Sync WhatsApp Conversation Data to Google Sheets: Leads, Response Times and Outcomes

If your sales team works out of WhatsApp and you report in Google Sheets, the honest answer to "which platform syncs the data" depends entirely on which WhatsApp you are using — and for one of the three, the answer is that no platform can.

• The WhatsApp Business app on a phone? Nothing can sync it. No API exists. Manual per-chat text export only.
• A shared-inbox platform (respond.io, Wati, Zoko, Trengo and similar)? Most offer a Sheets integration or a REST API you can pull on a schedule.
• The WhatsApp Business Platform (Cloud API)? Webhooks push every message to a URL you control. Point that at an Apps Script web app and your sheet fills itself, with no platform in the middle.

There is a second thing worth knowing before you pick: the metrics you want are not fields anyone exports. "Leads handled", "first response time" and "outcome" are things you derive from message traffic. That derivation is the actual work, and it is the same regardless of which route brings the messages in.

The three routes

	Business app (phone)	Inbox platform	Cloud API direct
Automatic sync	No	Usually	Yes
Setup effort	—	Minutes	A day
Monthly cost	Free	Per seat, typically $20–$60	Meta's message rates only
Data you get	Manual text export	Whatever they expose	Everything, raw
Custom metrics	No	Their fields	Anything you can compute
Who holds the data	The phone	The vendor	You

The middle route is the right call if you already pay for a shared inbox — check for a Sheets integration first, and an API second. Pull it on a schedule with UrlFetchApp exactly as you would any other API.

The rest of this covers route three, because it is the one that gives you the metrics nobody else's export has.

What lands in the sheet

With the Cloud API, Meta POSTs a webhook every time a message arrives or a status changes. Deploy an Apps Script web app (Deploy → New deployment → Web app, access Anyone) and register its /exec URL in your Meta app's webhook settings.

Meta verifies the endpoint first with a GET, so you need both handlers:

const VERIFY_TOKEN = 'pick-a-long-random-string';

function doGet(e) {
  // Meta's one-time subscription handshake
  if (e.parameter['hub.verify_token'] === VERIFY_TOKEN) {
    return ContentService.createTextOutput(e.parameter['hub.challenge']);
  }
  return ContentService.createTextOutput('forbidden');
}

function doPost(e) {
  const body = JSON.parse(e.postData.contents);
  const sheet = SpreadsheetApp.getActive().getSheetByName('Messages');
  const rows = [];

  (body.entry || []).forEach(entry => {
    (entry.changes || []).forEach(change => {
      const v = change.value || {};
      const agentNumber = (v.metadata || {}).display_phone_number || '';

      // Inbound: a customer wrote to you
      (v.messages || []).forEach(m => {
        rows.push([
          new Date(Number(m.timestamp) * 1000),
          'inbound',
          m.from,                              // customer's number
          agentNumber,
          m.type,
          (m.text || {}).body || '',
          m.id,
        ]);
      });

      // Outbound: delivery/read status of messages you sent
      (v.statuses || []).forEach(s => {
        rows.push([
          new Date(Number(s.timestamp) * 1000),
          'status:' + s.status,                // sent | delivered | read | failed
          s.recipient_id,
          agentNumber,
          '',
          '',
          s.id,
        ]);
      });
    });
  });

  if (rows.length) {
    sheet.getRange(sheet.getLastRow() + 1, 1, rows.length, 7).setValues(rows);
  }
  return ContentService.createTextOutput('EVENT_RECEIVED');
}

Two notes that save an afternoon. WhatsApp timestamps are Unix seconds, not milliseconds — multiply by 1000 or every row lands in 1970. And webhooks arrive batched, so handle arrays and write once with setValues; a busy hour with appendRow in a loop will time out. The general patterns are in our Apps Script webhooks guide.

Meta retries failed webhook deliveries, so return quickly. Do the parsing here and the analysis on a trigger — never make Meta wait while you compute a dashboard.

Turning messages into the three metrics

Raw messages are not a report. Run a second script on a trigger that rolls the Messages tab into a Conversations tab — one row per customer per conversation — and compute:

Leads handled. Count distinct customer numbers with at least one inbound message in the period. Decide once whether a returning customer is a new lead or the same one; both are defensible, inconsistency is not.

First response time. The gap between the first inbound message opening a conversation and the first human outbound reply. Two exclusions are essential:

• Do not count automated replies or template messages. An auto-acknowledgement is not a response. Count it and your median drops to four seconds and tells you nothing.
• Subtract time outside business hours. Otherwise one enquiry at 11pm ruins the week's average and your team is measured on when customers happen to write.

Outcome. This one cannot be derived, because it is a judgement about what happened. The pattern that works is a dropdown — Won, Lost, No Response, Not a Fit — that the agent sets on the conversation row, with everything else auto-filled around it. You can have an AI pass suggest an outcome from the transcript, and that is genuinely useful at volume, but keep a human confirming it. Our guide to grounding AI on your own data covers making those classifications reliable rather than plausible.

With the Conversations tab in place, the dashboard is formulas:

Metric	Formula
Leads this month	=COUNTIFS(Conversations!B:B, ">="&EOMONTH(TODAY(),-1)+1)
Median first response (mins)	=MEDIAN(FILTER(Conversations!E:E, Conversations!E:E>0))
Win rate	=COUNTIF(Conversations!F:F,"Won") / COUNTA(Conversations!F:F)
Unanswered over 1 hour	=COUNTIFS(Conversations!E:E, ">60")

That last one is the one people actually act on. A daily trigger that emails the list of conversations still unanswered after an hour changes behaviour in a way a monthly dashboard never does.

The constraints worth knowing up front

The 24-hour window. You can reply freely for 24 hours after a customer's last message. Outside it, only approved template messages — which are billed and which do not count as a response for your response-time metric. This shapes your whole follow-up process, not just your reporting.

Never automate the consumer app. Scraping WhatsApp Web or driving the desktop client breaks WhatsApp's terms and gets numbers banned. Losing the channel your customers use is a far larger cost than any dashboard is worth.

Treat transcripts as personal data. A conversation log is personal data under GDPR and similar regimes. Store the fields you report on rather than every message body, keep the sheet inside your Workspace domain rather than link-shared, set a retention period, and be able to delete one person's rows on request.

Meta's pricing moves. The billing model has been revised several times. Check the current rate card — and note that receiving messages and webhooks is the cheap side; business-initiated messaging is what costs.

Where this leads

Once conversations are in a sheet, the reporting is the small win. The larger one is that WhatsApp stops being a black box that only the person holding the phone can see into. Response times become manageable, leads stop being forgotten in an inbox, and the channel joins the rest of your sales reporting.

From there, the same message stream drives the rest: auto-logging leads into a WhatsApp AI CRM in Google Sheets, or automating the replies themselves as covered in automating WhatsApp sales with Google Sheets. If you are still choosing an approach, our comparison of the three ways to automate WhatsApp lays out what each really costs.

The reporting pipeline itself costs nothing beyond Meta's message rates — Apps Script is free, and the sheet is one you already own. The work is in defining the metrics honestly and handling the edge cases: business hours, automated replies, returning customers, deletions.

If you want that built against how your team actually works, tell us what you need to see each Monday morning. The feasibility review is free, and if your existing inbox platform already exports what you need, we will tell you to use it.