Myth session duration

“The length of stay is important.” Or: “A high session duration is good.” I hear such statements very regularly when it comes to online marketing, SEO or the quality of websites. Blog posts are published that reveal tips and tricks to increase the length of stay. Monologue webinars are held in which the session duration is described as one of the key indicators of a website. I am very surprised by such statements. And I am surprised that the vast majority of speakers repeat things that have already been expressed by an infinite number of voices in an infinite number of places. For me, the session duration is a myth, which I will question in the following.

The length of stay in real life

To say it right at the beginning: length of stay can be good, but it does not have to be. I would like to illustrate this with two examples from “real life”.

Imagine you are standing at a box office to buy a ticket for a film. How long are you prepared to wait in line? Two or three minutes? Five or maybe even 10 minutes? You probably don’t mind waiting “a few” minutes for your ticket, especially if you have been looking forward to the latest film by X with Y for quite a while. I’ll save myself the question of whether you’d rather wait 5 minutes or 2 minutes until the cashier asks you which film you want to see.

But: Now you are sitting in the cinema with your ticket. What happens next? You “may” watch commercials. In many cinemas in Berlin 25 minutes, in some cases even up to 35 minutes. 25 to 35 minutes, which many people would gladly do without. Finally, your film begins. The length of cinema films usually varies between approx. 90 and 120 minutes. I often have the impression that films would be better if they were shorter; but that is of course a matter of taste.

Conclusion: Although the length of time spent in the cinema is tolerated by many visitors, it is certainly NOT a quality factor. I will return to this idea in the online world.

Second example: Imagine you want to go out for dinner in a restaurant. What about the length of stay there? Is it perhaps a good signal for satisfaction and quality there? Actually, it depends on the concrete situation: If you have to wait five minutes for the waiter to bring the menu, if you wait over half an hour for your starter, if it takes forever to order a second drink or if you repeatedly ask the waiter for the bill – all these points increase the length of stay but not the quality of your visit.¹

Fortunately, there are other examples: excellent service, great atmosphere, delicious food, creative dessert – in such a case, the length of stay is indeed an indicator of a good time. In such an environment, guests are happy to stay a little longer, chat with each other and order one last drink. Interestingly, this is different in Germany than in the USA, for example. There, guests receive the bill immediately after the last course without being asked. Obviously, the length of stay is less important to restaurant operators than the option of offering the table to other guests and generating more turnover in comparison. I will also return to this idea in a moment.

The session duration in the online world

And what is the situation in the online world now? How can the examples be transferred and what are the results? Actually it is relatively simple:

• Are you prepared to wait online for 5 minutes before you can buy a ticket (to make it even clearer: you don’t have the ticket yet, the ordering process is just beginning!) No, of course not.² So is a long wait when buying a ticket a good quality mark? No, of course not. Would it be better if you bought your ticket after one instead of after five minutes? Yes, of course. Does this mean that a short retention time is better than a long retention time? Yes, of course.
• Similar to a cinema film, advertising messages are often placed on the internet before you can watch a video of your choice.³ Often even two commercials in a row. Maybe you try to skip the rest of a commercial like me. Maybe you are annoyed by the advertisement or simply ignore it. Is the extended length of stay a sign of quality? No, of course not.
• Is a 120 minute film better than a film that only lasts 90 minutes? Of course not. Is a four-minute video better than a video that is only three minutes long? Of course not. The pure duration of a film and/or video says nothing about the quality. It may be possible to make a quality statement by the ratio of the playing time to the total duration of a video, but even that is difficult in my opinion.⁴

I’ll cut a bit short at this point, because even a visit to a restaurant can – at least in part – be transferred to the online world in a similar way. A website that would take five minutes to provide essential information (the menu in a restaurant) would have to expect a high drop-out rate. The same applies if you trigger an action on a website (e.g. downloading a white paper or checking out in a webshop) and nothing happens for the next 30 minutes (online probably rather 1-3 seconds) etc.⁵

Measuring the time spent and the interest of the visitors

I would like to give you two more arguments to challenge the myth of session duration:

1. How do you actually measure the session duration?
2. And what interest do visitors to a website have?

Measuring the retention time sounds pretty simple, doesn’t it? A visitor comes to your website, she stays for a while (reads, looks, clicks etc.) and she leaves your website. All you have to do with the tool of your choice is measure the duration. And what happens if the visitor comes, calls up a specific page, and then receives a phone call or goes for a coffee, meaning she does not perform any action on your site? Should the duration still be calculated? Not if Google Analytics is involved. This widely used tool often shows average session duration of 0 seconds. A pity in fact.

The same applies, by the way, if the cookie content has not been confirmed and the website offers the possibility of being able to read the content anyway (even if this is most likely not GDPR-compliant). The session duration remains at 0 seconds.

Methodically, there are at least two ways to deal with this:

• Some website operators require visitors to actively click, for example, to continue reading an article. The click corresponds to an action, the length of stay is calculated and increases according to Google Analytics. Unfortunately, visitors are forced to take an action without any actual value.
• Other website operators activate a so-called heartbeat. Very simply put, a service sends a ping to the visitor’s browser and as long as the corresponding tab is open, the length of stay is calculated. Sounds good, but it is quite worthless, because many people do not close browser tabs, so the average session duration increases very quickly from minutes to hours.

And what does the visitor of your website actually want? Of course this varies depending on what you offer. I’m sure if you started a survey among your visitors you would hear a lot, but “I want to stay on your site for a long time to …” nobody would say.

Conclusion

For me, a high session duration as a quality feature of a website is a myth. Of course, the length of stay can be increased in real terms – e.g. by integrating videos or using info graphics – regardless of whether and how precisely it can be measured. And of course it can make sense to look at the pages without a stated retention time in order to integrate meaningful interactions for the visitors. But otherwise you should not try to increase the session duration in order to increase it or to achieve higher values in a tool. Instead, I would recommend that you adopt the view of your visitors. What does your visitor want from you and your website? How can you answer this interest in a way that makes the most sense to your visitors. If this increases the length of stay – fine. If this turns the visitor into a returning visitor⁶ or even a customer – much nicer!

Notes:

[1] Even though there is now a trend towards low food, it is certainly no coincidence that fast food chains specialising in the fast delivery of food continue to boom. The name “fast food” says it all: it’s all about speed, and therefore about short and not long duration.

[2] At the Berlinale it is customary for online ticket sales to start at a defined time. Since demand often exceeds supply, many interested parties are willing to wait online for the sales start. However, I find it difficult to evaluate this length of stay positively.

[3] In the meantime, there is also advertising on TV, for example on German TV station DSF. If you switch to this channel, you will always be shown an advertisement first, even if the programme is running normally.

[4] Example: 100 people watch two videos. Video A lasts three minutes, video B four minutes. Video A is watched by 100 people from start to finish. Video B is watched by 50 people for three minutes and by 50 people for four minutes. So the average viewing time is three minutes for Video A and 3:30 minutes for Video B. Video B has a higher display duration and a higher abort rate. Which video is better?

[5] There could be a small difference to a visit to a restaurant: online visitors rarely compete with each other for free seats. Exceptions can be the limited allocation of tickets to online events or webinars, for example if the organiser can only offer a certain number of tickets due to licensing restrictions.

[6] Unfortunately, the quota of returning visitors is also difficult or incomplete to measure. In principle, however, it is desirable for people to visit your website repeatedly, as this is often a sign of good quality.

Michael Schenkel has published additional posts in the t2informatik Blog, including 

What does t2informatik actually do? One click and you'll know.