tier"] || "").trim()] || 60;
  if (contact === null || today - contact > cadence) {
    score -= 10; why.push(contact === null ? "no check-in recorded" : "last check-in " + (today - contact) + " days ago");
  }

  if (!yes(r["Onboarding done"]) && start !== null && today - start > 30) {
    score -= 15; why.push("onboarding not finished after " + (today - start) + " days");
  }
  // A red flag caps the score: losing your main contact is never "Green".
  if (yes(r["Champion left"])) { score = Math.min(score - 20, 70); why.push("main contact has left"); }

  score = Math.max(0, score);
  var band = score >= 75 ? "Green" : score >= 50 ? "Amber" : "Red";
  var reasons = why.join("; ") + (gaps.length ? (why.length ? "; " : "") + "no data: " + gaps.join(", ") : "");
  var previous = String(r["Band"] || "").trim();

  log.push({ week: week, account: r["Account"], score: score, band: band, reasons: reasons });
  if (band !== previous) {
    changed.push({ account: r["Account"], score: score, band: band, previous: previous || "new",
      reasons: reasons, csm: r["CSM"] || "" });
  }
});

var order = { Red: 0, Amber: 1, Green: 2 };
changed.sort(function (a, b) { return order[a.band] - order[b.band]; });

return {
  week: week,
  log: log,
  changed: changed,
  changedCount: changed.length,
  expansionCount: expansion.length,
  digest: "Account health, week of " + week + "\n" +
    (changed.length ? changed.map(function (c) {
      return c.previous + " → " + c.band + ": " + c.account + " (" + c.score + ", " + c.csm + ") " + c.reasons;
    }).join("\n") : "No account changed band this week.") +
    (expansion.length ? "\nNearly full, talk about more seats: " + expansion.join(", ") : "")
};
  1. Loop, with List field log. Inside the loop, add a Google Sheets Update row on Accounts: Key column Account, Key value {{ $trigger.account }}, and set Score to {{ $trigger.score }}, Band to {{ $trigger.band }} and Reasons to {{ $trigger.reasons }}. Inside a loop, $trigger is the current item, not the workflow’s trigger.
  2. On the loop’s done branch, a Google Sheets Add row(s) on Health log in bulk mode: Upstream Array Source {{ $node["Account review"].log }}, mapping week, account, score, band and reasons to the five columns.
  3. Slack to #customer-success with {{ $node["Account review"].digest }}.

The step compares each account’s new band with last week’s, still in the Band column when the sheet is read, and reports only the changes, worst first. Against four test accounts, it produced:

Account health, week of 2026-10-20
Amber → Red: Kavya Logistics (35, Neha) only 8 of 20 seats active; 4 open tickets; NPS 6 (detractor); last check-in 41 days ago
new → Red: Arogya Clinics (45, Rahul) 3 of 5 seats active; no login for 16 days; no check-in recorded; onboarding not finished after 45 days; no data: tickets, NPS
Green → Amber: Metro Foods (70, Rahul) NPS 8 (passive); main contact has left
Nearly full, talk about more seats: Shree Textiles (10 of 10 seats)

The last line is the other side of health: accounts using 90% or more of their seats are expansion conversations, and the account owner should hear about them as early as the risks.

Update row stops with an error if no row matches the account name, and by default also if two rows match. Both are signs that the Accounts tab has a typo or a duplicate, and both are worth fixing rather than skipping.

Workflow 3: check-in reminders for every CSM

Workflow 2 tells the team which accounts look unwell. This one makes sure every account gets a conversation on schedule, healthy or not.

  1. Trigger: Schedule, every working day at 9:30 a.m. (cron 30 9 * * 1-5), Asia/Kolkata.
  2. Google Sheets, Read rows, on Accounts, with Raw values.
  3. Code step named Check-in gaps:
var CONTACT_DAYS = { High: 30, Mid: 60, Low: 90 }; // same cadence as the health check
var SOON = 7;                                        // warn this many days before a check-in is due

var today = Math.floor((Date.now() + 5.5 * 3600000) / 86400000) + 25569;
var byCsm = {};

