After your first webinar you suddenly have plenty of numbers, and they are not all equally useful. Good webinar metrics show you where prospects drop away and where an improvement will pay off most. This guide explains the key metrics with their formulas, works through a complete example and points out the figures that look impressive but tell you very little.
Start with the question
Metrics are answers, so you need the question first. For a sales webinar it is usually this: of the people who see my registration page, how many end up buying, and where do the others get stuck? A training webinar asks something different, such as how many people stay to the end.
Picture your attendees’ journey as a series of steps. You lose some people at every step, and each metric describes one of them. Once you know which step is weakest, you know where to work. The guide to structuring a sales webinar covers what happens in the content at each stage.
The key metrics along the way
| Metric | Formula | What it tells you |
|---|---|---|
| Registration rate | Registrations ÷ registration page visitors | How convincing the page and promise are |
| Confirmation rate | Confirmed registrations ÷ registrations | Whether the confirmation email arrives and makes sense |
| Attendance rate | Attendees ÷ confirmed registrations | How well the timing and reminders work |
| Retention to the offer | Attendees present at the offer ÷ attendees | Whether the content holds people until the pitch |
| Average watch time | Total viewing time ÷ attendees | How long people stay on average |
| Offer click-through rate | Clicks on the offer ÷ attendees present at the offer | How well the offer is presented |
| Conversion rate | Purchases ÷ attendees | How many attendees end up buying |
| Revenue per registrant | Revenue ÷ registrations | What one registration is worth on average |
| Cost per registrant | Ad spend ÷ registrations | What one registration costs you |
Two of these deserve extra attention. Revenue per registrant rolls every step into one number, which makes it ideal for comparing weeks or versions. Cost per registrant matters as soon as you pay for ads. Only the two together tell you whether a campaign pays for itself. Bear in mind that revenue is not profit. For an honest view, subtract product costs, payment fees and tax.
A worked example
The figures below are invented purely to show the arithmetic. Say you sell an online course for €290 and ran several sessions over a month.
| Step | Count | Metric |
|---|---|---|
| Registration page visitors | 2,000 | |
| Registrations | 560 | Registration rate 28% |
| Confirmed registrations | 504 | Confirmation rate 90% |
| Attendees | 202 | Attendance rate 40% |
| Still there at the offer | 121 | Retention to the offer 60% |
| Clicks on the offer | 30 | Click-through rate 25% |
| Purchases | 10 | Conversion rate 5% of attendees |
Revenue comes to 10 × €290 = €2,900. That gives revenue per registrant of €2,900 ÷ 560, roughly €5.18. If you spent €1,400 on ads, each registration cost you €1,400 ÷ 560 = €2.50, and each sale €140.
It gets interesting when you play with the numbers. If the attendance rate rises from 40 to 50 per cent and everything else stays the same, you would have 252 attendees instead of 202. At the same conversion rate, that means around 12 sales instead of 10, without spending another euro on ads. A better reminder plan can do more than a new advert. Here is a second sum. If 36 people click at the offer instead of 30, roughly 30 rather than 25 per cent, and a third of them buy as before, you also reach 12 sales. Both routes lead to the same result but take different amounts of work, and reworking reminders is often quicker than reworking the pitch. Quick sums like this help you pick your next improvement.
Reading the numbers well
Vanity metrics to treat with care
Some numbers feel good but rarely help you decide anything.
- Total registrations without a source. A thousand sign-ups sound great. If half come from a source whose people never show up, they are not worth much.
- Peak concurrent viewers. That captures a single moment. How many are still there for the offer matters more.
- Number of chat messages. A busy chat can mean enthusiasm or confusion. Read the messages instead of counting them.
- Email open rates. Some email apps preload content automatically, which reports opens that never happened. Clicks and actual attendance are more reliable.
Reading small numbers
With small volumes, metrics swing a lot. In the example above, a single extra sale would lift the conversion rate from just under 5 to roughly 5.4 per cent. One session with three sales more or fewer tells you very little.
Three habits protect you from jumping to conclusions:
- Group sessions together. Compare months or batches of sessions rather than individual days.
- Change one thing at a time. If you change the registration page, reminders and offer all at once, you won’t know what made the difference.
- Be patient with split tests. When you test two versions against each other, you need enough visitors per version before a difference means anything. Some tools calculate a statistical confidence level for this. Don’t decide after the first session.
Capturing metrics cleanly
A metric is only as good as its definition. Write these down once:
- What counts as an attendee. Anyone who enters the room, or only people who stay for a minimum time? Either is defensible, as long as you stay consistent.
- Whether you count people or visits. Someone who opens the registration page three times is one person but three visits.
- The window for purchases. Some attendees buy days later. A fixed window, say seven days after the session, keeps figures comparable.
- Where attendees come from. Tags in the links you use in ads and posts show you later which source brings the best attendees.
You don’t necessarily need tracking cookies for any of this. Many metrics can be counted in aggregate without building profiles. The article on webinar GDPR basics explains what to watch from a data protection angle, and reporting is one of the criteria for choosing webinar software.
A short weekly routine
Metrics only help if you look at them regularly. Fifteen minutes a week is often enough:
- Copy last week’s figures into a simple spreadsheet.
- Mark the weakest step compared with the previous month.
- Write down one guess as to why it is weak.
- Decide on exactly one change and note the date.
- After a few weeks, check whether that step has moved.
Frequently asked questions
What is a good webinar attendance rate?
There is no universal figure, because topic, audience, timing and reminders all play a big part. Your own baseline is far more useful than someone else’s benchmark. Measure it over a few weeks and then work on improving it.
Which metric should I improve first?
The step where you lose the largest share of people and that you can realistically influence. That is often attendance and retention, since reminders and content are in your hands. A quick calculation like the one in the example shows which change would have the biggest effect.
How long should I measure before changing something?
Long enough that single outliers no longer distort the picture. As a rule of thumb, several sessions or a few weeks. With very few attendees per session, it takes correspondingly longer.
Does webinar length affect the metrics?
Yes, especially retention to the offer and watch time. The article on how long a webinar should be helps you find the right running time.
Final thought
Good webinar metrics show you where your webinar loses people. Define them once, run through the journey from registration page to purchase regularly and change one thing at a time. That is how numbers turn into decisions you can explain.
