It's easy to judge a livestream by how many people watched. But viewer count is a vanity metric — a big audience that buys nothing is a bad stream. Here are the numbers that actually matter.

The metrics worth tracking

  • GMV (Gross Merchandise Value). The headline: how much you sold during the session. Everything else explains this number.
  • Conversion rate. Of the people watching, how many bought? Live sessions can hit 5–20% — well above feed averages — so a low rate signals a content or offer problem, not a traffic one.
  • Average order value (AOV). Bundles and upsells during a stream lift this; track whether your live offers actually raise it.
  • Watch time / retention. How long people stay. Drop-offs tell you where the stream loses energy — usually a slow patch between product blocks.
  • Add-to-cart rate. A leading indicator of intent, before the final purchase.
  • Return rate. High sales with high returns isn't a win. Watch this to keep growth healthy.

How to read them together

A stream with high views but low conversion = a traffic/entertainment win but a selling miss (tighten the offers and urgency). Low views but high conversion = your content is selling; now drive more viewers (boosting helps here). The goal is consistent GMV per hour, improved session over session.

Benchmarks to aim for

Numbers only mean something against a reference point. As rough targets for a healthy live-selling operation: conversion of 5–20% of live viewers (versus roughly 1–3% for ordinary feed content), steady or rising average watch time session over session, and an add-to-cart rate that comfortably exceeds your final purchase rate (a signal your offers create intent even when checkout lags). If your conversion sits below that band despite decent viewership, the problem is almost never traffic — it's the offer, the urgency, or the host, all of which you can fix.

A worked example: reading one session honestly

A brand runs a two-hour TikTok Shop Live: 8,000 unique viewers, 6% conversion, RM18,000 GMV, but a 30% return rate the following week. On the surface it's a win. Read it properly and two things jump out. First, conversion is solid — the selling worked. Second, the return rate is quietly eating the profit, which usually points to over-promising on camera or unclear sizing. The fix isn't "get more viewers"; it's tightening product claims and expectations mid-stream. Same data, very different action than a vanity-metric reading would suggest.

Common mistakes: chasing the wrong number

  • Celebrating peak viewers. A spike that doesn't convert is entertainment, not commerce.
  • Ignoring returns. GMV net of returns is the only revenue that counts.
  • No session-over-session tracking. Without a trend line you can't tell whether changes are working; you're just guessing from one noisy data point.

How SushiVid helps

With our Live Streaming Service, you track GMV and viewership performance across your hosted sessions — so you're optimising the numbers that move revenue, not chasing view counts. Need more viewers on a converting stream? Our Ad Boosting can boost Shopee Live and TikTok Shop Live sessions to drive viewership.

The takeaway

Measure GMV, conversion, AOV, retention and returns — not applause. Those are the numbers that tell you whether live selling is actually working, and where to improve.

Turning metrics into a weekly improvement loop

Numbers are only useful if they change what you do next week. Build a simple loop: after each session, log GMV (net of returns), conversion, average watch time and add-to-cart; find the single weakest number; form one hypothesis about why; and change exactly one thing next session to test it. Low watch time? Tighten the gaps between product blocks. Strong add-to-cart but weak conversion? Your checkout offer or urgency needs work. High returns? Fix the on-camera claims. Changing one variable at a time is what lets you attribute improvement to a cause rather than guessing — and over a quarter, those small, evidenced adjustments compound into a materially better GMV-per-hour.

Common questions

What's a "good" live conversion rate? As a rough band, 5–20% of live viewers, versus roughly 1–3% for ordinary feed content. Below that band with healthy viewership points to an offer or hosting problem, not a traffic one.

Why track returns? Because GMV net of returns is the only revenue that counts. A high-sales, high-return stream usually means the on-camera claims outran the product — a fixable messaging issue.

How many sessions before the data means anything? Look at trends across weeks, not single streams. One session is a noisy data point; a trend line tells you whether your changes are working.

SushiVid's own proof: in a SushiVid-run live campaign for Taiwan Select, live sessions drove over 4,000 product link clicks — the kind of action-based metric that matters far more than raw viewer count. (Case study)

Want live sessions measured on revenue, not applause? SushiVid's Live Streaming Service tracks GMV and viewership across your hosted sessions so you optimise what actually sells. Talk to us about live selling →

Sources: TikTok Shop & Shopee GMV tracker SEA 2026 — DigitalInAsia; TikTok Shop statistics 2026 — Marketing LTB.