(input.rows || []).forEach(function (r) {
  var account = String(r["Account"] || "").trim();
  var email = String(r["CSM email"] || "").trim().toLowerCase();
  if (!account || !email) return;
  var cadence = CONTACT_DAYS[String(r["Tier"] || "").trim()] || 60;
  var last = Number(r["Last contact"]);
  var left = r["Last contact"] === "" || isNaN(last) ? -1 : cadence - (today - last);
  if (left > SOON) return;
  var g = byCsm[email] || (byCsm[email] = { email: email, name: r["CSM"] || email, overdue: [], soon: [] });
  if (left < 0) g.overdue.push(account + (isNaN(last) || r["Last contact"] === "" ? " (never)" : " (" + (today - last) + " days)"));
  else g.soon.push(account + " (due in " + left + " days)");
});

return Object.keys(byCsm).map(function (k) {
  var g = byCsm[k];
  return {
    email: g.email,
    subject: g.overdue.length + " overdue, " + g.soon.length + " due soon: customer check-ins",
    body: "Hi " + g.name + ",\n\nOverdue:\n" + (g.overdue.join("\n") || "none") +
      "\n\nDue in the next " + SOON + " days:\n" + (g.soon.join("\n") || "none") +
      "\n\nLog each call or meeting in the Last contact column so this list stays right."
  };
});
  1. Loop, leaving List field blank so it loops over the step’s list directly. Inside it, an Email step to {{ $trigger.email }} with subject {{ $trigger.subject }} and body {{ $trigger.body }}.

Each CSM gets one email, and only on days when something is overdue or due within a week. In the test data, Rahul’s email listed Arogya Clinics as overdue (“never”) and Metro Foods as due in 5 days. The step lower-cases email addresses before grouping, so Rahul@ and rahul@ don’t produce two emails.

The reminder is only as good as the Last contact column. Make updating it part of every call; the closing line of the email is there to remind people.

Workflow 4: follow up every low NPS score

A detractor who hears nothing back after a survey has just been told their opinion doesn’t matter. This workflow makes sure the right person hears about every low score within minutes.

  1. Survey: a Google Form with a dropdown for the company (so names match the Accounts tab exactly), the 0–10 question “How likely are you to recommend us to a friend or colleague?”, and an optional comment. Link it to a sheet, and send it to your contacts from your usual email tool, quarterly or after onboarding.
  2. Trigger: Google Sheets, row added, on the form’s response tab. Choose No for Use column headers as IDs of the columns.
  3. IF the score field Is not empty, with Else unconnected. A respondent who skipped the question stops here, so a blank never overwrites the account’s previous score.
  4. Google Sheets, Update row, on Accounts: Key column Account, Key value {{ $trigger["Company"] }}, Last NPS {{ $trigger["How likely are you to recommend us to a friend or colleague?"] }}. Every score, good or bad, now feeds next Monday’s health check.
  5. IF the score Less than or equal 6, again with Else unconnected. The order matters: in WorkflowMitra a blank value passes “Less than or equal”, so without the empty check in step 3, an unanswered question would count as a detractor.
  6. Google Sheets, Find row, named Find account, on Accounts, key column Account and key value {{ $trigger["Company"] }}, to get the account’s CSM.
  7. Email to {{ $node["Find account"].row["CSM email"] }}: {{ $trigger["Company"] }} scored us {{ $trigger["How likely are you to recommend us to a friend or colleague?"] }}. Comment: {{ $trigger["Anything else you'd like to tell us?"] }}. Please call within two working days and note what you learn in Last contact. Copy the question titles exactly from your sheet’s header row.

If you collect many free-text comments, an AI step can sort them by theme before they reach the CSM. Our guide to AI sentiment analysis automation builds that with your own AI provider key. For a handful of responses a quarter, read them yourself.

Bain’s Net Promoter definition counts promoters as 9 to 10 and detractors as 0 to 6, and the score is the percentage of promoters minus the percentage of detractors. Track it per quarter on the form’s response tab, but act on each detractor individually: the follow-up is worth more than the number.

Workflow 5: renewals that aren’t healthy

Renewal reminders tell customers a renewal is coming. Renewal risk tells you which renewals need work first. Run it weekly, after the health score.

