Quick! 9 things to do when your website goes viral

I'll cut to the chase, because time is short. Here's what to do.

 

1. Check the time  

I'll explain more later. 

2. Contact your web host  

Let them know what's going on. You want them on your side; making demands can backfire in this situation. I'd say something like:

"I'm getting lots of traffic today. Is there anything you can do to keep the website stable until this passes?"

3. Check your analytics to understand where the traffic is coming from

If you have a relationship with the source of the traffic, it can help to get in touch with them. When things went viral at the University of Oxford, the traffic was coming from another department's website - the Wildlife Conservation Research Unit. When we contacted them they explained the connection with the Jimmy Kimmel Show.

4. Establish which page the traffic is landing on 

We'll come back to this later on.

5. Check the time again.  How long has it been since discovery of the traffic surge?  

If you work for an organisation you need to let your colleagues know at some point. This gives and also takes away. Other people can spot opportunities that come with a traffic surge. But, as other people learn of events they will come to you to ask questions. The extra interruptions can cause your practical actions to grind to a halt.

My advice is this: tell someone else, then ask them to handle communications with the rest of the organisation. 

6. Optimise the user journey

If you have clear answers to steps 3 and 4, now is the time to make the most of this event. Think of the mindset of these people -  what are they after? Is there anything you can promote that they'd be interested in? You could edit the page to link to that.

7. Improve the entry page

Sometimes traffic comes into a very old page. Check: is it up-to-date and accurate? I've seen traffic peaks hit pages more than 10 years old.

8.  Prepare for an outage

Your website might go down at any moment. If it does, what would users see? Now is a good time to optimise that message. You can help visitors fulfil their tasks in other ways.

For example, at Oxford we provided a phone number people could call to make donations. Our website did go down, and we still received gifts.

9. Take a break

These sort of events are wild. It's useful to take a short break to let the adrenaline subside and marshal your thoughts. You'll make better decisions afterwards.

Stepping back

When you go viral, it's a shock. It happens so rarely you're not clear what to do.  Hopefully this post will help you make the most of the moment.

Go easy on yourself. An event like this causes us to make hurried decisions. It's understandable if we make some mistakes.

More posts

Can Google Analytics give an early warning of going viral?

The danger of the Realtime Overview

Is Organic Social traffic healthy?

If you glance through the Acquisition Report of Google Analytics your eye might be caught by the phrase 'organic social'. What, pray tell, does that mean?

The quick answer

Users who came from Organic Social sources are those from typical posts on social media channels. 

The detailed answer

Imagine the following Google Analytics scenario. 

Cambridge City Council are thinking about reorganising the way they work. They want to ask residents for their views. They create a survey on their website and post about it on X, Facebook and Instagram.

Daniel sees the post on X. He clicks through to the survey, but then gets distracted and never fills it in.

Next month a member of Cambridge City Council staff checks their Google Analytics (GA) account. Daniel's visit is listed in the Acquisition Report, on the Organic Social row.

As with the Organic Search example, when you see 'organic' think natural, or normal. And social refers to social media, such as X, Facebook, LinkedIn. 

This is not paid traffic: in the example above, Daniel didn't click on an advert on X.

One word of caution... as discussed in a previous post, occasionally traffic from social media gets categorised in GA in the Referral row. 

More Google Analytics posts

What does 'Email' mean in Google Analytics, and why are those numbers so small?

The danger of the Realtime Overview

Can Google Analytics help us measure user satisfaction?

Do people like my website? Is it valuable? I ask these questions regularly, and I imagine they're on the mind of many website owners. 

Sometimes I think about this because I want to serve my audience well. Other times, I ask those questions because I need a website to be trusted in order for it to fulfil its purpose.  

Either way, user satisfaction is on my agenda. Can Google Analytics help me to measure it?

Here are some metrics that may provide the solution:

Average engagement time per active user

I think if someone appreciates a website they spend longer on it. And so average engagement time per active user seems a logical choice.

The number is unaffected by seasonal changes, making it a good choice if you don't have enough past-data for year-on-year analysis.

Take care though: as discussed previously, average engagement time per active user can mislead you if your website is focussed on quick interactions, or if your calls-to-action lead off-site.

This metric is found in Google Analytics (GA) in Reports -> Engagement -> Overview

Returning users

This measure is the number of people who come back a second time in the specified date range. Be aware that if you expand the date range in GA then returning users grows for two reasons: 

• you're catching multiple sessions from new users
• you're catching additional sessions from existing users, which redefines them as 'returning'.  

I like returning users because it's simple: we all understand the concept of repeat customers.

It does have drawbacks though. One is that GA only gives this number rounded to the nearest 100. That might frustrate you if you run a small website. 

Another drawback of returning users is the link to the date range. Imagine you have a website where you publish new content once a month. Many users will get used to the pace, and will only visit once a month. If you check this metric once a month those people will never get counted as returning users. 

Perhaps the answer is to use this measure with an eye to the content schedule of your website. Report on it for a time period significantly greater than the interval between changes. In the example above, maybe check returning users once a quarter.

Bear in mind that returning users does fluctuate with the time of year. 

This metric is found in Reports -> Retention

Active users

I sometimes forget that the most prominent metric in GA, active users, has audience satisfaction built-in. 

Google defines active users as users who:

visit for more than 10 seconds

or

view 2 or more pages

or

trigger a conversion event

or

make a first visit

Now, a person who stays for over 10 seconds, or visits several pages, is indicating some degree of satisfaction with the website. They certainly didn't "come, see, and puke" as Avinash Kaushik used to say. 

When we measure active users we're measuring some element of user satisfaction. So should we just track that number? I'm uneasy with that idea. I like my metrics to be more targeted. 

WAU/MAU

Catchy name, eh? The longer version is: Weekly Active Users / Monthly Active Users. It's expressed as a percentage. This metric is about how many people who visit each month also visit each week.

I find WAU/MAU a bit complicated, which puts me off using it. I don't want to have to remind myself what a metric means each time I check my analytics.

I'm sceptical of the value of this metric because it hinges so much on the frequency with which you update your website. You might get a low number because your website changes once a month.

As in the case of returning users, you'll find WAU/MAU does vary with the time of year.

This measure is found in Reports -> Engagement -> Overview

Websites designed for infrequent use

Sometimes people don't return to a website because the driver for visits occurs rarely. An example of this is the UK website for passport applications. However brilliant the user experience is, most people visit once every 5 or 10 years. 

Another example of a rarely-used website would be one that sells new cars. Do many customers buy a new car every quarter?

A context like these would significantly change the metrics you choose for user satisfaction.

Final thoughts

So, what do I use? I like returning users, for its simplicity.

Do you have views on the metrics I mentioned? Tell me more on Bluesky or Threads. 

More Google Analytics posts

Can Google Analytics give an early warning of going viral?

We don't talk enough about the seasons