  1. Trigger: Schedule, Mondays at 9:30 a.m. (cron 30 9 * * 1), Asia/Kolkata, so it reads this week’s bands.
  2. Google Sheets, Read rows, on Accounts, with Raw values.
  3. Code step named Renewal risk:
var WINDOW = 90; // days before renewal to start watching

var today = Math.floor((Date.now() + 5.5 * 3600000) / 86400000) + 25569;
var risky = [], fine = 0;

(input.rows || []).forEach(function (r) {
  var due = Number(r["Renewal date"]);
  if (r["Renewal date"] === "" || isNaN(due)) return;
  var days = due - today;
  if (days < 0 || days > WINDOW) return;
  var band = String(r["Band"] || "").trim();
  if (band === "Green") { fine++; return; }
  risky.push({
    days: days,
    line: r["Account"] + ": renews in " + days + " days, " + (band || "no band yet") +
      " (" + (r["Reasons"] || "no reasons recorded") + "), CSM " + (r["CSM"] || "unassigned")
  });
});

risky.sort(function (a, b) { return a.days - b.days; });
return {
  riskyCount: risky.length,
  message: "Renewals in the next " + WINDOW + " days that aren't Green:\n" +
    (risky.map(function (x) { return x.line; }).join("\n") || "none") +
    "\n(" + fine + " Green renewals in the same window are left to the normal reminders.)"
};
  1. IF {{ riskyCount }} Greater than 0, then Slack to the founder or CS lead with {{ message }}.

In the test data, it listed Metro Foods (renews in 20 days, Amber, main contact left) above Kavya Logistics (52 days, Red). Soonest comes first, because a Red account renewing in four months still has time for an adoption plan and an Amber one renewing in three weeks doesn’t.

Ninety days is about the shortest window in which a struggling account can be turned around; for annual contracts with a notice period, widen it. The customer-facing reminders, with a renewal quote approved before it’s sent, are covered in our guide to contract renewal reminders. The two workflows read the same Renewal date column.

Sand in a large hourglass piles up around a small office building signed Metro Foods, while a CS lead points at a board of renewals in the next 90 days that are not Green, with Metro Foods at 20 days above Kavya Logistics at 52 days
The renewals that need work show up months ahead, soonest first.

Workflow 6: a monthly brief for at-risk accounts

Before a review call with a struggling account, a CSM pieces together the same facts every time. An AI step can do the first draft, from the sheet only, using your own AI provider key.

  1. Trigger: Schedule, on the 1st of the month at 9 a.m. (cron 0 9 1 * *), Asia/Kolkata.
  2. Google Sheets, Read rows, on Accounts.
  3. Code step named At-risk accounts:
// Keep the accounts that need attention, and write down only facts from the sheet.
return (input.rows || []).filter(function (r) {
  return String(r["Account"] || "").trim() && String(r["CSM email"] || "").trim() &&
    String(r["Band"] || "").trim() !== "Green";
}).map(function (r) {
  return {
    email: String(r["CSM email"]).trim().toLowerCase(),
    account: r["Account"],
    facts: [
      "Account: " + r["Account"] + " (" + (r["Tier"] || "no tier") + " touch)",
      "Health: " + (r["Band"] || "not scored") + ", score " + r["Score"],
      "Reasons: " + (r["Reasons"] || "none recorded"),
      "Seats: " + r["Active users"] + " active of " + r["Seats"],
      "Open tickets: " + (r["Open tickets"] === "" ? "unknown" : r["Open tickets"]) + ", last NPS: " + (r["Last NPS"] === "" ? "none" : r["Last NPS"])
    ].join("\n")
  };
});
  1. Loop over the list (List field blank). Inside it:
    • AI step: pick your provider and model, set Max tokens to about 400, and put the rules in System prompt (optional): You prepare internal notes for a customer success manager. Use only the facts given. Write three short sections: what's going wrong, two questions to ask the customer, and one thing to offer. Don't invent numbers, names or events. Set the Prompt to {{ $trigger.facts }}.
    • Email to {{ $trigger.email }}, subject Account brief: {{ $trigger.account }}, body {{ text }}, which is the AI step’s reply.

The brief goes only to the CSM, never to the customer. The AI is told to stay within the facts, but it can still phrase things badly, so treat it as a draft for someone who knows the account. Two behaviours to know: a reply cut off by the Max tokens limit still counts as a successful step, so keep the limit comfortable; and if the provider refuses a request, the step fails and the run stops for that account.

What to do when an account turns Red

An alert with no agreed response is just noise. Agree on a short playbook with the team before the first Monday digest goes out, and tie each move to the reason the score gives:

Reason in the digestFirst move, within two working days
Few seats activeAsk the main contact who isn't using it and why; offer a 30-minute session for that team
No logins for two weeksCheck for a change on their side (new manager, project paused) before assuming dissatisfaction
Several open ticketsRead them, chase the oldest with support, and tell the customer who owns each one
Detractor scoreCall, listen, and write what you learned in the account's notes
Onboarding stalledFind the step they're stuck on and offer to do it with them
Main contact has leftFind the new owner, introduce yourself, and re-run a short kickoff

Whatever the move, it ends with Last contact updated. That’s how the account’s score recovers the following Monday, and how the check-in reminder stops. If an account stays Red for three weeks running, raise it with whoever owns the commercial relationship; by then it’s a renewal risk, not just a support one.

Test before Monday

  1. Copy the spreadsheet and point every workflow at the copy.
  2. Add five test accounts that cover the cases: one healthy, one with few seats active, one with no data at all, one whose champion has left, and one renewing in three weeks.
  3. Run workflow 2 by hand. Check that Score, Band and Reasons are written to every row, that five rows appear in Health log, and that the Slack digest lists only the accounts whose band changed. Run it again: the second digest should say nothing changed.
  4. Run workflow 3 and check that each test CSM gets one email, and that a CSM with nothing due gets none.
  5. Submit an NPS response of 4 and one with the score left blank. Only the first should update Last NPS and reach the CSM; the blank one should stop at the first IF.
  6. Run workflow 5 and check that the account renewing in three weeks is at the top.

The editor’s Run button performs a live run, so these tests really write to the sheet and really send emails and Slack messages. Use test addresses and a private channel, then switch the steps to the real ones.

The numbers to watch

Automation produces activity. These measures tell you whether it’s working. Most can be calculated from the Accounts and Health log tabs plus your billing records.

MeasureHow to calculate itWhat it tells you
Gross revenue retentionRecurring revenue from customers at the start of a period, minus what they cancelled or downgraded, divided by the starting revenueHow much revenue you keep, ignoring growth
Net revenue retentionAs above, but adding upgrades and extra seats from the same customersWhether existing customers grow faster than they shrink
Logo churnCustomers lost in a period divided by customers at its startHow many relationships end, whatever their size
Onboarding completionShare of accounts with Onboarding done = Yes within 30 days of the start dateWhether new customers reach first value
Check-in coverageShare of accounts whose last contact is within their tier's cadenceWhether the team is keeping up
Health distributionCount of Green, Amber and Red in this week's Health logThe direction of the whole book
Score accuracyOf the accounts that churned, how many were Red or Amber 60–90 days earlierWhether your weights are right
NPSPercentage of promoters (9–10) minus percentage of detractors (0–6)Loyalty trend, quarter to quarter

The last but one is the most useful and the least tracked. Once a quarter, filter the Health log for every account that churned or downgraded, and look at its band two to three months earlier. If most were Green, your score is measuring the wrong things; change the weights in the Code step and keep the log, so you can compare.

What it costs

At the rates checked on 20 October 2026, WorkflowMitra charges 1 credit for each trigger, Google Sheets step, Assign, Email, Slack and AI step; 1 credit per item a Loop processes; nothing for IF steps; and 2 thousandths of a credit per millisecond of Code, a tiny fraction here. For a team with 40 accounts:

WorkflowHow oftenCredits a runCredits a month
1. Handoff and kickoff4 new customers520
2. Weekly health score4–5 MondaysAbout 84 (40 loop items and 40 updates)About 360
3. Check-in reminders22 working daysAbout 6About 130
4. NPS follow-upAbout 15 responses2, or 4 for a detractorAbout 40
5. Renewal risk4–5 Mondays2, or 3 when it postsAbout 13
6. Monthly briefOnce, about 10 accountsAbout 32About 32

That’s roughly 600 credits a month, a small part of the Starter plan’s 10,000 credits for ₹999 a month. The health score is the biggest item because it updates every row; if your book grows to several hundred accounts, the Pro plan’s 20,000 credits for ₹1,499 a month covers it. AI calls are billed separately by your AI provider, because WorkflowMitra uses your own key. CSMs who only receive emails and Slack messages don’t need WorkflowMitra seats; only the people who build and change the workflows do (Starter includes 2 seats, Pro 5).

Using WhatsApp for customer check-ins

Many Indian B2B customers would rather hear from you on WhatsApp than by email, and WorkflowMitra’s WhatsApp step can send through ChatMitra or Meta’s Cloud API. Keep the rules in mind: customers must have opted in, and outside the 24-hour customer service window WhatsApp only delivers approved templates, which Meta charges for per delivered message, by category. A personal check-in is usually better sent by the CSM from their own WhatsApp Business inbox than by a workflow. Automated WhatsApp messages suit operational updates, such as “your onboarding session is confirmed for Thursday at 3 p.m.”. ChatMitra’s guide to WhatsApp utility templates explains which messages qualify, and current rates are on ChatMitra’s pricing page.

Common mistakes

  1. Scoring without reasons. A number with no explanation gets ignored by week three.
  2. Treating missing data as healthy. An account with no usage data isn’t Green; it’s unknown.
  3. Adding points without caps. One serious red flag can hide behind several good signals.
  4. Never checking the score against churn. The weights in the example are a guess until your own churned accounts confirm or correct them.
  5. Reading dates “As displayed”. The Code step needs raw serial numbers to subtract.
  6. Free-text company names on the survey. Update row and Find row need exact matches; use a dropdown.
  7. Forgetting the blank-score rule. Without “Is not empty” first, an unanswered NPS question looks like a detractor and wipes the last real score.
  8. Automating the conversation. Workflows should tell a person to call, not replace the call.
  9. Letting Last contact go stale. If CSMs don’t update it, the check-in reminders become noise and get ignored.
  10. Mixing in payment chasing. Overdue invoices are a finance workflow with its own tone and rules; keep them out of the health score’s messages.

Where WorkflowMitra fits

WorkflowMitra, powered by ChatMitra, is the layer that turns a sheet of accounts into a routine: it assigns new customers, scores every account with reasons, reminds each CSM what’s overdue, routes detractors, lists risky renewals and drafts briefs, with every run recorded in its execution history.

It isn’t a customer success platform. It has no built-in health score, no product-usage tracking, no customer portal and no CS dashboards; the score here is a Code step you own and can read. If you need in-app usage analytics, automated playbooks across a large team or detailed retention reporting, a dedicated CS platform does that. For a team of one to five people with tens or a few hundred accounts, a sheet and six workflows usually cover the work, and they move with you: the same signals can feed a platform later.

Start with the Monday score

Build the Accounts tab and workflow 2 first, and run it by hand on this week’s data. Read the reasons for every Red and Amber account with your team, and adjust the weights until the list matches your instinct about which customers are struggling. Then add the check-in reminders, and the rest once the sheet is trusted.

You can build all six on a 14-day WorkflowMitra trial with no card required, using the Google Sheets and Slack integrations. For the reporting side of the same data, our guide to KPI alerts automation shows how to send a message when a number crosses a line.

The Net Promoter Score question, scale, categories and calculation come from Bain & Company's Net Promoter System page; the WhatsApp opt-in, customer service window, template and per-message pricing rules from Meta's WhatsApp Business Platform pricing documentation. All were read on 20 October 2026. The retention measures are standard definitions written out in plain terms, not taken from one source. WorkflowMitra behaviour (the Google Sheets trigger and its row tracking, Read rows and its Read values options, Add row(s) in single and bulk mode, Find row and Update row matching and errors, Assign to the next person and its outputs and rotation, Loop and the current item in `$trigger`, the AI step's fields and output, the Email and Slack steps, the IF operators and blank values passing "Less than or equal", the previous-step rule for plain fields, date helpers using the server clock, plan seats and the credit rates per step) was checked in the product's source code, and all four Code steps were run in its sandbox against test accounts on 20 October 2026; plan prices come from WorkflowMitra's live plans the same day. The company, accounts, CSMs, scores and volumes in the examples are illustrations, not customer data or benchmarks. Sources: Bain & Company, measuring your Net Promoter Score; Meta, WhatsApp pricing.

FAQ

Frequently Asked Questions

Questions small B2B teams ask before automating customer success.

It's using workflows to handle the routine parts of looking after existing customers: assigning new accounts, watching usage and support signals, reminding the team about check-ins, following up on survey scores and flagging renewals at risk. The aim is that people spend their time on conversations, not on checking spreadsheets.

WorkflowMitra, powered by ChatMitra

Automate the work between your apps

Build workflows with triggers, conditions and actions, then test them before they go live. Try WorkflowMitra free for 14 days, no card required.